Amd a6 4455m gaming: AMD A-Series A6-4455M Notebook Processor

AMD A6-4455M Specs, Price, and FPS in Games


2 2 2.1 GHz 2.6 GHz
Cores Threads Base Frequency Turbo Frequency

Geekbench 5 Single-Core

254

9%

Geekbench 5 Multi-Core

386

1%

Availability

No items available

Graphics Card

Resolution

Select game resolution

Graphics Settings

Select game graphics

Offset

Apply Offset

Value Rating

Performance Rating

Value per FPS

You will receive

… FPS

432

FPS

Counter-Strike: Global Offensive

718

FPS

League of Legends

174

FPS

Grand Theft Auto V

305

FPS

VALORANT

184

FPS

Apex Legends

968

FPS

Minecraft

Alternatives for A6-4455M

1080p, High

No alternatives — this is the best option

Specifications

General
Release Date May 15th, 2012
Segment Mobile
Socket AMD Socket FP2
Collection A6
Codename Trinity
Performance
Cores 2
Threads 2
Base Frequency 2. 1 GHz
Turbo Frequency 2.6 GHz
Other
Power Consumption 17 W
Overclockable No
Integrated Graphics Radeon HD 7500G

Find out which of the 2 CPUs performs better, view a side-by-side specification comparison.

Ryzen 5 5600X

$193.89

A6-4455M

N/A Stock

Ryzen 7 5800X

$269.99

A6-4455M

N/A Stock

Core i9-12900K

$537.88

A6-4455M

N/A Stock

Core i9-9900K

A6-4455M

N/A Stock

Core i7-8700K

A6-4455M

N/A Stock

Core i7-12700K

$398.13

A6-4455M

N/A Stock

Refine results to filter 1558 processors by performance, release date, price, and value. Click on a CPU to view more in-depth specifications and game FPS.

Found 1558 CPUs.

Popularity

Processor

Performance

Ryzen 5 5600X

Release Date: Nov 5th, 2020


Performance Rating


$193. 89 on Amazon

In Stock


Ryzen 9 5900X

Release Date: Nov 5th, 2020


Performance Rating


$398.99 on Amazon

In Stock


Ryzen 7 5800X

Release Date: Nov 5th, 2020


Performance Rating


$269.99 on Amazon

In Stock


Ryzen 5 5600G

Release Date: Apr 13th, 2021


Performance Rating


$151.99 on Amazon

In Stock


Core i5-12600K

Release Date: Nov 4th, 2021


Performance Rating


$277.99 on Amazon

In Stock


Core i9-12900K

Release Date: Nov 4th, 2021


Performance Rating


$537.88 on Amazon

In Stock


Core i7-10700K

Release Date: Apr 30th, 2020


Performance Rating


$265.99 on Amazon

In Stock


Ryzen 5 3600

Release Date: Jul 7th, 2019


Performance Rating


$137.99 on Amazon

In Stock


Ryzen 7 5700G

Release Date: Apr 13th, 2021


Performance Rating


$207. 99 on Amazon

In Stock


Core i7-12700K

Release Date: Nov 4th, 2021


Performance Rating


$398.13 on Amazon

In Stock


Core i9-11900K

Release Date: Mar 16th, 2021


Performance Rating


$329.99 on Amazon

In Stock


Core i7-12700KF

Release Date: Nov 4th, 2021


Performance Rating


$355.48 on Amazon

In Stock


Ryzen 9 5950X

Release Date: Nov 5th, 2020


Performance Rating


$549 on Amazon

In Stock


Core i9-12900KF

Release Date: Nov 4th, 2021


Performance Rating


$525.99 on Amazon

In Stock


Core i9-10900K

Release Date: Apr 30th, 2020


Performance Rating


$349 on Amazon

In Stock


Core i7-11700K

Release Date: Mar 16th, 2021


Performance Rating


$269.99 on Amazon

In Stock


Core i5-10400F

Release Date: Apr 30th, 2020


Performance Rating


$133 on Amazon

In Stock


Ryzen 9 3900X

Release Date: Jul 7th, 2019


Performance Rating


$345 on Amazon

In Stock


Ryzen 7 3700X

Release Date: Jul 7th, 2019


Performance Rating


$260 on Amazon

In Stock


Core i7-11700KF

Release Date: Mar 16th, 2021


Performance Rating


$268. 36 on Amazon

In Stock


Ryzen 3 3200G

Release Date: Jul 7th, 2019


Performance Rating


$119.74 on Amazon

In Stock


Core i5-11600KF

Release Date: Mar 16th, 2021


Performance Rating


$209 on Amazon

In Stock


Core i7-9700K

Release Date: Oct 19th, 2018


Performance Rating


$435 on Amazon

In Stock


Core i3-10100

Release Date: Apr 30th, 2020


Performance Rating


$105 on Amazon

In Stock


Ryzen 7 3800X

Release Date: Jul 7th, 2019


Performance Rating


$257.99 on Amazon

In Stock


  • Previous
  • 1
  • 2
  • 3
  • 4
  • 61
  • 62
  • 63
  • Next

AMD A6-4455M review | 64 facts and highlights

40points

AMD A6-4455M

AMD A6-4455M

Why is AMD A6-4455M better than the average?

  • Thermal Design Power (TDP)?
    17Wvs48. 14W
  • L2 core?
    1MB/corevs0.44MB/core
  • Supported displays?
    4vs3.07
  • Turbo Core version?
    3vs2.4

Which are the most popular comparisons?

AMD A6-4455M

vs

Intel Core i3-2328M

AMD A6-4455M

vs

Intel Core i3-6100U

AMD A6-4455M

vs

AMD A8-4555M

AMD A6-4455M

vs

Intel Core i3-3110M

AMD A6-4455M

vs

Intel Core i5-3337U

AMD A6-4455M

vs

AMD A6-9225

AMD A6-4455M

vs

Intel Core i3-2350M

AMD A6-4455M

vs

Intel Core i7-3612QM

AMD A6-4455M

vs

Intel Core i5-3210M

AMD A6-4455M

vs

Intel Core i3-2348M

Price comparison

User reviews

Performance

1.CPU speed

2 x 2.1GHz

The CPU speed indicates how many processing cycles per second can be executed by a CPU, considering all of its cores (processing units). It is calculated by adding the clock rates of each core or, in the case of multi-core processors employing different microarchitectures, of each group of cores.

2.CPU threads

More threads result in faster performance and better multitasking.

3.turbo clock speed

2.6GHz

When the CPU is running below its limitations, it can boost to a higher clock speed in order to give increased performance.

4.Has an unlocked multiplier

✖AMD A6-4455M

Some processors come with an unlocked multiplier which makes them easy to overclock, allowing you to gain increased performance in games and other apps.

5.L2 cache

A larger L2 cache results in faster CPU and system-wide performance.

6.L3 cache

Unknown. Help us by suggesting a value.

A larger L3 cache results in faster CPU and system-wide performance.

7.L1 cache

A larger L1 cache results in faster CPU and system-wide performance.

8.L2 core

1MB/core

More data can be stored in the L2 cache for access by each core of the CPU.

9.L3 core

Unknown. Help us by suggesting a value.

More data can be stored in the L3 cache for access by each core of the CPU.

Memory

1.RAM speed

1333MHz

It can support faster memory, which will give quicker system performance.

2.maximum memory bandwidth

Unknown. Help us by suggesting a value.

This is the maximum rate that data can be read from or stored into memory.

3.DDR memory version

Unknown. Help us by suggesting a value.

DDR (Double Data Rate) memory is the most common type of RAM. Newer versions of DDR memory support higher maximum speeds and are more energy-efficient.

4.memory channels

More memory channels increases the speed of data transfer between the memory and the CPU.

5.maximum memory amount

Unknown. Help us by suggesting a value.

The maximum amount of memory (RAM) supported.

6.bus transfer rate

Unknown. Help us by suggesting a value.

The bus is responsible for transferring data between different components of a computer or device.

7.Supports ECC memory

✖AMD A6-4455M

Error-correcting code memory can detect and correct data corruption. It is used when is it essential to avoid corruption, such as scientific computing or when running a server.

8.eMMC version

Unknown. Help us by suggesting a value.

A higher version of eMMC allows faster memory interfaces, having a positive effect on the performance of a device. For example, when transferring files from your computer to the internal storage over USB.

9.bus speed

Unknown. Help us by suggesting a value.

The bus is responsible for transferring data between different components of a computer or device.

Benchmarks

1.PassMark result

This benchmark measures the performance of the CPU using multiple threads.

2.PassMark result (single)

This benchmark measures the performance of the CPU using a single thread.

3.Geekbench 5 result (multi)

Unknown. Help us by suggesting a value.

Geekbench 5 is a cross-platform benchmark that measures a processor’s multi-core performance. (Source: Primate Labs, 2022)

4.Cinebench R20 (multi) result

Unknown. Help us by suggesting a value.

Cinebench R20 is a benchmark tool that measures a CPU’s multi-core performance by rendering a 3D scene.

5.Cinebench R20 (single) result

Unknown. Help us by suggesting a value.

Cinebench R20 is a benchmark tool that measures a CPU’s single-core performance by rendering a 3D scene.

6.Geekbench 5 result (single)

Unknown. Help us by suggesting a value.

Geekbench 5 is a cross-platform benchmark that measures a processor’s single-core performance. (Source: Primate Labs, 2022)

7.Blender (bmw27) result

Unknown. Help us by suggesting a value.

The Blender (bmw27) benchmark measures the performance of a processor by rendering a 3D scene. More powerful processors can render the scene in less time.

8.Blender (classroom) result

Unknown. Help us by suggesting a value.

The Blender (classroom) benchmark measures the performance of a processor by rendering a 3D scene. More powerful processors can render the scene in less time.

9.performance per watt

This means the CPU is more efficient, giving a greater amount of performance for each watt of power used.

Features

1.uses multithreading

✖AMD A6-4455M

Multithreading technology (such as Intel’s Hyperthreading or AMD’s Simultaneous Multithreading) provides increased performance by splitting each of the processor’s physical cores into virtual cores, also known as threads. This way, each core can run two instruction streams at once.

2.Has AES

✔AMD A6-4455M

AES is used to speed up encryption and decryption.

3.Has AVX

✔AMD A6-4455M

AVX is used to help speed up calculations in multimedia, scientific and financial apps, as well as improving Linux RAID software performance.

4.SSE version

SSE is used to speed up multimedia tasks such as editing an image or adjusting audio volume. Each new version contains new instructions and improvements.

