AMD A6-3650 vs. Intel Core i3-2100 im Test: CPU + GPU für unter 90 Euro (Seite 4)
Inhaltsverzeichnis
Transcodierung
Moderne GPUs können weitaus mehr als nur die Grafik auf den Bildschirm zu zaubern. Ein Einsatzgebiet ist das Transcodieren von Videos, sei es in andere Auflösungen oder andere Codecs. Um die Geschwindigkeit auf AMDs neuer Fusion-Plattform zu testen, wandeln wir einen 1.080p-Trailer (3:30 Minuten) mittels des H.264-Codecs in die Auflösung 1.280×720 um.
Transcodierung – 1080p auf 720p
-
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
AMD A8-3850
-
Intel Core i3-2100
-
AMD A8-3850 @ 2,6 GHz
-
AMD A6-3650
-
Intel Pentium G840
-
-
-
Intel Pentium G840
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
Intel Core i3-2100
-
AMD A8-3850 + GT 430 (GPU)
-
AMD A8-3850 + HD 6670 (GPU)
-
AMD A8-3850 + HD 6570 (GPU)
-
AMD A8-3850
-
AMD A8-3850 @ 2,6 GHz
-
AMD A6-3650
-
AMD A8-3850 + HD 6450 (GPU)
-
Einheit: Minuten, Sekunden
Der Verlust vom 300 MHz kostet den A6-3650 beim Transcodieren eines Film-Trailers mit dem H. 264-Codec ziemlich genau eine Minute gegenüber dem A8-3850 – damit sind satte 16 Minuten notwendig, bis der Arbeitsaufwand erledigt ist. Unser auf 2,6 GHz herunter getakteter A8-3850 bestätigt diese Zeit.
Die Rechenzeit auf der GPU ist dagegen deutlich kürzer, wobei die optischen Ergebnisse von den CPU-Varianten abweichen. Hier liegt der Verlust des A6-Prozessor gegenüber dem größeren Bruder bei immerhin nur einer halben Minute – in neun Minuten und zwei Sekunden ist die Aufgabe erledigt. Die langsamere Radeon-Grafikeinheit macht übrigens nur die Hälfte der 30 Sekunden aus, da der A8-3850 mit 2,6 GHz acht Minuten und 49 Sekunden benötigt.
Der Intel Core i3-2100 benötigt auf der CPU rund 40 Sekunden weniger als der A6-3650, kann bei der GPU-Variante aber auftrumpfen. Dort sorgt Intels besserer ClearVideo-Algorithmus für einen Vorsprung von mehr als zwei Minuten.
Videobearbeitung mit vReveal
Mit Hilfe des Tools vReveal von MotionDSP lassen sich Videos bezüglich deren Qualität verändern. Sei es eine Rauschunterdrückung, das Verbessern von Farben, das Stabilisieren eines Videos, das nachhaltige Schärfen oder noch vieles mehr, vReveal macht ohne Hintergrundwissen sinnvolle Änderungen möglich. Wir testen die GPUs mit Hilfe der Version 3.0, wobei diese speziell auf AMD Vision-Produkte optimiert worden ist – dies sollte beachtet werden.
Als Testszenario schauen wir uns ein mit 30 Bildern pro Sekunde aufgenommenes Video in der Auflösung von 1.280×720 an und wenden die „One Click Fix“-Funktion auf dieses an. Die Herausforderung ist die Echtzeitanwendung und wie viele FPS noch übrig bleiben.
vReveal – Videooptimierung in Echtzeit
-
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
Intel Core i3-2100
-
AMD A8-3850
-
Intel Pentium G840
-
AMD A6-3650
-
AMD A8-3850 @ 2,6 GHz
-
-
-
AMD A8-3850 + HD 6670 (GPU)
-
AMD A8-3850 + HD 6570 (GPU)
-
AMD A8-3850 @ 2,6 GHz
-
AMD A8-3850
-
AMD A8-3850 + HD 6450 (GPU)
-
AMD A6-3650
-
AMD A8-3850 + GT 430 (GPU)
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
Intel Core i3-2100
-
Intel Pentium G840
-
Einheit: Bilder pro Sekunde (FPS)
Während der AMD A8-3850 bei der „One Click Fix“-Funktion in vReveal auf acht Bilder pro Sekunde kommt, wenn die Berechnungen auf der CPU durchgeführt werden, sind es beim A6-3650 nur sieben Frames per Second. Beides ist zu langsam, um die Funktion nutzen zu können.
