Logo   Banner   TopRight
TopUnder
Transparent
ATI Radeon HD 4890: The RV790 Unveiled
Transparent
Date: Apr 02, 2009
Section:Graphics/Sound
Author: Marco Chiappetta
Transparent
Introduction and Specifications

As anyone remotely in tune with the tech sector can attest, the rivalry in the PC graphics card market between AMD / ATI and NVIDIA is as intense as ever. It used to be that one of the two companies would release a new product, only to have the other release a competing offering a few weeks, or maybe a few months later. But even in these gloomy economic times, AMD and NVIDIA continue to fight on and today both graphics giants are releasing new graphics cards aimed squarely at one another. Not a few weeks apart, but simultaneously on the very same day. Don't believe us? See here for our NVIDIA GeForce GTX 275 coverage.

AMD is rolling out the ATI Radeon HD 4890 today, technically a new graphics card, but one that borrows heavily from the previous generation. The Radeon HD 4890 is based on an updated variant of the popular RV770 GPU which powers Radeon HD 4850 and 4870 cards, dubbed the RV790. We've got some technical details regarding the RV790 GPU below and have more particulars regarding the actual cards and performance on the pages ahead. Read on to see what AMD has in store with the brand new Radeon HD 4890...


Asus ATI Radeon HD 4890 (RV790)

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/GDDR5 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

    • 800 stream processing units

      • Dynamic load balancing and resource allocation for vertex, geometry, and pixel shaders
      • Common instruction set and texture unit access supported for all types of shaders
      • Dedicated branch execution units and texture address processors
    • 128-bit floating point precision for all operations
    • Command processor for reduced CPU overhead
    • Shader instruction and constant caches
    • Up to 160 texture fetches per clock cycle
    • Up to 128 textures per pixel
    • Fully associative multi-level texture cache design
    • DXTC and 3Dc+ texture compression
    • High resolution texture support (up to 8192 x 8192)
    • Fully associative texture Z/stencil cache designs
    • Double-sided hierarchical Z/stencil buffer
    • Early Z test, Re-Z, Z Range optimization, and Fast Z Clear
    • Lossless Z & stencil compression (up to 128:1)
    • Lossless color compression (up to 8:1)
    • 8 render targets (MRTs) with anti-aliasing support
    • Physics processing support
  • 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 support
  • 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 1920x1200 (single-link DVI) or 2560x1600 (dual-link DVI)
      • Secondary supports 18-, 24-, and 30-bit digital displays at all resolutions up to 1920x1200 (single-link DVI only)3
      • 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 2048x15363
    • DisplayPort output support

      • Supports 24- and 30-bit displays at all resolutions up to 2560x16003
    • HDMI output support

      • Supports all display resolutions up to 1920x10803
      • 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 outputs
    • ATI CrossFireX Multi-GPU Technology

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

     


  •   
    AMD RV790 / Radeon HD 4890 Details

    The new Radeon HD 4890 is technically based on a new GPU design, which was formerly codenamed RV790, but fundamentally it is very similar to the RV770-based Radeon HD 4870. The two GPUs share the same feature set, same compliment of shader processors (800), texture units, and ROP configurations. Because we've covered the architectural details of the RV770 in a previous article, we won't do the same again here. If you'd like a little refresher, however, we'd recommended checking out our Radeon HD 4800 series launch article from last year. All of the details necessary to understand what's going on under the Radeon HD 4890's hood are in that article.

    Although the features are technically very similar, the RV790 does differ from the RV770 in a number of ways. Some may be tempted to say the RV790 is simply an overclocked RV770, but that is not the case. The RV790 at the heart of the Radeon HD 4890 is a new chip that is comprised of more transistors than the RV770 (959M vs. 956M). The additional transistors are dedicated to updates made to the core to support higher clock speeds. According to AMD, the entire chip was re-timed and power distribution on the chip was altered. In addition, a decap ring--or ring of decoupling capacitors--was implemented around the chip to reduce signal noise. The changes to the core and the addition of the decap ring result in an increase in die and packaging size, but the end result is a GPU that can run at much higher clock speeds than the RV770.

    Transparent
    Asus Radeon HD 4890

    For the purpose of this article, we got our hands on a couple of Asus-built Radeon HD 4890 cards--one standard reference card and one overclocked edition.