5.Has F16C

✔AMD A6-4455M

F16C is used to speed up tasks such as adjusting the contrast of an image or adjusting volume.

6.bits executed at a time

Unknown. Help us by suggesting a value.

NEON provides acceleration for media processing, such as listening to MP3s.

7.Has MMX

✔AMD A6-4455M

MMX is used to speed up tasks such as adjusting the contrast of an image or adjusting volume.

8.Has TrustZone

✖AMD A6-4455M

A technology integrated into the processor to secure the device for use with features such as mobile payments and streaming video using digital rights management (DRM).

9.front-end width

Unknown. Help us by suggesting a value.

The CPU can decode more instructions per clock (IPC), meaning that the CPU performs better

Price comparison

Cancel

Which are the best CPUs?

Notebook Amd A6 4455m | MercadoLivre ?

Ordenar por

  1. Notebook Dell 15 3000 Amd R7 8gb 256gb Ssd W11 Preto

    por Dell

    em

    10x

    399 reais con 90 centavos R$399,90

    sem juros

    Patrocinado

  2. Chromebook Lenovo Ideapad 3 Amd Ddr4 32gb Ssd Original Nf

    em

    10x

    147 reais con 81 centavos R$147,81

    sem juros

    Frete grátis

  3. Processador Amd A-series A6-5350m 2.

    9ghz Fs1 — Am5350dec23hl

    em

    12x

    19 reais con 20 centavos R$19,20

    Frete grátis

  1. Processador Amd A-series A6-5350m 2.9ghz Fs1 — Am5350dec23hl

    em

    12x

    19 reais con 20 centavos R$19,20

    Frete grátis

  2. RECOMENDADO

    Notebook Lenovo IdeaPad S145 prateada 15.6″, AMD Ryzen 5 3500U 8GB de RAM 256GB SSD, AMD Radeon RX Vega 8 1366x768px Windows 10

    em

    10x

    369 reais con 99 centavos R$369,99

    sem juros

    Frete grátis

    Disponível em 2 cores

  3. Memória RAM Impact DDR4 color preto 16GB 1 HyperX HX426S15IB2/16

    em

    12x

    38 reais con 30 centavos R$38,30

    Frete grátis

  1. Pc Gamer Amd A6 7480 8gb Ram Ssd 240gb Gpu Radeon R5

    em

    10x

    175 reaisR$175

    sem juros

    Frete grátis

  2. Pc Gamer Completo Amd A6 3.

    8ghz 8gb Radeon R5 Ssd 480gb

    por 2Eletro

    em

    10x

    266 reais con 31 centavos R$266,31

    sem juros

    Frete grátis

  3. Processador Amd A6 4400m 2.7ghz Turbo 3.2ghz Oem

    em

    12x

    14 reais con 05 centavos R$14,05

    Frete grátis

  1. Memória RAM Impact color preto 8GB 1 HyperX HX432S20IB2/8

    em

    12x

    22 reais con 30 centavos R$22,30

    Frete grátis

  2. Notebook Dell 15 3000 Amd R5 8gb 256gb Ssd W11 + Mochila

    por Dell

    em

    10x

    374 reais con 90 centavos R$374,90

    sem juros

    Patrocinado

  3. Notebook Lenovo Ideapad 3 Amd Ryzen 7 8gb 256gb W11 15,6

    por 123 Comprou

    em

    10x

    339 reais con 90 centavos R$339,90

    sem juros

    Frete grátis

    Patrocinado

  1. Memória RAM Vengeance color preto 16GB 1 Corsair CMSX16GX4M1A2666C18

    em

    12x

    37 reais con 71 centavos R$37,71

    Frete grátis

  2. Pc Cpu Gamer Barato Amd, 8gb, Hd 500, R5 2gb Volta As Aulas

    em

    12x

    155 reais con 02 centavos R$155,02

    Frete grátis

  3. Kit Upgrade — Cpu A6 + Placa Mae + Placa Video 2gb + 8gb Ram

    em

    12x

    95 reais con 94 centavos R$95,94

    Frete grátis

  1. Notebook Hp Amd A6 — Hd500gb — Funcionando E Barato

    em

    10x

    118 reais con 92 centavos R$118,92

    sem juros

    Frete grátis

    Recondicionado

  2. Memória RAM Impact color preto 8GB 1 HyperX HX316LS9IB/8

    em

    12x

    12 reais con 70 centavos R$12,70

    Frete grátis

  3. RECOMENDADO

    Notebook Acer Aspire 5 A515-45 prata 15.

    6″, AMD Ryzen 7 5700U 8GB de RAM 256GB SSD, AMD Radeon RX Vega 8 (Ryzen 4000) 60 Hz 1920x1080px Windows 11 Home

    por Acer Brasil

    em

    12x

    316 reais con 93 centavos R$316,93

    Frete grátis

  1. Pc Gamer Facil Amd A6 3.8ghz 16gb Radeon R5 Ssd 240gb — 500w

    por 2Eletro

    em

    10x

    224 reais con 90 centavos R$224,90

    sem juros

    Frete grátis

  2. Cpu Gamer Amd A6 7480 3.8ghz, 8gb, 8 Núcleos! Promoção!!

    por Imperiums

    em

    10x

    161 reais con 91 centavos R$161,91

    sem juros

    Frete grátis

  3. Placa Mae Hp Probook 6465b A4 A6 Series Amd 658544-001 Fs1

    em

    12x

    33 reais con 83 centavos R$33,83

    Frete grátis

  1. Notebook Lenovo Ideapad 3 Ryzen 5-5500u 8gb 256gb W11 15,6»

    por 123 Comprou

    em

    10x

    359 reais con 90 centavos R$359,90

    sem juros

    Frete grátis

    Patrocinado

  2. Tela Frontal Lcd Touch Galaxy Tab A6 7.

    0 T285 3g 4g T 285

    em

    12x

    22 reais con 57 centavos R$22,57

    Frete grátis

    Patrocinado

  3. Memória RAM color verde 4GB 1 Orionas ORI1066S7/4

    em

    6x

    18 reais con 33 centavos R$18,33

    sem juros

    Frete grátis

  1. MAIS VENDIDO

    Notebook Ideapad 3 15.6» Ryzen 7 8gb 256gb Ssd W11 Lenovo

    por Mercado Livre

    em

    10x

    351 reais con 56 centavos R$351,56

    sem juros

    Frete grátis

  2. Computador Pc Gamer Amd A6 9500 3.8 Turbo 8gb Ddr4 Ssd 240gb

    em

    10x

    172 reais con 90 centavos R$172,90

    sem juros

    Frete grátis

  3. Pc Gamer A6 7480 Hdmi Wifi 8gb + 30 Jogos!

    por Imperiums

    em

    10x

    289 reais con 90 centavos R$289,90

    sem juros

    Frete grátis

  1. Notebook Hp Amd A6 4gb Ram Ssd 240gb Windows 10 Pró

    em

    10x

    205 reaisR$205

    sem juros

    Frete grátis

    Recondicionado

  2. Notebook Probook Amd A6 4gb Ram Ssd 240gb Windows 10 Pró !!

    em

    10x

    195 reaisR$195

    sem juros

    Frete grátis

    Recondicionado

  3. Memória RAM Fury Impact DDR4 color preto 8GB 1 Kingston KF432S20IB/8

    em

    12x

    23 reais con 07 centavos R$23,07

    Frete grátis

  1. Memória RAM Impact DDR4 color preto 32GB 1 HyperX HX432S20IB/32

    em

    10x

    75 reais con 60 centavos R$75,60

    sem juros

    Frete grátis

  2. Pc Cpu Completo Mancer, Amd A6, 8gb De Ram, Ssd 240gb

    por Pichau

    em

    10x

    271 reais con 98 centavos R$271,98

    sem juros

    Frete grátis

  3. Pc Computador Barato Placa Video 2gb / 8gb Ram / Cpu 3.

    8ghz

    em

    10x

    170 reais con 95 centavos R$170,95

    sem juros

    Frete grátis

  1. Notebook Probook Amd A6 4gb Ram Ssd 120gb Windows 10 Pró !!

    em

    10x

    175 reaisR$175

    sem juros

    Frete grátis

    Recondicionado

  2. Notebook Hp Amd A6 8gb Ram 480gb De Ssd Windows 10 Pró Oem

    em

    10x

    255 reaisR$255

    sem juros

    Frete grátis

    Recondicionado

  3. Notebook Acer Aspire 5 A515-45 prata 15.6″, AMD Ryzen 3 5300U 8GB de RAM 256GB SSD, AMD Radeon RX Vega 6 60 Hz 1920x1080px Windows 11 Home

    por Mobcom

    em

    10x

    332 reais con 25 centavos R$332,25

    sem juros

    Frete grátis

  1. Memória RAM Fury Impact DDR4 color preto 8GB 1 Kingston KF426S15IB/8

    em

    12x

    20 reais con 36 centavos R$20,36

    Frete grátis

  2. Computador Pc Gamer Amd A6 9500 3.

    8 Turbo 8gb Ddr4 Ssd 480gb

    em

    10x

    189 reais con 90 centavos R$189,90

    sem juros

    Frete grátis

  3. Pc Gamer Amd A6 9500 8gb Ddr4 Radeon R5 2gb Hd 500gb

    por 2Eletro

    em

    10x

    299 reais con 90 centavos R$299,90

    sem juros

    Frete grátis

  1. Notebook Hp Probook Amd A6 8gb Ram Hd 500gb Windows 10 Pró

    em

    10x

    195 reaisR$195

    sem juros

    Frete grátis

    Recondicionado

  2. Memória RAM Vengeance color preto 32GB 2 Corsair CMSX32GX4M2A3200C22

    em

    12x

    96 reais con 95 centavos R$96,95

    Frete grátis

  3. Memória RAM Fury Impact DDR4 color preto 16GB 1 Kingston KF426S15IB1/16

    em

    10x

    41 reais con 30 centavos R$41,30

    sem juros

    Frete grátis

  1. Notebook Hp Amd A6 8gb Ram 240gb De Ssd Windows 10 Pró Oem

    em

    10x

    225 reaisR$225

    sem juros

    Frete grátis

    Recondicionado

  2. Pc Completo Gamer A6 7480 3.

    8ghz, Wi-fi! Black Friday!

