Radeon 4850 power consumption: Power Consumption, Heat and Noise

HIS and Sapphire Radeon HD 4850 Face Off

With the recent run of newer and more affordable graphics cards from ATI, it’s almost easy to forget that there’s already a sub-$200 frame-rate cruncher called the Radeon HD 4850.  Based on the same RV770 chip as the more powerful, yet more expensive HD 4870, the HD 4850 ships with the same 800 stream processors, 40 texture units, and 16 ROPs that have made these cards such hot items.  The main area where they differ, other than clock speeds, is in regard to memory.  While the HD 4870 ships with high-end GDDR5 memory chips, the HD 4850 finds itself loaded with 512 MB of more mainstream GDDR3.

A major caveat with the Radeon HD 4850 that has arisen, however, is the excessive heat that gets produced by the RV770 GPU.  It seemed the single slot reference cooling solution had a tough time handling the heat output of the RV770.  With heat being a concern, and manufacturers having more time to tinker with their cards, we weren’t surprised to find out that two of ATI’s major partners, HIS and Sapphire, had two new cards waiting in the wings, each with their own ramped up clock speeds but with some superior cooling thrown in as well.   Before we get down to comparing these two head-to-head, let’s take a quick look at the basic specifications of the Radeon HD 4850. 

AMD ATI Radeon HD 4800 Series

Specifications and Features

  • 956 million transistors on 55nm fabrication process
  • PCI Express 2. 0 x16 bus interface
  • 256-bit GDDR3 memory interface
  • Microsoft DirectX 10.1 support

    • Shader Model 4.1
    • 32-bit floating point texture filtering
    • Indexed cube map arrays
    • Independent blend modes per render target
    • Pixel coverage sample masking
    • Read/write multi-sample surfaces with shaders
    • Gather4 texture fetching
  • Unified Superscalar Shader Architecture

  • Dynamic Geometry Acceleration

    • High performance vertex cache
    • Programmable tessellation unit
    • Accelerated geometry shader path for geometry amplification
    • Memory read/write cache for improved stream output performance
  • Anti-aliasing features

    • Multi-sample anti-aliasing (2, 4 or 8 samples per pixel)
    • Up to 24x Custom Filter Anti-Aliasing (CFAA) for improved quality
    • Adaptive super-sampling and multi-sampling
    • Gamma correct
    • Super AA (ATI CrossFireX configurations only)
    • All anti-aliasing features compatible with HDR rendering
  • Texture filtering features

    • 2x/4x/8x/16x high quality adaptive anisotropic filtering modes (up to 128 taps per pixel)
    • 128-bit floating point HDR texture filtering
    • sRGB filtering (gamma/degamma)
    • Percentage Closer Filtering (PCF)
    • Depth & stencil texture (DST) format support
    • Shared exponent HDR (RGBE 9:9:9:5) texture format support
  • OpenGL 2. 0
  • ATI PowerPlay
  • Advanced power management technology for optimal performance and power savings
  • Performance-on-Demand
    • Constantly monitors GPU activity, dynamically adjusting clocks and voltage based on user scenario
    • Clock and memory speed throttling

    • Voltage switching

    • Dynamic clock gating

    • Central thermal management – on-chip sensor monitors GPU temperature and triggers thermal actions as required

  • ATI Avivo HD Video and Display Platform

    • 2nd generation Unified Video Decoder (UVD 2)

      • Enabling hardware decode acceleration of H.264, VC-1 and MPEG-2
      • Dual stream playback (or Picture-in-picture)
    • Hardware MPEG-1, and DivX video decode acceleration

      • Motion compensation and IDCT
    • ATI Avivo Video Post Processor

      • New enhanced DVD upconversion to HD new!
      • New automatic and dynamic contrast adjustment new!
      • Color space conversion
      • Chroma subsampling format conversion
      • Horizontal and vertical scaling
      • Gamma correction
      • Advanced vector adaptive per-pixel de-interlacing
      • De-blocking and noise reduction filtering
      • Detail enhancement
      • Inverse telecine (2:2 and 3:2 pull-down correction)
      • Bad edit correction
      • Full score in HQV (SD) and HQV (HD) video quality benchmarks
    • Two independent display controllers