Deutlich besser läuft dagegen die GPU-Variante, wobei der A6-3650 mit 17 Bilder pro Sekunde drei FPS langsamer als das Llano-Flaggschiff ist. Der Unterschied kommt in dem Szenario ausschließlich durch die schwächere GPU zu Stande, der geringere CPU-Takt macht keinen Unterschied aus.
WebGL-Performance
Die wohl größte Neuerung von Firefox 4 (und Firefox 5) sowie dem Internet Explorer 9 ist eine neuartige GPU-Beschleunigung, die mittels der WebGL-API die 3D-Grafik des Browsers unterstützen kann. WebGL wurde primär von der Khronos Group entwickelt, wobei auch AMD und Nvidia mit von der Partie waren. WebGL muss von der Grafikkarte unterstützt werden, was nicht durchweg der Fall ist.
Die WebGL-Performance testen wir anhand der bekannten „Aquarium“-Demo. Wir haben die Einstellung auf 500 Fische erhöht, um die Recheneinheiten mehr zu fordern. Darüber hinaus schauen wir uns noch den „WebGL-Benchmark“ in der „Medium“-Einstellung an, der ein synthetischer Test ist und mehrere 3D-Szenarien durchlaufen lässt. Das Ergebnis wird in Punkten ausgegeben, während die Aquarium-Demo die gewohnten FPS-Werte liefert.
WebGL – Aquarium
-
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
Intel Core i3-2100
-
Intel Pentium G840
-
AMD A6-3650
-
AMD A8-3850 @ 2,6 GHz
-
AMD A8-3850
-
-
-
AMD A8-3850 + HD 6670 (GPU)
-
AMD A8-3850 + HD 6570 (GPU)
-
AMD A8-3850 + HD 6450 (GPU)
-
AMD A8-3850 + GT 430 (GPU)
-
AMD A8-3850
-
AMD A8-3850 @ 2,6 GHz
-
Intel Core i7-2600K
-
AMD A6-3650
-
Intel Core i5-2500K
-
Intel Core i3-2100
-
Intel Pentium G840
-
Einheit: Bilder pro Sekunde (FPS)
WebGL – Synthetischer Test
-
-
Intel Core i7-2600K
-
Intel Core i5-2500K
-
AMD A8-3850
-
AMD A8-3850 @ 2,6 GHz
-
AMD A6-3650
-
Intel Core i3-2100
-
Intel Pentium G840
-
-
-
AMD A8-3850 + HD 6570 (GPU)
-
AMD A8-3850 + HD 6670 (GPU)
-
AMD A8-3850
-
Intel Core i7-2600K
-
AMD A8-3850 @ 2,6 GHz
-
Intel Core i5-2500K
-
AMD A6-3650
-
AMD A8-3850 + HD 6450 (GPU)
-
AMD A8-3850 + GT 430 (GPU)
-
Intel Core i3-2100
-
Intel Pentium G840
-
Einheit: Punkte
Im Aquarium-Benchmark für die WebGL-API gibt es in der CPU-Version keinerlei Differenzen zwischen den AMD-CPUs, mehr oder weniger als acht Bilder pro Sekunde erreicht kein Testkandidat.
Bei der GPU-Variante zeigen sich deutlich höhere FPS-Werte, jedoch spielt die CPU immer noch eine wichtige Rolle. So kommt der A8-3850 auf 38 Frames per Second. Taktet man diesen auf 2,6 GHz herunter, messen wir dagegen nur noch 35 FPS. Der A6-3650 mit derselben CPU-Frequenz und einer langsameren GPU kommt auf 34 FPS – der Hauptunterschied kommt also durch den geringeren CPU-Takt zu Stande.
Der Intel Core i3-2100 ist in der CPU-Version satte 50 Prozent schneller als die AMD-APU, hat dafür das Nachsehen, wenn die GPU genutzt wird. Dann liegt der Rückstand bei fünf Bildern pro Sekunde.
Dasselbe Ergebnis zeigt sich im synthetischen WebGL-Test. Selbst in der GPU-Variante spielt die CPU eine entscheidende Rolle. Dort ist der A6-3650 genau zehn Prozent langsamer als der A8-3850, wobei aber nur zwei Prozent mit der langsameren GPU begründet sind. Der Core i3-2100 liegt im synthetischen Test interessanterweise sowohl in der CPU- als auch in der GPU-Variante deutlich zurück.