    por Imperiums

    em

    10x

    236 reais con 71 centavos R$236,71

    sem juros

    Frete grátis

  3. Cpu Pc Gamer Barato Computador Placa De Video 2gb / 16gb Ram

    em

    12x

    181 reais con 98 centavos R$181,98

    Frete grátis

  1. Notebook Hp Amd A6 8gb Ram 120gb De Ssd Windows 10 Pró Oem

    em

    10x

    205 reaisR$205

    sem juros

    Frete grátis

    Recondicionado

  2. Notebook gamer Acer Aspire Nitro 5 AN515-44 preta e vermelha 15.6″, AMD Ryzen 7 4800H 8GB de RAM 512GB SSD, NVIDIA GeForce GTX 1650 144 Hz 1920x1080px Linux Endless

    por Mobcom

    em

    10x

    543 reais con 90 centavos R$543,90

    sem juros

    Frete grátis

  3. Memória RAM Premier color verde 16GB 1 Adata AD4S2666316G19-S

    em

    12x

    41 reais con 20 centavos R$41,20

    Frete grátis

  1. Cpu Pc Gamer Barato Computador Placa De Video 2gb + Ram 16gb

    em

    10x

    188 reais con 66 centavos R$188,66

    sem juros

    Frete grátis

  2. Pc Completo Gamer Monitor 19.

    5 Led Hdmi Wifi 16gb + 30 Jogos

    por Imperiums

    em

    10x

    311 reais con 66 centavos R$311,66

    sem juros

    Frete grátis

  3. Notebook Hp Probook Amd A6 8gb Hd 1tb Windows 10 Pró Oem!

    em

    10x

    204 reais con 70 centavos R$204,70

    sem juros

    Frete grátis

    Recondicionado

  1. Memória RAM Fury Impact DDR4 color preto 16GB 1 Kingston KF432S20IB1/16

    em

    12x

    45 reais con 57 centavos R$45,57

    Frete grátis

  2. Notebook gamer Dell G15 Ryzen Edition preta 15.6″, AMD Ryzen 7 5800H 16GB de RAM 512GB SSD, NVIDIA GeForce RTX 3060 1920x1080px Windows 11 Home

    em

    10x

    996 reaisR$996

    sem juros

    Frete grátis

    Disponível em 3 cores

  3. Protetor De Cabo Carregador Macbook Magsafe 1 — Magsafe 1

    em

    2x

    5 reais con 44 centavos R$5,44

    sem juros

    Patrocinado

O frete grátis está sujeito ao peso, preço e distância do envio.

AMD APU for the FM2 platform

The entire current line in one article and its comparison with its predecessors and competitors

We already got acquainted with the results of the top APU for FM2 immediately after the announcement of this platform, but the range of Trinity desktops is not limited to the A10-5800K alone. Moreover, although this model is relatively inexpensive, the platform itself is aimed at the budget market, so many potential buyers are more interested in the results of even cheaper solutions. So, you need to test them too. By tradition, we will carry out this process in two stages: the performance of the graphics component will be considered next time, and today we will study the processor part of all new APUs using the «standard» method on a stand with a discrete video card. 3.2/3.7 3.4/4.0 Number of compute cores/threads 2/2 2/2 4/4 cache l1 (sums. ), KB, KB 64/32 64/32 128/64 128/64 cache L2, KB 1024 1024

1024

1024

1024

1024

1024

1024 1024

2×2048 2×2048 RAM 2×DDR3-1600 2×DDR3-1866 $108(60)

As with the FM1, most FM2 desktop APUs fit into a 65W TDP. But there are nuances: such models of all families are available for the new platform in retail, and under FM1 it was easy to buy only A4 or the youngest tri-core low-frequency A6-3500. Quad-core processors (such as A6-3600/3620 or A8-3800/3820) were bought up «on the vine» by large manufacturers, so retail stores almost didn’t get it. Now everything is simple: A4, A6, A8, and even A10 are available in an energy-efficient version.

As for the A6 and A4, there are only two of them in the updated model range, and both are single-module, so they are interesting in themselves: before, «construction equipment» in a similar configuration did not fall into our hands. Both are strictly 65 W, both belong to the budget family. More precisely, A4 has already migrated to the ultra-budget one: a locked multiplier and the lack of Dual Graphics support «lowers» these processors even below the level of old APUs of the same family. But the A6-5400K can offer the economical user a more powerful video core, and even overclocking freedom. In the latter, it surpasses the younger A8 and A10.

Although it is obvious that no matter how much you overclock a single-module processor, four computation threads look more fun in our time. And when buying a platform “for overclocking”, paying attention to the official heat pack is a dubious occupation. Therefore, for users demanding on the processor component, there are 5600K and 5800K. Minus — 100 watts. Plus — judging by the frequencies, the A8-5600K will often outperform the A10-5700 in terms of processor performance. The GPU is weaker, but even here the already studied A10-5800K looks more interesting, where the frequency of the integrated graphics is higher. However, we repeat, we will deal with the study of graphics performance later. So far — only the CPU, that is, to some extent, we are testing not A4/A6/A8/A10, but Athlon X2 and X4. And here the most interesting will be not only the comparison of 65 vs. 100W, but also how the new modules compare to the old cores. 4/4 cache L1, I/D, KB 128/128 192/192 256/256 CASH L2, KB 2 × 512

1

3 × 1024 4 × 1024 RAM 2 × DDR3-1600 2 × DDR3-1866 2 × DDR3-1866 9000

6530D Radeon HD 6550D Socket 2/2 2/2 2/4 cache L1, I/D, KB 64/64 64/64 64/64

900 9000

CASH L2, KB 2 × 256 2 × 256 2 × 256 CASH L3, MIB 2 3 999

Frequency UNCORE, GGC 9001

3. 1 3.3 RAM 2×DDR3-1066 9-10-9-28 / 9-9-9-24)

Testing

Traditionally, we divide all tests into a number of groups and show the average result for a group of tests/applications on diagrams (you can see separate article). The results on the diagrams are given in points, the performance of the iXBT.com reference test system of the 2011 sample is taken as 100 points. It is based on the AMD Athlon II X4 620 processor, but the amount of memory (8 GB) and the video card (NVIDIA GeForce GTX 570 1280 MB by Palit) are standard for all tests of the «main line» and can only be changed as part of special studies. Those who are interested in more detailed information are again traditionally invited to download a table in Microsoft Excel format, in which all the results are shown both in converted points and in «natural» form.

Interactive work in 3D packages

These tests are essentially single-threaded (which is why even the Celeron G440 didn’t look so bad here), but the second thread can help them a little, and more is not needed, so the difference between one and two there are no modules. And the benefits of the «classic» cores too — the A6-3500, due to the low clock frequency, was generally equal only to the Celeron G465. Accordingly, within the framework of the APU, new models lined up practically in terms of clock frequencies and, for the most part, overtook their predecessors. Especially revealing is that the all new A8 and A10 overtook the flagship for FM1! Three of them and the older Celeron G555 are quite tough, and only the A8-5500 is inferior not only to it, but also to the older G550, overtaking, however, the G530 and G540.

It would seem — is it a great honor to defeat Celeron? On the one hand, not really. On the other hand, solutions for FM1 were incapable of such things. Even the fastest. But FM2 is doing much better. Moreover, they are superior not only to Celeron, but also to junior Pentiums, which in essence belong to the same level. The elders, of course, are faster, but we are not inclined to criticize “growth above ourselves” even in such non-targeted areas of application.

Final rendering of 3D scenes

Previously, AMD’s «ladder» looked smoother due to the presence of three-core solutions in the company’s assortment (in a sense, inherited from AM2+), but there is a radical difference between one and two modules. Moreover, in this group of tests, two- and four-threaded Intel processors, for obvious reasons, are separated by a much smaller abyss: there are still two cores in both cases, and here some blocks that are very necessary for these tests exist in very different quantities. As a result, the new A4 and A6 not only compete with Celeron «does not shine» — they are comparable only with the previous generation A4. But the older lines, on the contrary, are quite interesting. And even though they fall short of both modern Core i3 and quad-core Llano, it’s okay: at least Pentium is handled without any stress.

Packing and unpacking

Only one test out of four can use more than two computational threads, and two tests even manage with one, which led to a kind of triumph of single-module processors in the budget segment. And no wonder: two high-frequency integer streams, a small but still shared L2 cache (what Llano and Athlon II lacked so much), fast memory support — all this leads to the fact that the A6-5400K not only outperforms any Celeron , but also got very close to the flagships for FM1. And A8/A10 get an additional boost from additional threads (albeit only in one test), and they have twice as much L2 per module (this affects everywhere), so here the rivalry can go seriously with older Pentiums (G870 and A8 -5600K, for example, show equal results), and even the younger Core i3 do not look unattainable.

Audio encoding

Two-module configurations continue to please in this field of activity: of course, they are slower than the “full-fledged” four cores (and even then not all: we recall that 100 points is the junior Athlon II X4 620, and 110 is Core 2 Quad Q8200), but initially no one doubted this — after all, these are somewhat different classes. “Cheap multi-core” was actually promoted by AMD only out of desperation, and it cost the company not at all cheaply, and Intel tried its best not to let quad-core processors into the segment below $150 in order to avoid problems with profitability, preferring games around SMT. So AMD eventually came to the same conclusion. But there is one difference: Intel has to strongly restrain the Core i3 with clock speeds so that their results in low-threaded applications do not interfere with Core i5 sales, while AMD does not have such a problem: two modules is the maximum for FM2. Above — at the moment the patrimony of FX and AM3 +, where positioning has to be handled carefully, and on FM2 hands are still untied. In particular, no one bothers to make «steroidal» modifications with a TDP of 100 W, overtaking any Core i3, as well as some old quad-core models.

But single-module devices, however, have a hard time as a result — their frequencies are almost the same and even lower (for those who have not yet understood why, re-read the text above). Accordingly, in direct accordance with the theory, the performance may turn out to be twice as low, and this is already a level lower than the «real» dual-core models of both Intel and AMD. The best we can say about the A4-5300 is that it’s just a tad faster than the A4-3400 or the Athlon II X2 215, while the A6-5400K manages to beat the A6-3500. In general, it’s also not bad to some extent, but, of course, we can’t talk about competition even with Celeron. More precisely, it can, but, as before, only with the involvement of integrated graphics in the «disassembly», and not at the expense of processor performance.