      • Drive two displays simultaneously with independent resolutions, refresh rates, color controls and video overlays for each display
      • Full 30-bit display processing
      • Programmable piecewise linear gamma correction, color correction, and color space conversion
      • Spatial/temporal dithering provides 30-bit color quality on 24-bit and 18-bit displays
      • High quality pre- and post-scaling engines, with underscan support for all display outputs
      • Content-adaptive de-flicker filtering for interlaced displays
      • Fast, glitch-free mode switching
      • Hardware cursor
    • Two integrated DVI display outputs

      • Primary supports 18-, 24-, and 30-bit digital displays at all resolutions up to 1920×1200 (single-link DVI) or 2560×1600 (dual-link DVI)
      • Secondary supports 18-, 24-, and 30-bit digital displays at all resolutions up to 1920×1200 (single-link DVI only)
      • Each includes a dual-link HDCP encoder with on-chip key storage for high resolution playback of protected content4
    • Two integrated 400MHz 30-bit RAMDACs

      • Each supports analog displays connected by VGA at all resolutions up to 2048×1536
    • DisplayPort output support

      • Supports 24- and 30-bit displays at all resolutions up to 2560×1600
    • HDMI output support

      • Supports all display resolutions up to 1920×1080
      • Integrated HD audio controller with up to 2 channel 48 kHz stereo or multi-channel (7. 1) AC3 enabling a plug-and-play cable-less audio solution
    • Integrated AMD Xilleon HDTV encoder

      • Provides high quality analog TV output (component/S-video/composite)
      • Supports SDTV and HDTV resolutions
      • Underscan and overscan compensation
      • MPEG-2, MPEG-4, DivX, WMV9, VC-1, and H.264/AVC encoding and transcoding
      • Seamless integration of pixel shaders with video in real time
      • VGA mode support on all display output
    • ATI CrossFireX Multi-GPU Technology

      • Scale up rendering performance and image quality with two GPUs
      • Integrated compositing engine
      • High performance dual channel bridge interconnect

      While the specs above give you a detailed technical overview of what makes the HD 4850 tick, it doesn’t cover all of the nuances that you might be looking for.  However, since we’ve already extensively covered the technology behind not only the HD 4850, but the HD 4870 as well, we won’t rehash that information here again. If you want to learn more about the technologies employed by ATI, please take a look at the following previous articles, which cover them in detal:

      • ATI Radeon HD 4850 and 4870: RV770 Has Arrived
      • ATI Radeon HD 4870 X2 — AMD Back On Top
      • ATI Radeon HD 3870 and 3850: 55nm RV670
      • ATI Catalyst 8.3 Sneak Peek: CrossFireX and More

      Once up to speed, we’ll take a closer look at our two entries in today’s HD 4850 Face-off: the HIS Radeon HD 4850 IceQ4 TurboX and Sapphire’s Toxic Radeon HD 4850.

  • GeForce GTX 950M vs Mobility Radeon HD 4850 : Which one is better?


    GPU Comparison

    NVIDIA GeForce GTX 950M vs ATI Mobility Radeon HD 4850


    NVIDIA GeForce GTX 950M
    ATI Mobility Radeon HD 4850

    We compared two Mobile platform GPUs: 4GB VRAM GeForce GTX 950M and 1024MB VRAM Mobility Radeon HD 4850 to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.

    Main Differences

    NVIDIA GeForce GTX 950M’s Advantages

    Released 6 years and 2 months late

    Boost Clock 1124MHz

    More VRAM (4GB vs 1GB)

    ATI Mobility Radeon HD 4850’s Advantages

    Larger VRAM bandwidth (54.40GB/s vs 28.80GB/s)

    160 additional rendering cores


    FP32 (float)

    GeForce GTX 950M

    1.439 TFLOPS

    Mobility Radeon HD 4850

    0.804 TFLOPS

    FP64 (double)