Vorige Seite
Seite 4/15
Nächste Seite
Alltagseinsatz
clx-set-amd-ryzen-5 — Google Suche
AlleShoppingVideosBilderMapsNewsBücher
Suchoptionen
Tipp: Begrenze die Suche auf deutschsprachige Ergebnisse. Du kannst deine Suchsprache in den Einstellungen ändern.
CLX SET AMD RYZEN 5 ESSENTIAL RTS 1 — CLX Gaming
www.clxgaming.com › gmsetrtsbase02 › configure
Bewertung 5,0
(4) · 649,99 $ · Auf Lager
CLX SET. AMD RYZEN 5 ESSENTIAL RTS 1; $ 649.99.
CLX SET AMD RYZEN 5 ESSENTIAL RTS 3 — CLX Gaming
www.clxgaming.com › configure
749,99 $ Auf Lager
CLX SET. AMD RYZEN 5 ESSENTIAL RTS 3; $ 749.99.
CLX Set Gaming PC AMD Ryzen 5 3600 3.6GHz 6-Core, B450 …
www.amazon.com › …
Bewertung 4,8
(7) · 1.039,99 $ · Auf Lager
CLX Set Gaming PC AMD Ryzen 5 3600 3.6GHz 6-Core, B450 MATX, 16GB DDR4, GeForce GTX 1660 6GB, 960GB SSD, WiFi, Black Mini-Tower RGB Fans, Windows 11 Home.
Graphics Coprocessor: Nvidia Ge Force Gtx 1660
Graphics Card Ram Size: 6 GB
RAM: 16 GB DDR4
Processor: 3. 6 GHz ryzen_5_3600
CLX Set Gaming Desktop — AMD Ryzen 5 5600X 3.7Ghz 6-Core …
www.amazon.com › CLX-Set-Gaming-Desktop-Pro…
1.649,99 $ Auf Lager
CLX Set Gaming Desktop — AMD Ryzen 5 5600X 3.7Ghz 6-Core Processor, 32GB DDR4 Memory, GeForce RTX 3070 8GB GDDR6 Graphics, 480GB SSD, 3TB HDD, WiFi, Windows 10 …
Graphics Card Ram Size: 8 GB
RAM: 32 GB DDR4
Processor: 3.7 GHz ryzen_5
Graphics Coprocessor: GeForce RTX 3070
CLX SET Gaming Desktop AMD Ryzen 5 3600 16GB Memory …
www.bestbuy.com › … › Gaming Desktops
Bewertung 5,0
(2) · 1.039,99 $ · Auf Lager
CLX — SET Gaming Desktop — AMD Ryzen 5 3600 — 16GB Memory — NVIDIA GeForce GTX 1660 — 960GB Solid State Drive — Black/RGB ; Model · TGMSETGXM9600BM ; SKU · 6360467.
CLX SET Gaming PC — AMD Ryzen 5 3600 3.6GHz 6-Core , 8GB …
www.newegg.com › … › Gaming Desktops › CLX
749,49 $ Auf Lager
Buy CLX SET Gaming PC — AMD Ryzen 5 3600 3. 6GHz 6-Core , 8GB DDR4 2666, GeForce GTX 1650 4GB, 480GB SSD, WiFi, Black Mini-Tower RGB, Windows 11 Home with …
Graphics Graphics: NVIDIA GeForce GTX 1650
Operating System Operating System: Windows 11 Home
Colors: Black
Power Supply: 500 Watts
AMD Ryzen 9 7950X — AMD Radeon RX 7900 XTX — 32GB RAM
www.amd.com › shop › Desktops
CLX SET Gaming Desktop … AMD Ryzen 5 5600X — AMD Radeon RX 6600 XT — 16GB RAM … AMD Ryzen 5 PRO 4650G — Integrated AMD Graphics — 1024 HDD — 8GB RAM.
AMD Ryzen 5 5600X 3.7Ghz 6-Core Processor, 16GB DDR4 Me
www.ebay.com › itm
1.549,99 $ Auf Lager
It’s driven by a 4th Gen 6-Core/12-Thread AMD Ryzen 5 5600X 3.7GH z processor (Max. And it isn’t just the insides that look great; the CLX SET is housed in …
Processor Speed: 5600X (3.70 GHz)
Type: Desktop
GPU: NVIDIA GeForce RTX 3070
Hard Drive Capacity: 240 GB
CLX SET Gaming Desktop — AMD Ryzen 5 5500 3.6GHz 6-Core . ..