Compilation

In this group, things are even worse for single-module models, since they begin to lose even to old dual-core models. A comparison with the A6-3500 is even more indicative: only a three-core processor, crushed by frequencies to the level of the plinth, is still one and a half times faster than the new single-modules. It is clear that not only the cores helped here, but also the cache memory, of which (albeit separate) it has a total of 3 MiB, while the A6 / A4-5000 has three times less — compilers do not forgive this — but the overall result is important to us. And, we repeat, it is indicative: if the A6-3500 is suitable for a student, for example, both to study programming and to play in pauses, then the A6-5400K for the same money can only offer to play 🙂 Or still pay extra — then and it will be possible to play with greater comfort, and not only, since dual-module processors look good against the background of “their” competitors. Especially if we consider modifications with a TDP of 100 W: as we see, while the second digit of the model number actually duplicates the name of the subfamily, although it would be more logical to “endow” it with information about the performance of the processor component. That is, to swap 5600K and 5700 — there would be less confusion among buyers, in our opinion. And that the graphics part of the 5700 is more powerful — it’s obvious from the fact that it’s A10: there’s no need to underline this with a number;)

Mathematical and engineering calculations

Another set of tasks, for the most part gravitating towards single-threading, and therefore the “extra” cores are just superfluous here. Accordingly, AMD’s old budget developments could not really outperform even the Celeron G465, but the new ones are quite competitive, since they have results above a hundred points. Of course, now this is the level of only dual-core Celerons, but against the backdrop of the situation in recent years, it is quite tempting to achieve. In addition, more expensive processors are not so far ahead of this class in terms of performance.

Bitmap graphics

As we have already said more than once, this group of programs refers to multithreading in a mixed way, and those who need it perceive various manifestations of SMT (both AMD’s x86 core and Intel’s Hyper-Threading) at least favorably than real kernels. Which, of course, is good for Trinity. Budget modifications of the new APUs are noticeably ahead of their predecessors, and everything is not bad at the maximum level — already 65 W of the A8-5500 is enough to overtake any models for FM1 and reach the level of Pentium, not Celeron for LGA1155.

Vector graphics

Here in this group with archaic optimization for Core 2 Duo, only the older APU modification can bypass the Celeron. But there is some progress inside the class: Llano was greatly hindered by the lack of a common cache memory for inter-thread interaction, and additional cores (albeit in this family and absolutely “honest”) could not correct the situation, since they were not used. Trinity, judging by the results, is doing better than its predecessors, since even the budget ( is already budget) A6-5400K performs at the level of A8-3850, i.e. the second oldest processor for FM1.

Video encoding

Again a multi-threaded (at least up to four computation threads) group of tests, where, of course, single-module configurations fail — they can’t even cope with the A6-3500. But the presence of two modules allows you to easily bypass all Pentiums. “Steroid” modifications are already controlled with any Llano, Propus, old Core 2 Quad, younger Deneb (recall: all of the above are equipped with four “real” cores) and reach the second generation Core i3 level. Particularly noteworthy is the ability to compete with the last three groups, which turned out to be even a little unexpected, but … Core 2 Quad Q9650 (the oldest of the non-extreme models) “gains” 123 points in video encoding, which is only one point more than the A8-5600K; the Phenom II X4 955, which became popular after the last price cut, is even slightly behind it; and the results of the Core i3-2130 and A10-5800K are the same — 127 points. The transition to Ivy Bridge, of course, allowed the last family to recoup, but the performance of these solutions is at least comparable. Of course, this does not mean that a person who seriously works with video should focus on FM2 — in this field of activity, the costs of buying a Core i7 (not to mention i5 or FX-8000) are quite justified. However, if such a task arises only occasionally, then the older modifications of the new APUs will cope with it no worse than the once top processors. An important result from a practical point of view 🙂

Office software

Talking about the fact that any modern processor is redundant for office use, has long been hackneyed. However, no matter how much you repeat “halva”, it will not become sweeter in your mouth: other things being equal, any buyer will prefer what is faster. Even if he is sure in advance that it will not be possible to estimate the degree of this “quickness” by eye.

That’s why we state with deep satisfaction that the worst processor for FM2 in this group of tests is already equal to the best for FM1. The reason is clear: the vast majority of tests are single-threaded. However, the A8-3870K did not save the multi-threaded (moreover, very multithreaded) FineReader. And there is little chance of improving the results of multi-core processors — except that in Excel you can heroically squeeze something out, but not in Word or PowerPoint. A couple of threads are supported by any APU, and starting from A8, all four threads, so even with very active background loading (from the same antivirus), the maximum that can be achieved is a slight deterioration in the results of budget A4 and A6. And this is only in comparison with the old «senior» A6 and A8, which also have a TDP of 100 watts.

Intel processors are a little faster, of course. But we can talk about any sustainable advantage only in the case of Pentium. And here it will already affect, that any modern processor is redundant for office use , and a radical difference in the video part. That is, if you buy a computer for office, Internet and play , relying on an integrated IGP (as is often done with limited finances), then the same A6-5400K is fundamentally better than Pentium (not to mention Celeron). Just because in the office and on your Internet , it will be slightly slower (or even not slower at all; but, in any case, you won’t be able to detect the difference without tests), and the superiority of the Radeon HD 7540D over the HDG of the third and, even more so, the second generation in games will already noticeable to the naked eye.

Java

AMD claims its implementation of SMT (in the form of x86 cores) is better than Intel’s Hyper-Threading. SPECjvm seems to hate both versions equally, so despite good multithreading support, the A8-5500 barely beats the Pentium G2120. However, the picture is similar to rendering, and it may very well be that the problem lies in the insufficient number of vector blocks, of which there is one per module. But from a practical point of view, for a family home (especially if you remember the applicability of Java in this area), the performance of two-module models is sufficient. With single-modules, the situation is certainly worse: their performance is at a lower level than even the power-efficient Celeron T-series (but slightly better than the Celeron E3000 family for LGA775). But other results by this time, however, were difficult to expect.

Games

Not the most pleasant (once again) situation for single-module processors — even strongly “crushed” in terms of frequencies, but the tri-core A6-3500 is faster than them. And if rendering or compilation could be attributed to a non-target area of ​​​​use, then you can’t say the same about games: why else would you buy an APU, except for massively used 3D applications, of which there are almost none other than games? And here it turns out that not everything is so smooth. However, as everyone knows for a long time, first of all, games are demanding on video, and 9The 0599 Powerful Integrated Graphics and the Powerful Integrated Graphics are two very different graphics, so these features are not necessarily going to ruin the user’s life. But whether this is true or not, we will check next time.

There are no complaints about older models — at least they are not inferior to Athlon II X4 and outperform Pentium. And if we talk only about game engines with multithreading support, then the performance is comparable even with many Phenom II, Core i3, Core 2 Quad and even junior FX. Moreover, all this despite the fact that a common (for all computation threads) cache memory has not appeared, and its capacity is rather small at the present time. And if, again, we remember that from a global point of view, the Radeon HD 7560D/7660D can only be attributed to budget video cards, then the bottleneck in the final performance will most likely be here, and not at all in the processor part.

Multitasking environment

The quintessence of much of what has been said above (and before). First, four “cores” are not four cores after all. But with a successful selection of parameters, it’s a little worse: dual-module models are organically integrated between Pentium and Core i3. Secondly, two «cores» with a multi-threaded diverse load are very unattractive — no matter how hard the manufacturer tries, even the A6-5400K loses to the «classic» dual-core models of AMD itself, but the A6-3500 still steadfastly repels all Celeron attacks. By the way, while the 3500 and 3870K continue to be the cheapest A6 and A8, respectively, so if you are not afraid of the «moral death» of the FM1 platform and there is a need for a more or less serious computing load, the choice is obvious.

Total

Let’s start with global totals. As you can see, despite all our fears, the single-module design has shown its viability. Of course, such APUs do not shine in multi-threaded applications, but there are not so many of them in the domestic sphere. And taking them into account, the performance of the new A4 / A6 is at least higher than that of the old A4-3000 or even A6-3500 (of course, there are only three cores and the frequencies are low, but the availability of faster A6 / A8 with a TDP of 65 W until the last time left a lot to be desired). For desktop computers, this is only a basic level of performance, although AMD will be able to make good money in this segment as well — if these models are really made on crystals of their own design, and do not turn out to be mostly discarded older APUs. Well, in the mobile market, this line generally has every chance of becoming bestsellers, and in two segments at once. In budget laptops — because the A4-4300M and A6-4400M with a TDP of 35 W are unlikely to be inferior in performance to any A4-3000M and even MX, overtaking them at the same time and in graphics. And Llano previously did not fit into netbooks and ultra-thin laptops at all — until recently, they had to use the Brazos platform, which is acceptable only against the background of Atom, but not in any way in comparison with Celeron / Pentium CULV modifications. Here A4-4355M or A6-4455M, perhaps, will be able to compete with the latter. Even if they lose in the processor component — but the graphics are good. If only… If only the partners didn’t spoil everything by adding discrete video adapters to such models with a worthy better use (which negatively affects prices), and AMD itself would be able to provide the necessary supply volumes and the «correct» prices. But the groundwork for success, we repeat, is. With desktop models, it’s even easier. The A6-5400K has a performance equal to the Core 2 Duo E8200 (many people still have enough of the slower Pentium E5000 and even weaker ones) and slightly less than the Celeron G530, but in terms of integrated graphics, even the HDG 4000 (the best that can offer Intel, and not at all in this price class) «does not reach» A6.

Now let’s look at the other end — two models with a 100W TDP. Here is the meaning of their release is not too obvious. That is, it is clear that they are slightly faster than models with 65 W — so what? The increase in performance cannot be called radical, i.e. in fact, 5600K and 5800K are the same class as 5500/5700, respectively, and in the first pair the video part is absolutely the same. Did they do it because for FM1, for the most part, APUs with such a TDP were shipped? Doubtful explanation. FM2 is a new platform, so it would be a strong decision to «clamp» the heat pack at 65 W and promote it as a competitive advantage. Did you come up with these two models to compete with the Pentium and Core i3? But it still doesn’t work out: it would be nice if it were possible to “reach out” at least for the old Core i5-750, which can always be effectively beaten in advertising ( quad-core Cores are slower than our new APUs — no matter what), but not like now. Demonstrate superiority over Llano? More like the truth, but already the A8-5500 is faster than the entire line for FM1, with the exception of the A8-3850 and A8-3870K, and the A10-5700 is enough to replace these models — the top ones in the previous family and 100-watt ones, i.e. even lagging behind them could be successfully beaten, not to mention parity. Disposal of not the most successful crystals, unable to meet 65 watts? This is the closest thing to the truth. Moreover, the prices in pairs 5500/5600K and 5700/5800K are almost identical. And it seems to be a completely logical situation: who needs it cooler — those 65 W, and who needs it faster and with the possibility of overclocking (in which, for obvious reasons, official TDP levels do not matter) — 100 W. For the same money, so that no one will leave offended. But if you think about it, then the «offended» may well be found — just among overclockers. As soon as they understand that they are offered “for overclocking” not “the best grains” (for Intel, as we have already seen, the top-end Core i7-3770K, due to the selection of the best specimens, may turn out to be more economical than the “plebeian” and weaker in terms of TTX Core i5- 3450), but what “not quite” turned out.