    GeForce GTX 950M

    0. 045 TFLOPS

    Mobility Radeon HD 4850

    0.161 TFLOPS

    GeForce GTX 950M


    Mobility Radeon HD 4850

    Graphics Processor


    GPU Name



    GPU Variant

    M98 L2







    28 nm

    Process Size

    55 nm

    1,870 million


    956 million


    Die Size


    Graphics Card

    Mar 2015

    Release Date

    Jan 2009

    GeForce 900M






    PCIe 3. 0 x8

    Bus Interface

    PCIe 2.0 x16

    Clock Speeds


    Base Clock



    Boost Clock



    Memory Clock




    Memory Size



    Memory Type



    Memory Bus


    28. 80 GB/s


    54.40 GB/s

    Render Config


    Shading Units



    SM Count



    Tensor Cores



    RT Cores


    64 KB (per SMM)

    L1 Cache

    16 KB (per CU)


    L2 Cache


    Theoretical Performance

    17. 98 GPixel/s

    Pixel Rate

    8.048 GPixel/s

    44.96 GTexel/s

    Texture Rate

    20.12 GTexel/s


    FP16 (half)


    1439 GFLOPS

    FP32 (float)

    804.8 GFLOPS

    44.96 GFLOPS

    FP64 (double)

    161. 0 GFLOPS

    Board Design





    Suggested PSU


    No outputs


    Portable Device Dependent


    Power Connectors


    Graphics Features

    12 (11_0)


    10.1 (10_1)

    4. 6













    Shader Model


    Related GPU Comparisons


    NVIDIA GeForce GTX 950M vs
    NVIDIA GeForce GTX 1050


    NVIDIA GeForce RTX 4060 Mobile vs
    NVIDIA GeForce GTX 950M


    NVIDIA GeForce MX250 vs
    NVIDIA GeForce GTX 950M


    NVIDIA GeForce 940MX vs
    NVIDIA GeForce GTX 950M


    NVIDIA GeForce GTX 950M vs
    AMD Radeon Pro 570X


    NVIDIA GeForce MX550 vs
    NVIDIA GeForce GTX 950M


    NVIDIA GeForce GTX 950M vs
    NVIDIA GeForce GT 520MX


    NVIDIA GeForce GTX 950M vs
    NVIDIA GeForce 9800M GTS


    NVIDIA GeForce GTX 950M vs
    NVIDIA GeForce RTX 2050 Mobile


    NVIDIA GeForce GTX 950M vs
    AMD Radeon HD 6530M

    © 2023 — TopCPU. net  

    Contact Us

    Compare NVIDIA GeForce 210 PCI and ATI Radeon HD 4850

    Comparative analysis of video cards NVIDIA GeForce 210 PCI and ATI Radeon HD 4850 according to all known characteristics in the categories: General information, Specifications, Video outputs and ports, Compatibility, dimensions, requirements, API support, Memory.
    Performance analysis of video cards by benchmarks: PassMark — G3D Mark, PassMark — G2D Mark, GFXBench 4.0 — T-Rex (Frames), GFXBench 4.0 — T-Rex (Fps).

    NVIDIA GeForce 210 PCI


    ATI Radeon HD 4850


    Reasons to choose NVIDIA GeForce 210 PCI

    • Newer graphics card, release date difference 1 year(s) 3 month(s) 9002 0
    • A newer technological process for the production of a video card allows it to make it more powerful, but with less power consumption: 40 nm vs 55 nm
    • 3. 5 times less power consumption: 31 Watt vs 110 Watt
    Issue date 12 October 2009 vs 25 June 2008
    Process 40 nm vs 55 nm
    Power consumption (TDP) 31 Watt vs 110 Watt

    Reasons to choose ATI Radeon HD 4850

    • About 32% more core clock: 625 MHz vs 475 MHz
    • 6.6 times more texturing speed: 25 GTexel / s vs 3.8 GTexel / s
    • 50 times more shader processors: 800 vs 16
    • 28.4 times greater floating point performance: 1,000.0 gflops vs 35.2 gflops
    • 2.5 times greater memory frequency: 1986 MHz vs 800 MHz
    Core frequency 625 MHz vs 475 MHz
    Texturing speed 25 GTexel/s vs 3. 8 GTexel/s
    Number of shaders 800 vs 16
    Floating point performance 1,000.0 gflops vs 35.2 gflops
    Memory frequency 1986 MHz vs 800 MHz

    Benchmark comparison

    GPU 1: NVIDIA GeForce 210 PCI
    GPU 2: ATI Radeon HD 4850

    Name NVIDIA GeForce 210 PCI ATI Radeon HD 4850
    PassMark — G3D Mark 937
    PassMark — G2D Mark 94
    GFXBench 4. 0 — T-Rex (Frames) 3344
    GFXBench 4.0 — T-Rex (Fps) 3344