www.walmart.com › … › CLX Gaming Desktops
899,99 $
Arrives by Mon, Feb 6 Buy CLX SET Gaming Desktop — AMD Ryzen 5 5500 3.6GHz 6-Core Processor, 8GB DDR4 Memory, Radeon RX 6400 4GB GDDR6 Graphics, 500GB SSD, …
Processor Type: AMD Ryzen 5 5500
Processor Speed: 3.6 GHz
RAM Memory: 8 GB
Operating System: Windows 11 Home 64-bit
CLX — SET Gaming Desktop — AMD Ryzen 5 5600G — 8GB Memory …
www.supersalediscounts.com › products
449,99 $
Get an edge on your online gaming competition with this CLX SET gaming desktop computer built with an AMD Ryzen 5 5600G 3.9GHz processor that has 6-Cores …
Ähnliche Suchanfragen
CLX SET AMD Ryzen 5 essential rts 1 review
CLX SET Gaming desktop
CLX SET Gaming Desktop — AMD Ryzen 5 3600 8GB Memory NVIDIA GeForce GTX 1650 480GB SSD Black/RGB
CLX SET Case
CLX SET VR-Ready AMD Ryzen 9 3.7GHz 32GB RAM 480GB SSD, 3TB HDD Gaming Desktop
CLX gaming PC setup
Cybertron CLX Custom PC desktop computer AMD Ryzen 7 3800X 256gb 2TB RTX 3070
CLX Cart
Comparison AMD A6-3500 vs Intel Core i3-530 which is better?
Home / Processor / AMD A6-3500 VS Intel Core i3-530
AMD A6-3500
2%
Devicelist
VS
Intel Core I3-530
2%
9000 We have compared the specifications of AMD A6-3500 and Intel Core i3-530 and compiled a list of advantages and a comparison table for you. Find out which one to choose in 2023.
AMD A6-3500 Benefits
Core |
3 1 (50%) better than vs 2 |
Value for money |
91.5% 51.2% (127%) better than vs 40.3% |
Power consumption (TDP) |
65 W -8 W (-11%) better than vs 73W |
Benefits of Intel Core i3-530
Comparison winner
Threads |
4 1 (33.3%) better than vs 3 |
Base frequency |
2.933 GHz At 0.833 GHz (39.7%) better than vs 2. 1 GHz |
Maximum frequency |
2.93 GHz 0.53 GHz (22.1%) better than vs 2.4GHz |
Number of transistors |
382 million 381 million (38100%) better than vs 1 million |
64 bit support |
vs |
Passmark |
1523 142 (10.3%) better than vs 1381 |
General | |
Type |
|
Desktop | Desktop |
Architecture code name |
|
Llano | Clarkdale |
Cores A large number of cores improves performance in multi-threaded applications. |
|
3
1 (50%) better than |
2 |
Threads More threads help the cores process information more efficiently. Real performance will be noticeable in very specific tasks (video editing, databases). |
|
3 | 4
1 (33.3%) better than |
Base frequency |
|
2.1 GHz | 2.933 GHz
Better than at 0.833 GHz (39.7%) |
Process |
|
32 nm | 32 nm |
Chip size |
|
228 mm2 | 81 mm2 |
Number of transistors |
|
1 million | 382 million
381 million (38100%) better than |
Maximum frequency Faster clocked processors perform more calculations per second and thus provide better performance. |
|
2.4GHz | 2.93 GHz
Better than at 0.53 GHz (22.1%) |
Support 64 bit |
|
Max. number of processors in configuration |
|
1 | 1 |
Socket |
|
FM1 | FCLGA1156 |
AMD-V |
|
Release price |
|
n/a | 60 $ |
Value for money The sum of all the advantages of the device divided by its price. The more%, the better the quality per unit price in comparison with all analogues. |
|
91.5%
51.2% (127%) better than |
40.3% |
Maximum core temperature |
|
n/a | 73 °C |
vPro |
|
TXT Intel Trusted Execution Technology for hardware-based malware protection. For each protected program, the processor allocates its own isolated section of RAM. |
|
PCI Express revision |
|
no data | 2.0 |
Number of PCI-Express lanes |
|
n/a | 16 |
Demand Based Switching |
|
n/a | — |
PAE |
|
n/a | 36 bit |
FDI |
|
n/a | + |
Level 1 Cache The fastest level of cache that works directly with the core. The larger the cache, the better the performance. |
|
128 KB (per core) | 64 KB (per core) |
Level 2 cache |
|
1MB (per core) | 256 KB (per core) |
Level 3 cache |
|
n/a | 4 MB (total) |
Power Demand (TDP) Calculated thermal power indicates the average heat dissipation in load operation, |
|
65 W
-8 W (-11%) better than |
73W |
EDB |
|
n/a | + |
Permissible core voltage |
|
n/a | 0.65V-1.4V |
Benchmarks | |
Passmark |
|
1381 | 1523
142 (10.3%) better than |
Technology and additional instructions | |
Extended instructions |
|
n/a | Intel® SSE4.2 |
Turbo Boost |
|
n/a | — |
Idle States |
|
Enhanced SpeedStep (EIST) Intel’s technology that allows the processor to drop to its lowest frequency to conserve power when the processor is idle. |
|
Hyper-Threading Intel hardware technology that allows multiple threads to be processed on each processor core. For server applications, the performance improvement is up to 30%. |
|
RAM options | |
RAM types |
|
DDR3 Dual channel |
DDR3-1066, DDR3-1333 |