In general, a little strange processors. Although interesting for a «regular» system: it is easy to remove not only 100 W from a typical microATX case, but also more heat, so it’s not necessary to «squeeze» too much. But a little faster than the «regular» series, and for the same money and with high-quality video, which we will study next time. But, in principle, the data already obtained is enough to have no doubts about the noticeable superiority of the new line of APUs (almost any) over LGA1155 processors, which compensates for both the higher performance of the latter in «classic» applications, and the lower heat pack — after all, to achieve parity in graphics, Intel processors will have to be used together with a discrete video card, and the latter cost extra money, and are far from being fed by the holy spirit (therefore, they also emit heat). That is, as we already wrote, APUs themselves are needed if and only if both components are planned to be used: both the CPU and the GPU . This was true of the first generation as well, while the second became a little more attractive in terms of technical characteristics, but without any global breakthroughs in performance. In any case, in terms of the performance of the processor part, it has been slightly improved, but this does not pull on a revolution (especially since there are areas of application where the new architecture is still losing out to the old one). As for graphics, as we have already seen, A10 is really a new level for integrated solutions, although there is no full-fledged competition (except for intra-company) in this sector. And how it looks «in the old positions» — we will study in the next article, fortunately, we repeat, for APU this issue is no less (or even more) important than processor speed.

We thank the company Corsair , Palit , “ F-Center ” and “ Yulmart
for help in the set of tests

are presented new processors AMD Ryzen 5000 with graphics VEGA.

General structure of Ampere GA102 and RDNA 2 Navi 21

The diagrams below don’t necessarily show how everything is physically arranged, but they do give a clear idea of ​​what components the processors are made of.

In both cases, the layouts are very familiar as they are essentially extended versions of their predecessors. Adding more components improves performance, which is especially useful given the high resolution rendering workloads in today’s 3D applications that involve a huge amount of parallel computing.

Such diagrams are useful, but it is more interesting for our analysis to see where the various components are located within the crystals themselves. When designing a large-scale processor, it is common to want shared resources such as controllers and cache to be centrally located to ensure that all components have the same path to them.

Interface systems like local memory controllers or video outputs should be located at the edges of the chip to make it easy to connect them to the thousands of individual wires that connect the GPU to the rest of the card.

False color images of AMD Navi 21 and NVIDIA GA102 dies below. Both images have been cleaned up and show only one layer inside the chip, but still give a good idea of ​​the insides of a modern GPU

The most obvious difference between the designs is that NVIDIA does not follow a centralized approach to chip layout: all system controllers and the main cache are at the bottom, and logic blocks are arranged in long columns. They have done this before, but only with models of the middle and lower price segment.

For example, the Pascal GP106 (used in the GeForce GTX 1060) was literally half the size of the GP104 (from the GeForce GTX 1070). In the earlier version, the die size was larger, and the cache and controller were located in the middle. At the younger brother, they moved to the side.

Pascal GP104 and GP106

For all previous top-end GPUs, NVIDIA used the classic centralized layout. Why change the approach? Interfaces have nothing to do with it, because memory controllers and PCI Express work on the edge of the crystal. This is not related to thermal problems either, because even if the cache part or the die controller gets hotter than the logical sections, you will probably want more heat-absorbing silicon in the middle of the circuit. While the reason for this change is not entirely clear, it is suspected to be related to the implementation of Render Output Units (ROPs).

We’ll look at these in more detail later, but for now, let’s just say that while the layout change looks weird, it doesn’t have a significant performance impact. This is due to the fact that 3D rendering is accompanied by a large number of long delays — usually due to the need to wait for data. Thus, the extra nanoseconds added by having some logical units further away from the cache are hidden in the overall chip design.

Before we continue, it’s worth noting the engineering changes AMD made to the Navi 21 layout compared to the Navi 10 found in the Radeon RX 5700 XT. Despite the fact that the new chip is twice the size of the previous one, both in area and in the number of transistors, the developers managed to improve clock speeds without a significant increase in power consumption. For example, the Radeon RX 6800 XT has a base frequency and boost frequency of 1825 and 2250 MHz, respectively, at a TDP of 300W. Same performance for Radeon RX 5700 XT: 1605 MHz, 1905 MHz and 225 W.

A performance-per-watt study of Ampere and RDNA 2 cards showed that both manufacturers have made significant improvements in this area, but AMD and TSMC have achieved something quite remarkable — compare the difference between the Radeon RX 6800 and Radeon VII in the graph above. The

Radeon VII is AMD and TSMC’s first collaboration using 7nm technology, and in less than two years they have increased their performance per watt by 64%. This begs the question: how much better could the Ampere GA102 be if NVIDIA stayed with TSMC.

Performance

Trinity’s compute performance is on average slightly better than Llano’s (+5-10%), although given the notable architectural differences, the difference may vary depending on the applications used. In some cases, first-generation APUs with four full cores can even be faster than a pair of dual-core modules running at a much higher frequency. In applied tasks, Trinity is not lost against the background of dual-core Intel Core i3, offering quite decent performance for its price. In single-threaded tasks, the processor from Intel will definitely have an advantage, the phenomenal efficiency of the Intel Core architecture makes itself felt. But in multi-threaded tasks, the number of computing units decides a lot, and here AMD’s quad-core CPUs have an advantage. Of course, Intel processors with the same number of cores are even more productive, but they are significantly more expensive.

During the test of the new APU, we also decided to evaluate the effectiveness of the CPU + GPU combination in applied tasks, using the Musemage graphics editor for this purpose, which uses the resources of the graphics core to perform various operations. The list of stages included the SVPMark benchmark, which also knows how to connect graphics for video processing.

The range of programs interspersed with heterogeneous computing is gradually expanding. Moreover, this is not only synthetic software for tests, but also applied applications. The pace, of course, leaves much to be desired, but there is hope that such developer initiatives will be strongly encouraged by hardware manufacturers. This is the rare case when the interests of both competitors coincide. Intel is also focusing more and more on the performance and capabilities of its integrated video with each successive architectural iteration. The Ivy Bridge chips here have noticeably succeeded in comparison with their predecessors, and in the expected Haswell, the graphics core should receive an even more significant increase in performance. In the meantime, AMD has a much stronger position here.

In 3D synthetics, Trinity has a very solid performance gain of 40-45%. Of course, in the overall standings, the increased performance of the x86 block is also taken into account, but this is not bad after all. 6000 points in 3DMark Vantage is almost the level of the Radeon HD 6570, that is, a discrete graphics card, which is now offered for $50-60. The indicators of Intel HD Graphics 2500 look noticeably more modest against the background of «embeddings» from AMD.

Intel offers separate processor modifications equipped with Intel HD Graphics 4000 graphics. In the case of dual-core models of the Ivy Bridge line, this is the Core i3-3225. It also has an operating clock speed of 3.3 GHz, like the Core i3-3220, but is equipped with a full-fledged graphics module with 16 compute units (the HD Graphics 2500 has only six), although it costs $ 20-25 more. At the time of writing, we did not have such a model, however, in order to include in the review not only the results of the Intel HD Graphics 2500, but also the results of Intel’s most powerful integrated graphics solution at the moment, we used the Core i7-3770K. It appears only in gaming tests with embedded video. This will allow a more balanced assessment of the current position and potential capabilities of the integrated GPUs of both companies.

In real games, the A10-5800K is again very confidently ahead of the A8-3850. The advantage is no longer as great as in the case of tests from Futuremark, however, an increase of 25-35% can also be considered an excellent result. In addition, the average 30 fps at 1920×1080 resolution already allows not only to view pictures in not the most simple games.

Solutions from Intel are expectedly less hasty, especially in the case of lightweight GPUs. It would seem that Intel HD Graphics 4000 has just managed to get close to Llano’s performance, as Trinity chips again make this mission impossible. We hope that with the release of Haswell there will be intrigue again.

Integrated video performance is highly dependent on the performance of the memory subsystem. Let’s see how RAM bandwidth affects gaming performance in the case of the A10-5800K.

If an integrated video core is used, DDR3-1866 is highly recommended. High-speed or overclocked memory modules to such frequencies will provide an 18-24% increase in performance in games.

Piledriver architecture

The Piledriver architecture is an upgraded version of the Bulldozer used for Zambezi (AM3+) chips.

Improved branch prediction blocks, data prefetching, improved L2 cache efficiency, increased L1 TLB size, improved INT and FP module loading scheduler. In addition, the new F16C instruction sets are now supported, as well as FMA3, which Intel plans to add to its Haswell chips. AVX kits are now available for new APUs, which were not supported by Llano chips. In general, Piledriver does not fundamentally differ from the Bulldozer architecture, it is a modified version with a number of improvements and cosmetic optimizations.

Accessories

Motherboards

  • HP Laptop 17-ca0xxx
  • Lenovo

  • F0AN003NTA
  • Gigabyte GA-G1.Sniper Z97
  • Toshiba WT8-A
  • Acer, Aspire 7530G
  • HP ENVY x360 Convertible 15-cp0xxx
  • Asus GRYPHON Z97

Video cards

  • FirePro V4900
  • GeForce 9500GS
  • GeForce 7100GS
  • RTX 2060S (Super)
  • GTX 980
  • GTX 650 Ti Boost
  • GTX 1050-Ti
  • GTX 960
  • Radeon R7 Graphics

SSD

We have put together a list of parts that users most often choose when building a computer based on A10-5800K APU (2012 D. Tr). Also with these components, the best results in tests and stable operation are achieved.

The most popular config: motherboard for AMD A10-5800K APU (2012 D.Tr) — HP Laptop 17-ca0xxx, video card — FirePro V4900.

APU Trinity 9 range0670

At the time of the launch of the new platform, the Trinity chip line includes six models. Two quad-core processors A10 and A8, as well as one each A6 and A4. As you can see, the number of x86 blocks is not reflected in the name of the APU series. At the same time, there is a dependence of the chip belonging to one or another line, which is determined by the number of computing cores of integrated graphics: A10 — 384, A8 — 256, A6 — 192, A4 — 128. This is another clear example of how the manufacturer wants emphasize the importance of the graphic component.