    Performance comparison

    NVIDIA GeForce 210 PCI ATI Radeon HD 4850
    Architecture Tesla 2.0 TeraScale
    Codename GT216 RV770
    Production date October 12, 2009 25 June 2008
    Place in the ranking not rated 1209
    Type Desktop Desktop
    Price at first issue date $199
    Core frequency 475MHz 625MHz
    Floating point performance 35. 2 gflops 1,000.0 gflops
    Process 40nm 55nm
    Number of shaders 16 800
    Texturing speed 3.8 GTexel/s 25 GTexel/s
    Power consumption (TDP) 31 Watt 110 Watt
    Number of transistors 486 million 956 million
    Video connectors No outputs 2x DVI, 1x S-Video
    Interface PCI PCIe 2. 0 x16
    Length 145mm 246 mm
    Additional power connectors None 1x 6-pin
    DirectX 10.1 10.1
    OpenGL 3.3 3.3
    Maximum memory size 512MB 512MB
    Memory bandwidth 6.4GB/s 63.6 GB/s
    Memory bus width 64 Bit 256 Bit
    Memory frequency 800MHz 1986 MHz
    Memory type DDR2 GDDR3

    AMD Radeon HD 4770 — resistance is useless! / Video cards

    Together with the awakening of nature, manufacturers of graphic solutions are showing unprecedented activity this spring. The theater of competitive battles is moving from one price category to another, various marketing and technical methods are being used, and there is no end to these competitions. To our delight and to spite each other, AMD and NVIDIA are releasing more and more new solutions to the market, one is better than the other. At the same time, it is very difficult, and sometimes simply impossible, to determine the absolute leader in a particular niche. It seems that AMD has decided to break the vicious circle and consolidate its leadership in at least one of the price categories. To achieve this goal, AMD is launching a new graphics accelerator on the market, the name of which is the Radeon HD 4770. The novelty is based on the industry’s first 40-nm RV740 graphics processor.

    An interesting fact — for the first time in a long time, the technological process used in the production of a video chip turned out to be «thinner» than the technological process by which the most advanced models of x86-compatible central processors are made. According to AMD itself, the new HD 4770 graphics card is the fastest in the «under $100» category. Well, a very intriguing statement, which we will definitely check today. But first, let’s look at the summary table of technical characteristics of modern AMD solutions:

    Radeon HD4850 ​​

    Radeon HD4830

    Radeon HD4770

    Radeon HD4670






    Number of transistors, million





    No. of universal processors










    ROP blocks





    Core frequency, MHz





    Memory frequency, MHz





    Video memory type





    Memory capacity, bit





    Theoretical memory bandwidth, Gb/s


    57. 6



    Memory size

    512 MB/1 GB



    512 MB/1 GB

    Power consumption, W





    Average price according to Market.3DNews, rub

    No data

    No data

    No data

    No data

    The table shows that the HD4770 graphics card strongly resembles the HD4830 in terms of its technical characteristics — the same number of universal processors, the same number of ROPs and TMUs. However, the new product contains a number of changes that, in our opinion, should have a positive impact on the performance level of the HD4770 compared to the HD4830.

    New GPU workflow

    The transition to a thinner process technology has significantly reduced the size of the substrate and chip. Area RV740 — 140 sq. mm vs. 260 sq. mm at 55 nm RV770, which, coupled with a decrease in the number of transistors in the graphics core, had a positive effect on the final cost of the finished product. In addition, according to AMD, the final power consumption of the HD4770 video card was kept within 80 W versus 110 W for the HD4830, and this is taking into account a significant increase in clock frequencies.

    Use of GDDR-5

    The first motherboards from AMD with installed GDDR-5 video memory appeared quite a long time ago. At first, only top solutions at that time received support for this memory — Radeon HD4870, HD4870X2 and, subsequently, HD4890. Now AMD has decided to introduce support for GDDR-5 video memory even in budget solutions, and the Radeon HD4770 is an example of this. It is worth saying that GDDR-5 video memory is more expensive than GDDR-3, but this did not affect the final cost of the HD4770, since the layout of the board has become a little simpler, and, consequently, the PCB itself turned out to be cheaper.

    We got a reference sample of the Radeon HD 4770 for testing.