        

    Asus calls their Radeon HD 4890 cards the EAH4890 (standard reference card) and the EAH4890 TOP (overclocked edition). Before we get to the particulars of each, we wanted to detail the accessory and software bundle Asus offers with the cards. Included with both the EAH4890 and EAH4890 TOP are driver and utility discs, a leather mouse pad, an HD component output dongle, an S-Video to Composite adapter, a CrossFire bridge connector, a dual 4-pin to Molex to 6-pin PCI Express power adapter, and DVI to VGA and HDMI (with audio) adapters. You'll also notice a large "Voltage Tweak" badge emblazoned on the box. Asus is also offering an updated version of their SmartDoctor utility with the cards, which gives users the ability to alter the GPU voltages for extreme overclocking. Cool stuff. 

        

        

    Physically, the Radeon HD 4890 is similar to the Radeon HD 4870. The Radeon HD 4890 features a dual-slot cooling solution, and requires a pair of PCI Express 6-pin power feeds. Max board power has been increased to 190 W, but idle power should be somewhat lower than the 4870 according to AMD.

    Along the top edge of the PCB you can see the dual CrossFire connectors, which allow the cards to be used in 2-, 3-, or 4-way CrossFireX configurations, and the outputs on the 4890 are identical to the 4870--dual Dual-Link DVI outputs and an HDTV output.

    Clock speeds on stock reference Radeon HD 4890 cards like the EAH4890 will be 850MHz for the GPU and 975MHz for the GDDR5 RAM (3.9 Gbps effective). At those clocks, cards will offer 1.36 TFLOPS of compute performance, a pixel fillrate of 13.6 Gpixels/s, and 124.8GB/s of memory bandwidth, on a 256-bit memory bus. Overclocked cards like the EAH4890 TOP will also be available from a few of AMD's add-in board partners, which will likely increase GPU and memory clocks to the 900MHz+ and 1000MHz+ plus range. The EAH4890 TOP, for example, has a GPU clock of 900MHz and a memory clock of 1000MHz.

        

    With the cooler removed, you can get a good look at the RV790 GPU at the heart of the Radeon HD 4890, and the eight Qimonda GDDR5 memory chips surrounding the GPU. The cooler features a large copper heatsink GPU, with heat-pipes linked to secondary aluminum heatsinks. The memory and the GPU both make contact with the cooler, which exhausts hot air from the system through vents in the card's case bracket.

    Transparent
    Our Test System and 3DMark Vantage

    HOW WE CONFIGURED THE TEST SYSTEM: We tested the graphics cards in this article on an Gigabyte GA-EX58 Extreme motherboard powered by a Core i7 920 quad-core processor and 6GB of OCZ DDR3 RAM. The first thing we did when configuring these test system was enter the system BIOS and set all values to their "optimized" or "high performance" default settings. Then we manually configured the memory timings and disabled any integrated peripherals that wouldn't be put to use. The hard drive was then formatted, and Windows Vista Ultimate SP1 was installed. When the installation was complete we fully updated the OS, and installed the latest DX10 redist and various hotfixes, along with the necessary drivers and applications.

    HotHardware's Test Systems
    Core i7 Powered

    Hardware Used:
    Core i7 920 (2.66GHz)

    Gigabyte EX58-Extreme
    (X58 Express Chipset)

    Radeon HD 4890 OC (EAH4890 TOP)
    Radeon HD 4890 (EAH4890)
    Radeon HD 4870 1GB
    GeForce GTX 285
    GeForce GTX 275
    GeForce GTX 260 Core 216

    6GB OCZ DDR3-1066 C7
    (3 X 2GB)

    Integrated Audio
    Integrated Network

    Western Digital "Raptor" 150GB
    (10,000RPM - SATA)

    Relevant Software:
    Windows Vista Ultimate SP1
    DirectX March 2009 Redist

    ATI Catalyst v9.3b
    NVIDIA GeForce Drive v185.63

    Benchmarks Used:

    3DMark Vantage v1.0.1
    FarCry 2
    Crysis*
    Left 4 Dead*
    Enemy Territory: Quake Wars v1.5*

    * - Custom benchmark

    Futuremark 3DMark Vantage
    Synthetic DirectX Gaming


    3DMark Vantage

    The latest version of Futuremark's synthetic 3D gaming benchmark, 3DMark Vantage, is specifically bound to Windows Vista-based systems because it uses some advanced visual technologies that are only available with DirectX 10, which y isn't available on previous versions of Windows.  3DMark Vantage isn't simply a port of 3DMark06 to DirectX 10 though.  With this latest version of the benchmark, Futuremark has incorporated two new graphics tests, two new CPU tests, several new feature tests, in addition to support for the latest PC hardware.  We tested the graphics cards here with 3DMark Vantage's Performance preset option, which uses a resolution of 1280x1024 no anti-aliasing and trilinear filtering.



    The NVIDIA powered cards sweep the top three spots in our 3DMark Vantage tests, besting all of the AMD offerings. NVIDIA's latest Release 185 drivers give the cards a nice boost in performance here, that allows even the GeForce GTX 260 Core 216 to pull ahead of the brand new Radeon HD 4890 in 3DMark Vantage.