As for the cost of the new APUs, Trinity chips are actually in the same price segment as their predecessors — $50-130. At the same time, the pricing system is interesting. Both A10s are priced at $122. Both the model with an unlocked multiplier and the chip with a lower clock speed and a blocked KU have one recommended cost, which nevertheless has a TDP of 65 W, instead of 100 W for the flagship. The situation is exactly the same with the A8 line of APUs — both models are offered at the same price of $101. For some, higher performance has value, for others, more economical options are preferable. And for those and for others, suitable processors will cost the same price.

As is the case with Llano processors, as well as devices from a competitor, models with the “K” index have an unlocked multiplier. Curiously, now the most affordable model with this feature costs only $67, while the price of the previous generation APU with a free multiplier started at $80. However, the A6-3670K is a quad-core model, while the A6-5400K only has one module with a couple of dependent modules.

For Socket FM2, processors with disabled graphics core will also be available, which will complete the line of Athlon chips. Given the general concept of APU, it is obvious that separate crystals will not be produced for such models (although, given the area occupied by the GPU, this would make sense), for such processors chips will be used first of all, with certain problems in the graphic part, and if there are fewer of them than the market requires, then full-fledged crystals with a deactivated GPU will also be used.

Speed ​​of numerical operations

54.4

Minimum Medium Maximum
64 Memory: 79 90

Memory
83.6

46 1 core: 68 78

1 core
39.9

77 2 cores: 125 147

2 cores
39.1

26. 1

Minimum Average Maximum
124 4 cores: 211 251

4 cores
34.9

134 8 cores: 213 250

8 cores
17.8

3.9

Minimum Medium Maximum
131 All cores: 213 251

All cores
3.9

Different tasks require different CPU strengths. A system with few fast cores and low memory latency will be fine for the vast majority of games, but will be inferior to a system with a lot of slow cores in a rendering scenario.

We believe that a minimum of 4/4 (4 physical cores and 4 threads) processor is suitable for a budget gaming PC. At the same time, some games can load it at 100%, slow down and freeze, and performing any tasks in the background will lead to a drop in FPS.

Ideally, the budget shopper should aim for a minimum of 4/8 and 6/6. A gamer with a big budget can choose between 6/12, 8/8 and 8/16. Processors with 10 and 12 cores can perform well in games with high frequency and fast memory, but are overkill for such tasks. Also, buying for the future is a dubious undertaking, since in a few years many slow cores may not provide sufficient gaming performance.

When choosing a processor to work with, consider how many cores your programs use. For example, photo and video editors can use 1-2 cores when working with filtering, and rendering or converting in the same editors already uses all threads.

Data obtained from tests by users who tested their systems both with overclocking (maximum value in the table) and without (minimum). A typical result is shown in the middle, the more filled in the color bar, the better the average result among all tested systems.

History

Komodo Platform

Leaked roadmaps revealed ten-core Piledriver processors are part of the Komodo platform. Komodo was scheduled to launch in 2012 on socket FM2, but this never happened. AMD kept the AM3+ socket for the FX series and placed the Piledriver-based APU on the FM2.

FX, Athlon, and Opteron series

In 2010, AMD revealed that the 2nd generation was scheduled for release in 2012; AMD called this generation Enhanced Bulldozer . This later generation of the Bulldozer core was codenamed Piledriver .

On June 11, 2013, AMD announced two additional octa-core Piledriver FX series processors, FX-9590 and FX-9370, running at maximum turbo speeds of 5.0 GHz and 4.7 GHz, respectively, making AMD the first company to which has ever released 5GHz CPU commercially. AMD specifies that the 9xxx series processors require «reliable liquid cooling» due to their high thermal design power (TDP).

Trinity and Richland Athlon processor series — Desktop Budget market: Socket FM2 Athlon X4 730, 740, 750K and 760K processor models have four Piledriver core Trinity microarchitectures but lack on-chip integrated graphics. Athlon X2 340 is a dual-core model. Socket FM2 Richland based Athlon X4 760K and Athlon X2 370K processors were expected, both non-GPU and quad- and dual-core respectively.

For the server market, it has been stated that three versions are in development:

  • Web Service, Web Hosting & Microserver Market (1 CPU) : Opteron 3200 series ( Zurich ; 4 or 8 cores) was to be replaced by Delhi (4 or 8 cores) using Socket AM3+ format from the Desktop FX series. line. The memory controller had to support a dual-channel DDR3 memory configuration.
  • Market cost/power efficient servers (1 to 2 processors) : Opteron 4200 series ( Valencia ; 6 or 8 cores) was to be replaced by Seoul (6 or 8 cores). The Seoul will continue to use the Socket C32 format. The memory controller will support dual channel DDR3 memory configuration.
  • Enterprise/Mass Server Market (2 to 4 processors) : Opteron 6200 series ( Interlagos ; 4, 8, 12 and 16 cores) was to be replaced by Abu Dhabi (4, 8, 12 and 16 cores) )). Abu Dhabi will continue to use Socket G34. The memory controller will support a quad-channel DDR3 memory configuration.
    Trinity ending with «K» denotes processors with an unlocked CPU. multiplier . The Trinity APU line was praised for superior integrated graphics performance, but fell short of comparable Intel processor models in most demanding tasks.
  • Trinity A-Series APU — Notebook Mainstream and Performance market ( Comal platform): Laptops with picture Trinity APUs are shipping as early as June 2012. The mobile Trinity series has four APUs: A10-4600M, A8-4500M, A6-4400M and A4-4300M. In March 2013, AMD announced two more mobile models: A8-4557M and A10-4657M.

In January 2013, AMD officially introduced a new series of APUs codenamed Richland . A total of six new APUs are introduced in the series. The fastest model, the A10-6800K, featured two Piledrivers, a running at 4.1GHz and 4.4GHz in turbo mode, and an integrated HD 8670D GPU with 384 stream processors running at 844MHz. Only the A10-6800K has official support for DDR3-2133 memory. The A10-6800K offered approximately a 5% performance boost in applications and 3D gaming over its A10-5800K-based predecessor Trinity , mainly due to the higher clock speeds of the Richland and higher overclocking potential than the Trinity. On March 12, 2013, AMD officially unveiled four Richland mobile APUs. On June 4, 2013, AMD officially announced six Richland desktop APUs.

Chipsets

Despite the fact that Socket FM1 and Socket FM2 are incompatible with each other, the chipsets used on the previous generation platforms are quite suitable for the new one. We will see AMD A55 and AMD A75 chips in motherboards for Socket FM2. In general, there is nothing to be surprised here. Considering that the core functions of chipsets are taken over by central processors, their role in modern platforms is largely reduced to servicing peripherals. And here innovation does not happen so often. If there are already certain complaints about the functionality of the AMD A55 (lack of SATA 6 Gb / s), then the AMD A75 cannot be called obsolete. The latter became the first chipset in the industry with an integrated native USB 3.0 controller. And the rest of the «kit» is quite on the level.

To make the announcement of Socket FM2 even brighter, AMD also introduced a new chipset that will be used for this platform — AMD A85X. One of its key differences from the A75 is the ability to split the PCI-E x16 bus into two devices (x8 + x8), and, as a result, the ability to create CrossFire configurations with a pair of discrete video cards. In addition, the A85X now supports 8 instead of 6 SATA 6 Gb/s ports and allows you to create RAID 5 disk arrays. It also provides FIS-Based Switching channelization capabilities. In terms of support and configuration of the USB bus, no changes: 4 USB 3.0 ports, up to 10 USB 2.0 ports and up to two USB 1.1.

The Socket FM1 platform did not provide the ability to use two graphics adapters in the system. Such configurations are the lot of rather enthusiastic gamers or experienced crunchers. Obviously, in the case of Socket FM2, AMD wants to make the most versatile platform that could interest users with different needs in terms of performance and functionality.

Test results AMD Pumory

Finally, the most exciting moment has come — the tests. We divided them into three parts — a comparison of the performance of the CPU, video cores built into the processor, and testing the battery life. For greater clarity, we decided to compare AMD Pumori with three laptops at once.

The first opponent is the MSI GT60 gaming laptop. Of course, we will not put Pumori integrated graphics (Radeon HD 7660G) on a par with the discrete NVIDIA GeForce GTX 670M, which is installed inside the MSI GT60. We disabled the use of the latter in the driver settings, and the results of the MSI GT60 graphics tests show the performance of only the Intel HD 4000 video core integrated in the Intel Core i7-3610QM, an Ivy Bridge processor. Its TDP is 45W, which is 10W more than the AMD A10-4600M.

The next participant in the race is the Samsung NP900X3C ultrabook with a low-voltage Intel Core i7-3517U processor, also belonging to the Ivy Bridge generation. The TDP of this CPU is 17W.

Finally, a very compact laptop Samsung NP535U4C, or, more precisely, its modification with AMD A6-4455M, the thermal package of which also does not exceed 17 watts. In terms of dimensions, this laptop can be attributed to the ultra-compact.

Once again we note that in laptops we tested only the performance of the CPU and graphics integrated into processors.

General information

1.Supports 64-bit system
AMD A10-5800K

A 32-bit operating system can support up to 4 GB of RAM. 64-bit allows more than 4 GB, which improves performance. It also allows you to run 64-bit applications.

2.size semiconductors

32nm

A smaller size indicates a newer chip creation process.

3.GPU clock

800MHz

The graphics processing unit (GPU) has a higher clock speed.

4.Design requirements for heat dissipation (TDP)

100W

The heat dissipation requirement (TDP) is the maximum amount of energy that the cooling system will have to dissipate. A lower TDP also usually means less power consumption.

5. PCI Express (PCIe) version

2

PCI Express (PCIe) is a high-speed expansion card standard that is used to connect a computer to its peripherals. Newer versions support higher throughput and provide better performance.

6.CPU temperature

74°C

If the processor exceeds the maximum operating temperature, an accidental reset may occur.

7th DirectX version

11

DirectX is used in games with a new version that supports better graphics.

8.number of transistors

1303 million

A higher transistor count usually indicates a newer, more powerful processor.

9.OpenGL version

4.2

The newer version of OpenGL, the better graphics quality in games.

Features

Architecture name Trinity
Production date 26 September 2012
Price at first issue date $122
Place in the ranking 1366
Series AMD A-Series (Desktop)
Applicability Desktop
Support 64 bit
Crystal area 246 mm
Level 1 cache 128 KB (per core)
Level 2 cache 4096KB
Process 32nm
Maximum case temperature (TCase) 74°C
Maximum frequency 4. 2 GHz
Number of cores 4
Number of threads 4
Number of transistors 1303 Million
Unlocked
Supported memory types DDR3
Maximum number of processors in configuration 1
Supported sockets FM2
Power consumption (TDP) 100 Watt
AMD Virtualization (AMD-V)

Conclusion

More than a year ago, with the release of Llano processors, AMD surprised many with how well the integrated graphics of their APUs cope with their duties, and even raised the bar for the performance of the video core adjacent to the processor cores. Now that Intel has jumped into the race with Ivy Bridge-based Intel HD 4000 processors, it’s not easy for AMD’s Trinity line to win the mobile CPU segment. But one thing is for sure — the new APU series from AMD turned out to be very solid, which means that the market will have healthy competition and more reasonable prices than what we would expect if products of only one well-known company dominated.