    The board is made on a red textolite, a classic solution for most AMD reference video cards. The cooling system has a very elegant shape, reminiscent of its lines of a spaceship or a racing car.

    Compared to the Radeon HD4830, the PCB length of the HD4770 has decreased by almost 3 cm (20.7 cm for the HD4770 versus 23.2 cm for the HD4830).

    The Radeon HD4770 requires only one 6-pin PCI-Express power connector to operate.

    The GPU power subsystem is built according to a 3-phase scheme, the memory is 1-phase.

    The board has GDDR5 video memory manufactured by Quimonda, marked IDGV51-05A1F1C-40X, nominal for which is a clock frequency of 4 GHz DDR (actual frequency is 1 GHz)

    The video card cooling system consists of two components. To remove heat from the memory subsystem, a heatsink with heat-conducting pads glued to it is used, and a turbine with two heat pipes is used to cool the GPU and the power subsystem, transferring heat from the aluminum base to the heatsink. In order to prevent direct contact of the metal base of the cooling system with the elements of the board, the base is covered with a thin layer of transparent polymer material that does not conduct current.

    Test participants

    The list of participants in today’s testing is very large. We tried to take into account all the «neighbors» of the HD 4770, both from the «native» and from the nearest price categories.

    From the side of AMD, the following take part in testing:

    • AMD Radeon HD 4670
    • AMD Radeon HD 4830
    • AMD Radeon HD 4850

    Honor NVIDIA uphold:

    • NVIDIA GeForce 9800GT
    • NVIDIA GeForce GTS 250

    The following configuration was used as a test bench:

    CPU Intel Core i7 920 @ 4.0GHz (200×20)
    Cooling system ProModz Cooled Silence Extreme Package
    Motherboard ASUS P6T Deluxe Palm OS Edition
    RAM DDR3 OCZ Platinum 3x2GB @ 1600MHz @ 7-7-7-24
    Power supply IKONIK Vulcan 1200W
    Hard disk Western Digital Raptor 36GB + Seagate 160Gb
    Body IKONIK RA 10X SIM
    Operating system Windows Vista Home Basic x64 SP1
    Driver version for NVIDIA/AMD cards ForceWare 182. 50 WHQL/Catalyst 9.4

    A few words about overclocking

    At the time of writing this review, full-fledged overclocking of the HD4770 in both single and CrossFireX configurations is supported only by proprietary AMD drivers. A significant drawback of the standard OverDrive utility is the small range of settable frequencies for the core and memory. However, very soon there will be third-party utilities that will solve this problem. Already now, if desired, it is quite possible to use the RivaTuner 2.24 program, for this you just need to make changes to the rivatuner.cfg configuration file. Open the rivatuner.cfg file, find the line «RV770 = 9440h-9443h,944Ch» and add a comma to the end of the «94B3h» line, after which, by launching RivaTuner, you will have access to changing the core and memory clock frequencies, as well as to the cooler control functions. In addition, RivaTuner will allow you to monitor the GPU temperature .

    Today’s review also includes an overclocked version of the Radeon HD4770. During testing, we limited ourselves to using standard overclocking methods, as a result, the set frequencies of the Radeon HD4770 turned out to be the maximum possible for AMD drivers and amounted to 830/850(3400) MHz.

    List of test packages and test modes

    Test package

    Test mode

    3DMark Vantage Performance, High, Extreme
    For NVIDIA video cards, PhysX technology is disabled.
    CRYSIS version 1.2.1 DX10 / VeryHigh /1280×1024, 1600×1200, 1920×1200 / noAA, noAF
    PhysX technology disabled for NVIDIA video cards.
    FarCry 2 Benchmark DX10 / VeryHigh /1280×1024, 1600×1200, 1920×1200 / noAA, noAF
    PhysX technology disabled for NVIDIA video cards.
    HAWX built-in Benchmark DX9 / All High /1280×1024, 1600×1200, 1920×1200 / noAA, noAF
    PhysX technology disabled for NVIDIA video cards.
    Call of Juarez DirectX10 Benchmark DX10 / Max Detail /1280×1024, 1600×1200, 1920×1200 / noAA, noAF
    For NVIDIA video cards, PhysX technology is disabled.
    World in Conflict DX10 / all graphics options set manually to maximum quality /1280×1024, 1600×1200, 1920×1200 / noAA, noAF
    For NVIDIA video cards, PhysX technology is disabled.
    Unique Tropics DirectX10 Benchmark DX10 / High Detail /1280×1024,1600×1200,1920×1200 / noAA, noAF
    For NVIDIA video cards, PhysX technology is disabled

    First, let’s look at the temperature performance of the Radeon HD4770 video card in comparison with the Radeon HD4830. The turbine was controlled in automatic mode.