    Transparent
    Enemy Territory: Quake Wars

    Enemy Territory: Quake Wars
    OpenGL Gaming Performance


    Enemy Territory:
    Quake Wars

    Enemy Territory: Quake Wars is Based on a radically enhanced version of id's Doom 3 engine and viewed by many as Battlefield 2 meets the Strogg, and then some.  In fact, we'd venture to say that id took EA's team-based warfare genre up a notch or two.  ET: Quake Wars also marks the introduction of John Carmack's "Megatexture" technology that employs large environment and terrain textures that cover vast areas of maps without the need to repeat and tile many smaller textures.  The beauty of megatexture technology is that each unit only takes up a maximum of 8MB of frame buffer memory.  Add to that HDR-like bloom lighting and leading edge shadowing effects and Enemy Territory: Quake Wars looks great, plays well and works high end graphics cards vigorously.  The game was tested with all of its in-game options set to their maximum values with soft particles enabled in addition to 4x anti-aliasing and 16x anisotropic filtering.

    The NVIDIA powered cards had another strong showing in our Enemy Territory: Quake Wars testing. Here, the Radeon HD 4890 overclocked edition is able to pull ahead of the GeForce GTX 260 Core 216, as is the reference clocked card at 2560. But the reference GeForce GTX 275 is able to pull ahead of both--the margin of victory at 2560x1600 is exactly 1FPS.

    Transparent
    Crysis v1.21

    Crysis v1.21
    DirectX 10 Gaming Performance


    Crysis

    If you're at all into enthusiast computing, the highly anticipated single player, FPS smash-hit Crysis, should require no introduction. Crytek's game engine produces some stunning visuals that are easily the most impressive real-time 3D renderings we've seen on the PC to date.  The engine employs some of the latest techniques in 3D rendering like Parallax Occlusion Mapping, Subsurface Scattering, Motion Blur and Depth-of-Field effects, as well as some of the most impressive use of Shader technology we've seen yet.  In short, for those of you that want to skip the technical jib-jab, Crysis is a beast of a game.  We ran the full game patched to v1.21 with all of its visual options set to 'Very High' to put a significant load on the graphics cards being tested  A custom demo recorded on the Ice level was used throughout testing.

    Score one for AMD here, sort of. In our custom Crysis benchmark, the Radeon HD 4890 OC bests the GeForce GTX 275. The reference clocked Radeon HD 4890 also outpaces the new GTX 275, but only by the smallest of margins. The Radeon HD 4890 and GeForce GTX 275 couldn't be more closely matched then they are in Crysis.

    Transparent
    FarCry 2

    FarCry 2
    DirectX Gaming Performance


    FarCry 2

    Like the original, FarCry 2 is one of the more visually impressive games to be released on the PC to date.  Courtesy of the Dunia game engine developed by Ubisoft, FarCry 2's game-play is enhanced by advanced environment physics, destructible terrain, high resolution textures, complex shaders, realistic dynamic lighting, and motion-captured animations.  We benchmarked the graphics cards in this article with a fully patched version of FarCry 2, using one of the built-in demo runs recorded in the "Ranch" map.  The test results shown here were run at various resolutions with 4X AA and No anisotropic enabled concurrently.

    Our FarCry 2 testing also placed the GeForce GTX 275 ahead of AMD's brand new Radeon HD 4890, whether or not it is an overclocked edition.

    Transparent
    Left 4 Dead

    Left 4 Dead
    DirectX Gaming Performance


    Left 4 Dead

    Left 4 Dead is a co-operative, survival horror, first-person shooter that was developed by Turtle Rock Studios, which was purchased by Valve part-way into development. Like Half Life 2, the game uses the Source engine, however, the visual in L4D are far superior to anything seen in the Half Life universe to date. The game pits four Survivors of an apocalyptic pandemic against hordes of aggressive zombies. We tested the game at resolutions of 1,920 x 1,200 and 2,560 x 1,600 with 4x anti-aliasing and 16x anisotropic filtering enabled and all in game graphical options set to their maximum values.

    Although the framerates at both resolutions are somewhat higher, the results of our Left 4 Dead testing look very similar to those of FarCry 2 on the previous page. Once again, the GeForce GTX 275 outpaces the Radeon HD 4890 overclocked and reference clocked editions here.

    Transparent
    Overclocking The Radeon HD 4890

    For our next set of performance metrics, we spent some time overclocking the new Radeon HD 4890 using the Overdrive utility built into ATI's Catalyst drivers.