Despite the fact that in the processor benchmarks, the CPU from AMD showed results below the competitor, for most users this performance will be more than enough. Of course, if you’re making money with a CPU, then it’s worth shelling out for something more powerful. If you need a computer to surf the Internet, watch videos and listen to music, then you won’t find a better AMD Trinity platform for the same money.

Results

The Socket FM2 platform and Trinity processors are quite an interesting option for building sufficiently powerful multimedia PCs. Compared to its predecessors, the performance of Piledriver computing units has not increased as much, while the integrated graphics capabilities have been improved by a third, reaching the performance of entry-level discrete graphics cards. At the moment, this is a serious advantage of AMD solutions. At the same time, the range of Trinity chips is exactly the same as that of Llano. Considering the balanced price, they will look very organically as part of inexpensive universal solutions “for everything”. And although recently mobile systems have been increasingly purchased for such tasks, new APUs in the desktop version will also find their customers.

Test bench configuration

Processor Intel Core i7-3770K Intel, www.intel.ua
Intel Core i3-3220 Euro Plus, www.eplus.kiev.ua
AMD A8-3850 AMD, www.amd.com
Cooler Thermalright Archon Rev.A «1-Incom», www.1-incom.com.ua
Motherboard MSI Z77A-GD65 (LGA1155, Intel Z77 Express) MSI, www.msi.com
ASRock A75 Pro4 (Socket FM1, AMD A75) ASrock, www. asrock.com
Gigabyte GA-F2A85X-UP4 (Socket FM2, AMD A85X) Gigabyte, www.gigabyte.ua
RAM Team Xtreem TXD38192M2133HC9KDC-L (2×4 GB DDR3-2133) DC-Link www.dclink.com.ua
Video card Radeon HD 7970 3GB (925/5500 MHz) AMD, www.amd.com
Accumulator Intel SSD 520 (SSDSC2CW240A3, 240 GB) Intel, www.intel.ua
Power supply Thermaltake Toughpower Grand TPG-1200M (1200 W) Thermaltake, www.thermaltakeusa.com

Compare Processors AMD A6 4400M vs. AMD A10 4600M

The AMD A6-4400M is a dual-core notebook processor based on the Trinity microarchitecture. The A6-4400M is configured with two CPU cores and an HD 7520G GPU. The new processor is designed to work in conjunction with the AMD A70M chipset and also has support for DDR3-1600 memory. The base frequency of the A6-4400M is 2. 6 GHz, the maximum is 3 GHz. These numbers are 18% and 15% higher than the highest performing dual-core model of the current generation, the A4-3330MX. The heat dissipation of the new processor is within 35 watts.

The processor is based on the Trinity architecture, which is an improved version of the Bulldozer architecture. Also, this APU has one module with two cores for integer operations, and one core for floating point operations, although it is itself considered a dual-core processor. The performance of the new processor is enough to perform everyday tasks, such as office programs and surfing the Internet, but for computer games, the power of the Radeon HD 7520 graphics core is clearly not enough. For these purposes, it is best to use dual graphics solutions, when the integrated graphics core in the processor and a discrete graphics card from AMD are used.

See prices in your city

Go to the store
The Radeon HD 7520 graphics core has 192 shader units and operates at a frequency of 497 MHz — 686 MHz. The built-in UVD3 (Avivo HD) video decoder performs full decoding of Multi-View Codec (MVC), MPEG-4 Part 2 (DivX, xVid), MPEG-4 AVC/H.264 and Adobe Flash HD video standards. The performance of the HD 7520G is slightly lower than the AMD Radeon HD 6630M discrete graphics card. Radeon HD 7520G is best suited for laptops with a diagonal of 13-14″ and larger.

benchmark

General result

Based on 7 tests:
Intel Core i5 2400 faster 188%

Intel Core i5 2400

288

AMD A6 4400M

100

GEEKBENSH 3 (MulTi)

Intel Core i5 2400 faster 296.84%

Intel Core i5 2400

8298

AMD A6 4400M

2091

9000 Geekbench 3 (Single Core)

Intel Core i5 2400 9000

004 Intel Core i5 2400

2638

AMD A6 4400M

1434

Geekbench 3 (Aes Single Core)

2400 2400 9000

1680000

Geekbench (32-bit)

Intel Core i5 2400 308.61%

Intel Core i5 2400

8258

AMD A6 4400M

9000 9000 9000 9000 9000 Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core Core SPE

Intel Core i5 2400

10769

AMD A6 4400M

3018

Complex A6 4455M with similar processors 9000 9000.

Tested with: PCMark 8 Home 3.0 Accelerated, PassMark, Geekbench 3 Multi-Core.

See also: MSI h410M PRO-VD PLUS {S1151v2, h410, 2xDDR4, PCI-Ex16, 2xPCI-Ex1, D-SUB, DVI, SATA III, GB Lan, USB

A6 4455M 4.9 Of 10
Core i3 3217u 5 .

Processor tested on: PassMark (Single Core), Geekbench 3 Single Core, Geekbench 3 AES Single Core.

000, Production Production from 1 BU uses electricity.

Tested with: Sky Diver, Cloud Gate, CompuBench 1.5 Bitcoin mining, CompuBench 1.5 Face detection, CompuBench 1.5 Ocean Surface Simulation, CompuBench 1.5 T-Rex, CompuBench 1.5 Video composition, PCMark 8 Home 3.0 Accelerated, PassMark, Geekbench 3 Multi -Core, PassMark (Single Core), Geekbench 3 Single Core, Geekbench 3 AES Single Core, TDP.

A6 4455M 5.3 out of 10
Core i3 3217U

Tested on: CompuBench 1.5 Bitcoin mining, CompuBench 1.5 Face detection, CompuBench 1.5 Ocean Surface Simulation, CompuBench 1.5 T-Rex, CompuBench 1.5 Video composition.

A6 44555m 5.7 of 10
Core i3 3217u 6.2 Of 10
Core I5 ​​3210M

5.9 of 10
A6 4455M 9.3 out of 10
Core i3 3217U 8.8 out of 10
Core i5 3210M 5. 6 out of 10

Price-performance ratio

How much you overpay for performance.

Tested on: Sky Diver, Cloud Gate, CompuBench 1.5 Bitcoin mining, CompuBench 1.5 Face detection, CompuBench 1.5 Ocean Surface Simulation, CompuBench 1.5 T-Rex, CompuBench 1.5 Video composition, PCMark 8 Home 3.0 Accelerated, PassMark, Geekbench 3 Multi -Core, PassMark (Single Core), Geekbench 3 Single Core, Geekbench 3 AES Single Core, System Price (adjusted).

9000 9000

A6 4455M 9.2 Of 10
Core i3 3217u 7

53

Benefits of the Intel Core i5 2400

  • High frequency (3. 1 GHz vs 2.7 GHz) means more operations per second
  • High turbo frequency: 3.4 GHz vs. 3.2 GHz
  • More cores (4 vs. 2) allows more processes to run simultaneously, speeding up computer performance and processing speed
  • More L3 cache (6 MB vs. 0.5 MB) allows the processor to access faster to data
  • Larger L3 cache per core: 1.5 MB/core vs. 0.25 MB/core
  • Higher integrated video processor frequency: 850 MHz vs. perform an overclocking procedure or compare characteristics with other models. The built-in tools of the operating system do not allow this, so you need to use third-party software. Popular representatives of such software offer a choice of several analysis options.
Min.
A6 4455M 7.2 of 10
Core i3 3217u 6.4 of 10
Core 3210M 5.9 of 10
Intel Core i5 2400 AMD A6 4400M
Frequency 3.1 GHz 2.7 GHC 900 9000

3,10 3,1 2 GHC
Number 4 2
Unlocked multiplier No
Lumber of flows 4 2
cache 2 levels 1 MB 1 MB
CASH 2 levels on the kernel 0. 25 MB/Nucleus 0.5 MB/Core
CASH 3 levels 6 MB 0.5 MB
CASH 3 levels on the kernel 1.5 MB/Nucleus 0.25 MB/Core
Process 32 nm

32 nm 31 27
Average Max.
20 1 core 45 59
32 2 Nuclei 70 86
min. Average Max.
38 4 cores 72 86
41 8 cores 73 86

Different tasks require different CPU strengths. A system with few fast cores is great for gaming, but will be inferior to a system with a lot of slow cores in a rendering scenario.

We believe that a processor with at least 4 cores/4 threads is suitable for a budget gaming PC. At the same time, individual games can load it at 100% and slow down, and performing any tasks in the background will lead to a drop in FPS.

Ideally, the buyer should aim for a minimum of 6/6 or 6/12, but be aware that systems with more than 16 streams are currently only applicable in professional applications.

Data obtained from tests by users who tested their systems both with overclocking (maximum value in the table) and without (minimum). A typical result is indicated in the middle, with a colored bar indicating the position among all tested systems.