    At rest, the core temperature of the RV740 is only slightly lower than that of the RV770LE.

    As soon as it came to heavy graphics, the new workflow showed itself in all its glory! The difference in temperature between the HD4830 and HD4770 reached 21 degrees Celsius, and that’s at noticeably higher clock speeds. Excellent result!

    FurMark does not spare video cards, warming them up to the highest possible temperature. The leadership of the RV740 is quite obvious here too — 92 degrees for the RV770LE against 71 for the RV740.

    During testing of temperature conditions, we were able to identify the strange behavior of the HD4770 turbine. Even when the core temperature does not reach the maximum of 71 degrees Celsius we noted, sometimes, literally for 1-2 seconds, the turbine speeds up, then calms down again. We cannot yet explain this fact.

    Finally, let’s move on to performance testing and analysis of the results.

    Let’s start with 3DMark Vantage.

    The first test is a surprise. The Radeon HD4770 performs on an equal footing with the GTS 250 and is quite a bit behind the more expensive fellow HD4850, leaving behind the HD4830 and even more so the HD4670, which turned out to be a complete outsider. Overclocking allows the HD4770 to take over the non-overclocked HD4850, which is very good considering the price of the novelty.

    The increase in resolution, detail and activation of full-screen anti-aliasing do not change the overall picture. In terms of performance, the Radeon HD4770 is almost on par with the GTS250 and slightly behind the HD4850.

    Flying over a tropical island did not change the picture much, except that the HD4770 edged out its older brother HD4850 ​​with a slight margin and lost a little to the GeForce GTS 250. In fairness, it must be said that the difference in the results is extremely small. Perhaps the increase in resolution will change the situation, we will check.

    Indeed, with increasing resolution, the scatter of results became noticeably larger. Formally, the leader was the HD4770, which outperformed the Radeon HD4850 ​​by only six points. Most likely, this test depends more on the clock frequency of the graphics core than on the memory bandwidth, which is theoretically higher in the HD4850. NVIDIA video cards lost their positions with the growth of resolution. Now the GTS 250 can compete on equal terms only with the HD4830, while 9The 800GT is noticeably ahead of the HD4670, while lagging behind the GTS 250 by more than 100 points.

    The heaviest mode brought the HD4770 and HD4850 ​​to the lead, which, in fact, turned out to be equal in performance. Then comes the GTS 250, HD4830 and 9800GT, and the HD4670 closes the circle. Synthetic tests outlined only part of the picture, the most important part of the plot will open only after passing real gaming tests, to which we turn.

    Crysis knows how to load the video system. Even at such a relatively low resolution, our testers are not able to provide a comfortable game. In terms of the minimum FPS, the Radeon HD 4850 was about 20% ahead of the HD 4770, and only overclocking allowed the newcomer to reach approximate parity with its older brother. Radeon HD 4830 is trying hard not to give up under pressure 9800GT, and he does it well. Well, the GTS 250 is on par with the HD4850.

    As the resolution grew, the overall alignment of forces changed little, only the absolute values ​​fell to an indecent level, which, however, is quite excusable for inexpensive video cards, especially considering the severity of graphic settings. The GeForce GTS 250, though slightly, is still ahead of the Radeon HD4770, while being on the same level as the HD4850, the 9800GT still fights with the HD4830, however, with an increase in resolution in terms of the minimum FPS, the latter turned out to be a little faster.

    Colorful and spectacular combat clashes in World in Conflict make you completely immerse yourself in the virtual world. The 1280×1024 resolution allows almost all test participants to provide an acceptable level of performance, only the HD4670 gave up ahead of time. GTS 250 is ahead of all competitors. The HD 4850 is second, followed by the Radeon HD4770, which even overclocked did not help to take the lead.

    Complicating the task for test participants, we observe their impotence in high resolutions. The minimum FPS is below the critical point, except that the GTS 250 survived and did not give up. The Radeon HD4770 is a little behind the HD4850, but overclocking equalized their chances.