Allowable memory The maximum amount of RAM that can be used with this processor. |
|
n/a | 16.38 GB |
Number of memory channels |
|
n/a | 2 |
Virtualization Technologies | |
VT-x |
|
EPT |
|
Integrated graphics | |
Maximum number of monitors |
|
n/a | 2 |
Clear Video HD |
|
Video core The presence of the video core allows you to use your computer without using a video card. |
|
Radeon HD 6530D | Intel® HD Graphics for Previous Generation Intel® Processors |
Review and testing AMD APU A6-3500 GECID.com. Page 1
::>Processors
>2011
> AMD APU A6-3500
11/30/2011
Page 1
Page 2
One page
We all see that the modern processor market is just full of various rumors and discussions of products from leading manufacturers, as well as their further plans for the development of their products. Not so long ago there were announcements of new Bulldozer architectures from AMD and Sandy Bridge-E from Intel. If the former caused disappointment among many AMD fans due to unfulfilled hopes for a record level of performance, then the latter, focused on maximum performance, confuses the total cost of the system as a whole. Despite everything, both architectures somehow represent the so-called Hi-End class solutions for performing resource-intensive tasks. If you are building a system for use in so-called «domestic» tasks, which means playing multimedia, office work, web surfing and running not too demanding applications, there is no need for serious financial costs, because. each of the manufacturers is able to satisfy the wishes in the so-called budgetary sector.
AMD hybrid processors can be considered as one of such solutions. We have already introduced you to the quad-core APUs (AMD A6-3650 and A8-3850), which are able to provide a fairly high level of performance. However, if the buyer is either limited by financial resources, or simply does not need record performance, the developers also offer somewhat simplified models. One such option is the triple-core AMD APU A6-3500, which we will introduce you to in more detail today.
Appearance and packaging
Traditionally, let’s get acquainted with the boxed version of the processor from the packaging. We see that its graphic design is bright and catchy and, by the way, is absolutely identical to the more productive models considered earlier.
On the left side wall there is a transparent window through which you can easily get acquainted with the processor and its marking. Naturally, the developers focus the buyer’s attention on the advantages of a hybrid processor: support for the built-in DirectX 11 graphics core, energy efficiency, small dimensions. Below is a logo with the class designation AMD VISION A6. APUs in this class are designed for advanced multitasking, photo editing, HD video playback, and not too demanding games.
On the upper side there is a familiar sticker with the key characteristics of the proposed processor: model ( A6 3500 ), clock frequency ( 2.10 GHz ), cache memory size ( 3 MB ), processor socket ( FM1 ), serial number and product code.
You can see that apart from indicating the processor series (A6), there are no other identifying marks on the box that would allow specifying its model, which, in fact, we have seen before in other boxed versions of AMD APUs. This is quite logical, because. it is much easier to use a packaging design that is “generalizing” for a series and pack several product options into them. First of all, this approach can somewhat reduce the cost of printing due to flexible «manoeuvring» in the market, depending on the demand for certain models of hybrid CPUs.
In the boxed version of the package you will see a cooling system, instructions for installing the processor, which also contains information about the warranty, and a branded sticker for the PC case indicating the model range.
The complete cooling system is marked DKM-7D 52A-A1-GP. Unfortunately, we were not able to find out more detailed information about the characteristics of this system, because manufacturer has not been identified. In short, it can be described as a budget solution that will provide the temperature regime required for a hybrid processor. The design is absolutely identical to a number of coolers supplied in boxed versions of AMD’s CPUs. The noise background during operation is no different from solutions of a similar class and cost. Therefore, if the system unit is placed close to the user over time, it may need to be replaced with a more efficient and quiet cooling system.