Instructions supported

900 9000

Intel Core i5 2400 AMD A6 4400M
mmx
sse
sse2
SSE3
SSSE3
SSE4 SSE41
sse4a
sse42
fma3
fma4
AMD64
AMD_V
xop

AVX

✔0010 ✔
avx_11
f16c
aes
cvt16
CLMUL

characteristics AMD A6 4455M

General parameters

clock frequency 2. 1 GHZ
Maximum clock frequency 2.6 GHZ
Nuclei 2
Nuclei unlocks NO

Functions

SPIC

Yes

Yes

Yes

Yes Yes
Instructions supported AES AMD-V AMD64 AVX AVX 1.1 CLMUL CVT16 F16C FMA3 FMA4 MMX SSE SSE2 SSE3 SSE4.1 SSE4.2 SSE4a Supplemental SSE3 XOP

Details and features

9000

Streams 2
second -level cache (L2) 2 MB
Cash of the second level (L2) 1 MB/ker Third -level cache (L3) 1 MB
of the third level cache on the kernel (L3) 0.5 MB/Core
Technological process 32 nm
The number of transistors

PassMark

A6 4455M 1,277
Core i5 3210M 3,823
Core i3 3217U 2,299

PassMark (Single Core)

A6 4455M 762
Core i5 3210m 1. 524
Core i3 3217u

879

AMD A4 and A6 for SOCKET FM2

Basic level solutions, but with occasional play

After the publication of the article on processors for the AMD AM1 platform, some readers rightly pointed out to us that drawing parallels between A4/A6 for the FM1 and FM2 platforms is not entirely correct due to the fundamentally different processor parts of them. Still, the «old» A4 contained two «classic» cores similar to the Athlon II, and the A6 even more than that. At the same time, the transition to FM2 led to the fact that both trademarks went to «dual-half-core» (single-module) processors, albeit with improved graphics. At the same time, the desktop Kabini models that we have found are sometimes able to outperform even dual-core A4 for FM1, so their confrontation with models for FM2 (especially inexpensive ones — to compete directly with AM1, processors should cost less, since motherboards are more expensive) is somewhat ambiguous. . Well, today we will deal with the question of who is stronger: an elephant or a whale? Moreover, low-cost platforms are always of interest to users, but at times of market upheaval, they are paid even more attention than usual. 128 128 STD/MAX frequency, MHz 720 760 $ 34 (), T-10387646 $ 47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 (), T-47 ( with A4. Getting acquainted with FM2 two years ago, we studied model number 5300 on the Trinity core. Since that time, the range of processors in this design has expanded significantly, but almost all interesting models belong to the updated Richland family. As for the older models, everything is clear here, but the state of affairs in the budget segment is easily explained — at the time the platform entered the market, it was successfully occupied by the remnants of the A4-3300/3400 for FM1, but after their disappearance from the distribution network, the «hole» had cover something. AMD decided to «soak» the 40×0 line serial numbers for special artificially slow APUs. Firstly, clock frequencies — if the A4-5300 already worked at 3.4 / 3.6 GHz, and the transition to Richland made it possible to reach 4 GHz in turbo mode (this is just the A4-6320 we took), then the A4- 4000 is noticeably slower: only 3.0 / 3.2 GHz. Not so long ago, it was replaced by the A4-4020, but it can only boast of frequencies of 3.2 / 3.4 GHz, i.e. also slower than even the old A4-5300. Another significant limitation is that the maximum supported memory frequency is only 1333 MHz, although previously all A4s (starting with FM1) were able to work with DDR3-1600. In general, slowly. But the prices are at the level of Sempron 2650, so before the advent of AM1, the A4-4000 was the cheapest AMD processor. But also the slowest solution for FM2, of course, and it still remains so. 92 STD/MAX frequency, MHz 760 800 Price $ 47 (), T-8470929 $ 63 (), T-10737510

9000, now when the platform was changed, the concept was changed: if multi-core models fell into this family on the FM1 platform, then on the FM2 A6 it was just the “A4 Black Edition”, but, of course, with an improved video core. It is clear that overclock a single-module processor or do not overclock, and you will not conquer any global peaks, but this is at least possible — to the delight of especially greedy overclockers. As a result, the most interesting in this family is the same A6-5400K, which we met two years ago. Well, in order to roughly estimate what can be obtained from a slight overclocking, we also took the older model in the family, namely the A6-6420K.

A small digression. The older A6-6420K is only within the «canonical» version of FM2. Buying a board with FM2 + makes the younger model on the Kaveri core available to the user — A6-7400K. Unfortunately, it suffers from the same misfortune as older models on this core — too expensive: at the level of Pentium, from where everything is nothing to the younger A8. In general, the only “killer feature” is the ability to fit into a TDP of 45 W if necessary, but in normal mode, the purchase is not very justified, while there are still old models in stock. 2/2 CASH L1, I/D, KB 128/128 128/128 64/64 CASH L2, KB 2 × 512

2

2 9001

2048 2 × 256 L3 cache, MIB — — 3 RAM 2 × DDR3-1600 1600 TDP W 65 25

Test methodology

To evaluate performance, we used our performance measurement methodology using iXBT Notebook Benchmark v.1.0 and iXBT Game Benchmark v.1.0. We normalized all test results in iXBT Notebook Benchmark v.1.0 relative to the results of Pentium G3250 with 8 GB of memory and Intel 520 240 GB SSD, while the method of calculating the integral result remained unchanged. Another program that we, like last time, added to the test suite is the Basemark CL 1.0.1.4 benchmark, created to measure the performance of OpenCL code.

iXBT Notebook Benchmark v.1.0

As we have noted many times, in this resource-intensive application new-generation surrogate processors have a chance to compete on equal terms with younger desktop models of past years — after all, four cores against two. This is not enough to keep up with the Pentium, but two «Athlon II» cores or one «building» module can lose. What we are seeing is that all A4/A6 are noticeably behind the Athlon 5350. This is especially true of the old A4-3400 and the A4-4000 designed to replace it — these models are really almost equal. With an increase in clock frequencies, the situation becomes less sad, but even 4 GHz cannot completely fix it. By the way, we also note that despite the formal support for OpenCL by this proposal, in the first approximation there is no difference between A4 and A6.

But in these programs the A4-4000 has fallen behind not only the Athlon 5350, but also the A4-3400. However, slightly, and indeed — against the background of the Pentium G2130 (especially the Pentium G3250 we took for 100 points), all the AMD processors we took today are somehow the same.

Photoshop is not too fond of additional cores, however, it seems to have a good attitude towards OpenCL: here the younger A6 is kept on a par with the older A4, which has a higher clock frequency (by the way, the G2130 lags behind the reference G3250 to the maximum and to improve only its architecture don’t write at all). Well, the older of the younger APUs for FM2 are already one and a half times ahead of the best model for AM1: they lag behind the Pentium to a lesser extent.

Everything here is boring because it is proportional to the clock frequency on FM2, and the two AMD platforms “with ones at the end” are approximately equal in jumping to the slowest A4 for FM2, all together falling behind the Pentium G2130 twice.

Perhaps one of the few cases when it is clearly seen that even high-frequency «semi-nuclei» are not yet nuclei. And that the return of the “half-core” per gigahertz frequency is only at the level of that of the Kabini core: four at 2 GHz is exactly equal to two at four. Running on the spot is generally reconciling ©

In this case, having a single-threaded subtest still pushes FM2 forward, but not too far.

It’s interesting that in this test (we didn’t leave it in vain), other things being equal, dependence on the memory frequency begins to be observed. In general, at least this way “single-modules” can be leaders in their class.

What do we end up with? Athlon 5350 = A4-3400 = A4-4000. Other models for FM2 are a little faster, but in terms of processor speed they are not competitors of Pentium and hardly even Celeron, but in any case they are cheaper than the first ones. Also A4 ≈ A6, i.e. if you don’t play games and don’t overclock anything, you can buy A4. But we will deal with the video part a little later.

OpenCL

However, if you come across a task that is too heavily tied to OpenCL, A6 may be a better choice without any games. The «correct» A6, of course, and not a notebook Kabini.

Games

And here is the finest hour of our heroes — as expected. Moreover, we note that even the old A4 under FM1 does not pull the game, but already the A4-4000 with all its cuts is enough to give out 30 FPS on average. Well, any A6 is enough even with a margin, and considerable.

In principle, the Athlon 5350 could also be considered conditionally suitable, but this is not so important given the fact that adult APUs are certainly suitable for this mode of this game

Although formally their graphics are not much better than in Kabini, but a dual-channel memory controller with such a load is very relevant. And it’s not worth talking about the A6 — we have a large margin and an approach to the second limit of comfort.

The game is very processor-dependent and single-threaded, but with the selected settings it is not too demanding on the graphics core — as a result, the triumph of Pentium and the complete defeat of the Athlon 5350 (in principle, you can play with clenched teeth, but it’s better not to). Junior processors for FM2 occupy an intermediate position — it is already possible, but not yet very good. However, for WoT it is not in vain that Intel products are recommended — the performance of one computation thread is fundamental to the game. So the engine is written

As we have already said, the game is not very demanding, and the lack of results for some of the processors was due to the fact that the benchmark does not run everywhere. But in general, you can play even on AM1 (and not only the older model for this platform), and FM2 allows you to do it comfortably.

A6 are already approaching 30 FPS even in Metro (albeit in light mode), so you can try to play this game on them. On the rest — only if you lower the resolution even more, although there is practically no margin there, so this may not be enough.

Well, it’s too early for all users of low-end processors with integrated video to take aim at Hitman. Moreover, the game is clearly not indifferent to the number of computational threads, so there may be problems from this side as well.

In general, detailed comments are unnecessary. A6 is the minimum level to play occasionally. Not in all games, even with low picture quality, but here, at least, there is something to evaluate. And A4 is the minimum below which you can’t go down at all if you plan to run at least some games on your computer at least sometimes. Of course, there are exceptions in life, and sometimes very popular ones — such as WoT, but even they do not allow us to perceive Intel processors without discrete graphics as at least conditional gaming solutions. And surrogate platforms are all the more unsuitable for this role. None.

Total

In general, testing with predictable results: single-module processors for FM2 are typical desktop solutions. Yes, they are quite slow compared to budget Intel processors, but for the most part they are cheaper and better suited for at least occasionally running some games. True, Celeron and Pentium on Haswell should already reach the A4 level, but we will try to check this in one of the following articles, and A6 is still clearly ahead. In addition, the A6 will also be of interest to those who like to experiment, since buying it is one of the few methods of “budget overclocking” that have survived at the present time (however, one should not expect serious practical benefits from it, since this is initially a very limited solution). Well, the younger A4, given their price, is an excellent solution in cases where performance is not too important, but just the price is required. Moreover, in terms of performance, everything is not so bad either, since in any case they are not inferior to surrogate platforms both in terms of the processor part (especially in typical still low-threaded software) and in terms of gaming performance. At the same time, the younger A4s are quite competitive in price, and they really provide the possibility of modernization, unlike solutions for AM1: over time, the A4-4000, bought as a “socket plug”, can be exchanged even for the top-end A10 — when the money appears .

Of course, we repeat once again, all our praises for these solutions are determined by their price — otherwise, this is just a basic level. Quite basic. If possible, it’s better to pay extra. Especially if you’re interested in gaming, we still believe that, despite advances in the IGP space, it’s not possible to have a gaming PC without a discrete graphics card for at least $100 (and in this case, too, we will only be talking about a basic gaming PC). But if games are not an end in themselves, but only a side area of ​​activity, then you can limit yourself to integrated graphics within FM2. Only, of course, you should not buy A4 / A6, but at least A8. Moreover, these processors are not so much more expensive. And how much faster — we will check in one of the next articles.