    The test results for the HD 4850 and HD 4770 are a little surprising. There is no difference in speed. Most likely, the lack of functional units in the RV740 is compensated by the high core clock, which made it possible to catch up with the HD4850 ​​in speed.

    Looking at these results, the hand does not rise to write about a noticeable gap between 4770 and HD 4850. The difference is within the margin of error. But the GTS 250 is slightly faster than the HD4770 in terms of the minimum FPS.

    The H.A.W.X game engine is quite well optimized, since even the highest resolution does not reveal any problems with a comfortable game for the weakest representatives of our test subjects. Leadership behind NVIDIA solutions in all resolutions is simply amazing. Even the 9800GT outperformed all its competitors. As a result, AMD video cards can only compete with themselves. Among AMD solutions, the HD 4850 wins with a minimum margin in high resolutions, although it’s hard to call it a victory… HD4830 and HD4670, respectively, close the ranks.

    The Call of Juarez engine brought the HD 4850 to first place, and the difference with the closest competitor in the face of the HD4770 is more tangible than in some other tests.

    Well, perhaps, the victory in this game will be credited to AMD, but it’s too early to sum up the results.

    From the summary table with technical characteristics, it is clearly seen that the AMD Radeon HD 4770, although equipped with the most modern type of video memory of the GDDR-5 standard, has a video memory bus width of only 128 bits. Not so long ago, video cards with such a memory bus width were considered almost completely unsuitable for games with activated full-screen anti-aliasing and anisotropic filtering modes, all due to insufficient video memory bandwidth. Now we have to check how efficiently the 128-bit AMD Radeon HD 4770 memory controller works. All the same participants act as a benchmark, with the exception of the Radeon HD 4670, which, as practice has shown, is a clear outsider in today’s race.

    The inclusion of full-screen anti-aliasing significantly affected the performance of all test participants, however, the decrease in the performance of the Radeon HD4770 cannot be called critical. The lag behind the older brother Radeon HD 4850 with a wider memory bus is small. The GeForce GTS 250 is slightly ahead of the HD4770, while being slightly behind the Radeon HD4850 ​​and the overclocked Radeon HD4770.

    The increase in resolution did not change the situation in any way, except that the results of NVIDIA video cards slowly dropped to the same performance level as the Radeon HD4830.

    These results only serve as an additional confirmation of the previously obtained data. The performance of the Radeon HD4770 is approximately on the same level as the Radeon HD4850, and overclocking completely eliminates this meager gap. The results of the GeForce GTS 250 are on the same level with the Radeon HD4770. In its turn, the 9800GT lags far behind its competitor.

    The rise in resolution in Crysis makes the gap between the HD 4850 and the Radeon HD 4770 more pronounced. It’s just that you definitely won’t be able to play comfortably, just like on the GTS 250, the performance of which is only enough to stand on the same level as the HD 4850.

    The inclusion of full-screen anti-aliasing and anisotropic filtering in this game again confirms that there is no big difference between the Radeon HD 4850 and HD 4770. In the heavy graphics mode World in Conflict, NVIDIA solutions feel somewhat more confident than AMD products. Here the Radeon HD 4770 shows the results at the level of the GeForce 9800GT, while the GTS 250 is at the level of the HD 4850 or slightly ahead of it.

    FarCry2 with anti-aliasing and anisotropic filtering also doesn’t show anything new in the balance of power between AMD video cards. When comparing the performance of AMD and NVIDIA cards, it is clear that the GTS 250 outperforms not only the HD 4770, but also the HD 4850, well, 9800GT, on the contrary, is inferior to all Radeons.

    The alignment of forces between the decisions of the opposing graphics camps in this game is already familiar. Well, the relative positions of the results of Radeon HD 4850 and HD 4770 are preserved here, the difference is rather small.

    The Call of Juarez engine has a different opinion on the balance of power than HAWX. Here, all AMD solutions are ahead, including the Radeon HD 4830, while the performance of the HD 4770 still does not cause concern and is about or slightly lower than the HD 4850. Now we can sum up.


    The AMD Radeon HD 4770 really made an impression, and a very positive one at that. The latest technical process, low heat, good overclocking potential, coupled with decent performance make the Radeon HD4770 a good contender for purchase.