The heat-spreading cover of the processor shows: processor model, marking (AD3500OJZ33GX) and place of production (Malaysia). You can decipher the marking as follows:
- A — the processor belongs to the AMD Athlon family;
- D — the scope of this processor — workstations;
- 3500 — model number;
- OJ — thermal package of the processor 65 W;
- Z — packaged processor in case 905 pin Socket FM1;
- 3 — total number of active cores;
- 3 — 1024 KB L2 cache per core and no L3 cache ;
- GX is the core of the LN-B0 stepping processor.
The back of the APU has 905 Socket FM1 pins, so you’ll need a motherboard with the appropriate CPU socket to work.
Specification
Model |
AMD APU A6-3500 |
Marking |
AD3500OJZ33GX |
Processor socket |
Socket FM1 |
Clock frequency, GHz |
2.1 |
Multiplier |
21 |
Bus frequency, MHz |
100 |
L1 Cache Size (Instructions Data), KB |
3×64 3×64 |
L2 cache size, KB |
3×1024 |
L3 cache size, KB |
— |
Core |
Llano |
Number of thread cores |
33 |
Instruction support |
MMX, 3DNow!, SSE, SSE2, SSE3, SSE4A, x86-64 |
Supply voltage, V |
— |
Power dissipation, W |
65 |
Critical temperature, °C |
70. 9 |
Process |
32 nm |
Technology support |
Dual Graphics |
Built-in memory controller |
|
Maximum memory, GB |
64 |
Memory types |
DDR3 (up to 1866 MHz) |
Number of memory channels |
2 |
Maximum throughput, GB/s |
29.8 |
ECC support |
No |
Integrated Radeon HD 6530D |
|
Stream Processors |
320 |
SIMD |
4 |
Texture blocks |
16 |
Blocks of raster operations |
2 |
Z/Stencil ROP Blocks |
32 |
Color ROP blocks |
8 |
GPU clock frequency, MHz |
443 |
Peak computing power, Gigaflops |
284 |
Instruction support |
DirectX 11 (Tessellation, ShaderModel 5. 0, DirectCompute 11) |
Video decode acceleration |
3rd Generation Video Decoder (UVD3) |
After getting acquainted with the AMD APU A6-3500 specification, we can once again state the fact that almost all manufacturers try to sell everything that comes off the assembly line. So this processor is no exception, because. most likely he himself appeared as a rejected version of a full-fledged quad-core AMD APU A6-3650 with a slightly slower clock speed and activated Turbo Core technology, which ultimately affected the TDP, which became equal to 65 watts.
According to the CPU-Z utility, we have a processor made using a 32nm process technology based on the Llano architecture. The clock frequency of the APU is 2100 MHz, while the core voltage was 1.404 V. Please note that we have a three-core processor at our disposal. The maximum TDP, which, according to the utility, is determined to be 88 W is not correct, because. the manufacturer in the specification indicates a value of 65 watts.
Cache memory is distributed as follows: 64 KB per L1 cache core with 2-line associativity, which are equally divided into data and instruction caching; L2 cache — 1024 KB per core with 16-line associativity. You can see that based on the memory associativity, AMD APUs are related to AMD Athlon II and AMD Phenom II CPUs of the previous generation.
The DDR3 memory controller operates in dual channel mode and can support up to DDR3-1866 RAM.
The Radeon HD 6530D graphics core has 320 shader pipelines and 16 texture units. A distinctive feature of the AMD APU is the manufacture of both the computing component and the graphics component according to the 32 nm process technology. For the most complete information about the performance and capabilities of the graphics cores of hybrid processors in comparison with discrete graphics cards, as well as their operation in Dual Graphic mode, you can find out in the article «Testing integrated graphics cards AMD Radeon HD 6410D, Radeon HD 6530D and Radeon HD 6550D».
Graphics core performance
To determine the performance ratio of the graphics cores of the APUs we have already studied in comparison with similar solutions from Intel, a series of tests was performed. Please note that DDR3-1866 memory modules were used in the system to get the maximum performance of the graphics cores.
After testing, we see that the competition in the performance of integrated graphics cores is observed exclusively among AMD solutions. Between themselves, they line up in a clear chain according to the performance of the graphics part of the APU, and then according to the power of the computing component. When compared with Intel products, we observe at least a twofold lag of Intel HD Graphics 2000/3000 in synthetic tests, while the situation is aggravated by the lack of support for DirectX 11.
In gaming tests at sufficiently high screen resolutions, AMD APU shows a decent level of integrated graphics performance.