    Overclocking The Radeon HD 4890
    Pedal To The Metal



    Using the auto-tune utility built right into the Overdrive tab, we were very easily able to take the reference Radeon HD 4890 (Asus EAH4890) up from its default CPU and Memory clock speeds of 850MHz and 975MHz, respectively, to an impressive 975MHz and 1035MHz. With the card overclocked, we re-ran a couple of benchmarks and saw significant performance improvements, as you might expect. The framerate in our ET:QW benchmark went up by 7.9 FPS (8.8%) and our L4D result increased by 8.74 FPS (10%). With this kind of headroom in a reference cards, we're eager to see what some board parters are able to do with custom designs that feature more elaborate cooling solutions.

    We should also point out that Asus' Radeon HD 4890 offerings also include a copy of the company's SmartDoctor Utility that gives uses the ability to tweak voltages for even better overclocks. We weren't able to test SmartDoctor here due to time constraints, but suspect GPU speeds in excess of 1GHz are possible with some additional effort.

    Transparent
    Power Cosumption and Noise

    We'd like to cover a few final data points before bringing this article to a close. Throughout all of our benchmarking and testing, we monitored how much power our test systems were consuming using a power meter. Our goal was to give you an idea as to how much power each configuration used while idling and under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the motherboards alone.

    Total System Power Consumption
    Tested at the Outlet

    Although AMD claimed the Radeon HD 4890 would consume up to 30 fewer watts at idle and 30 more watts under load than the 4870, our testing did not prove out those claims. According to our tests, the Radeon HD 4890 actually consumed slightly less power at idle and under load. The differences were minimal, however.

    While the power consumption numbers put the 4890 and 4870 in a similar league, we can say the 4890 runs much cooler than the 4870. Our Asus Radeon HD 4890 cards idled at under 50'C and peaked in the mid 80's under load. But where we noticed the biggest difference was when we removed the cards from the test system. After testing, the Radeon HD 4890 felt very warm to the touch when removing it from the test rig, but the Radeon HD 4870 was so much hotter, it was almost too hot to handle without giving it a few minutes to cool down. Not exactly a scientific data point, but something we wanted to mention nonetheless.

    We'd also like to mention that the Radeon HD 4890 is a relatively quiet running card. At idle, the fan on the card is barely audible over the other components in our test system. Under load, the fan does spin up and is audible over the other components, but we would not consider it loud by any means.

    Transparent
    Our Summary and Conclusion

    Performance Summary: Summarizing the performance of the ATI Radeon HD 4890, as it relates to AMD's previous generation HD 4870, is easy. Due to the its increased GPU core and memory clock speeds, the Radeon HD 4890 is the fastest, single-GPU powered graphics card AMD / ATI has released to date. In every test, the Radeon HD 4890 (Asus EAH4890) was faster than the 1GB Radeon HD 4870, and the overclocked 4890 (Asus EAH4890 TOP) simply increased the card's overall lead. In comparison to competing offerings from NVIDIA, the Radeon HD 4890 is faster than the GeForce GTX 260 Core 216 overall, but it didn't quite keep pace with the just announced GeForce GTX 275.

     

    Exact pricing for the various ATI Radeon HD 4890 cards that should start to become available sometime later today is yet to be nailed down. According to the most recent information we have, pricing for reference clocked standard edition Radeon HD 4890 cards should be about $249, but mail in rebates should bring the final priced down to about $229. We are still waiting to confirm what AMD's add-in board partners will be selling Radeon HD 4890 overclocked edition parts for, but we're hearing $269 is the target, with mail in rebates bringing the final price down to about $249. Considering what we've seen from the Radeon HD 4890, the cards offers a lot of bang for the buck, especially if you're the overclocking type. The RV790 GPU powering the Radeon HD 4890 looks to have plenty of frequency headroom left for overclocking, which will surely appeal to the modders out there.

    It is not a completely new design, but the Radeon HD 4890 is an exciting product nonetheless. To put it simply, the Radeon HD 4890 is the fastest, single-GPU powered graphics card AMD has ever produced. And its competitive pricing and overclocking headroom should further its appeal amongst enthusiasts. It may not be the fastest card in AMD's graphics card line--that distinction still belongs to the Radeon HD 4870 X2--but could a dual-RV790 GPU based graphics card be that far out for AMD? Time will tell, of course. For now, the battle between AMD and NVIDIA rages on with a fresh pair of combatants that offer value propositions that are sure to please potential consumers.

         
    • Strong performer
    • DX10.1 Support
    • Great overclocker
    • Relatively quiet
    • Competitive pricing
    • Not quite as fast as the GTX 275
    • We prefer lower streat prices to MIRs



    Content Property of HotHardware.com