ATI Radeon HD 5850 Crossfire Review :: Features

11-04-2009 · Category: Hardware - Video Cards

By Benjamin Sun
  • 2.15 billion 40nm transistors
  • TeraScale 2 Unified Processing Architecture
  • 1440 Stream Processing Units
  • 72 Texture Units
  • 128 Z/Stencil ROP Units
  • 32 Color ROP Units
  • GDDR5 interface with 128.0 GB/sec of memory bandwidth
  • PCI Express 2.1 x16 bus interface
  • DirectX® 11 support
  • Shader Model 5.0
  • DirectCompute 11
  • Programmable hardware tessellation unitAccelerated multi-threading
  • HDR texture compression
  • Order-independent transparency
  • OpenGL 3.2 support
  • Image quality enhancement technology
  • Up to 24x multi-sample and super-sample anti-aliasing modes
  • Adaptive anti-aliasing
  • 16x angle independent anisotropic texture filtering
  • 128-bit floating point HDR rendering
  • ATI Eyefinity multi-display technology
  • Three independent display controllers
  • Drive three displays simultaneously with independent resolutions, refresh rates, color controls, and video overlays
  • Display grouping
  • Combine multiple displays to behave like a single large display
  • ATI Stream acceleration technology
  • OpenCL 1.0 compliant
  • DirectCompute 11
  • Double precision floating point processing support
  • Accelerated video encoding, transcoding, and upscaling
  • ATI CrossFireX™ multi-GPU technology
  • Dual, triple, and quad GPU scaling
  • Dual-channel bridge interconnect
  • ATI Avivo™ HD Video & Display technology
  • UVD 2 dedicated video playback accelerator
  • Advanced post-processing and scaling
  • Dynamic contrast enhancement and color correction
  • Brighter whites processing (Blue Stretch)
  • Independent video gamma control
  • Dynamic video range control
  • Support for H.264, VC-1, and MPEG-2
  • Dual-stream 1080p playback support
  • DXVA 1.0 & 2.0 support
  • Integrated dual-link DVI output with HDCP
  • Max resolution: 2560x1600
  • Integrated DisplayPort output
  • Max resolution: 2560x1600
  • Integrated HDMI 1.3 output with Deep Color, xvYCC wide gamut support and high bit-rate audio
  • Max resolution: 1920x1200
  • Integrated VGA output
  • Max resolution: 2048x1536
  • 3D stereoscopic display/glasses support
  • Integrated HD audio controller
  • Output protected high bit rate 7.1 channel surround sound over HDMI with no additional cables required
  • Supports AC-3, AAC, Dolby TrueHD and DTS Master Audio formats
  • ATI PowerPlay™ power management technology
  • Dynamic power management with low power idle state
  • Ultra-low power state support for multi-GPU configurations
  • Certified drivers for Windows 7, Windows Vista, and Windows XP

HD 5850 Specification

  • Engine clock speed: 725 MHz
  • Processing power (single precision): 2.09 TeraFLOPS
  • Processing power (double precision): 418 GigaFLOPS
  • Polygon throughput: 725M polygons/sec
  • Data fetch rate (32-bit): 209 billion fetches/sec
  • Texel fill rate (bilinear filtered): 52.2 Gigatexels/sec
  • Pixel fill rate: 23.2 Gigapixels/sec
  • Anti-aliased pixel fill rate: 92.8 Gigasamples/sec
  • Memory clock speed: 1 GHz
  • Memory data rate: 4.0 Gbps
  • Memory bandwidth: 128 GB/sec
  • Maximum board power: 151 Watts
  • Idle board power: 27 Watts

ATI’s HD 5850 card is based upon the Cypress Pro chip. This chip has 2.15 Billion transistors and is manufactured on TSMC’s 40nm fabrication process. The chip is around 334mm2 which is quite small for such a large number of transistors. ATI says that their HD 5850 has up to 2.09 Teraflops of computing power single precision. Two of the HD 5850s in CrossfireX mode output up to 4.18 TeraFlops.


ATI Radeon HD 5850 Crossfire Review Drivers

Drivers

ATI Radeon HD 5850 Crossfire Review Drivers

Drivers

ATI Radeon HD 5850 Crossfire Review Drivers

Drivers

ATI Radeon HD 5850 Crossfire Review Drivers

Drivers


The HD 5850 is clocked at 725MHz for the core and 4GHz effective for the memory. The key feature here is that the memory is clocked at 1GHz, but doubled for double data rate and doubled again for GDDR5 memory. This provides each card 128GB of memory bandwidth. The Texel fillrate is 52.2 Gigatexels and the pixel fill rate is 23.2 Gigapixels a second.

ATI introduced Crossfire in 2005 with the release of the X800 cards. The first generation of Crossfire required a Master and a Slave video card, an external Crossfire cable and a Crossfire compliant motherboard with two PCIe slots as the cards needed a composite chip to run in Crossfire mode. The X850XT Master card used a Silicon Image compositing chip (Sil 163 TMDS) and the maximum resolution of Crossfire first generation was 1600x1200 at 60Hz or 1920x1440 at 52Hz meaning that most people needing high resolution gaming at high refresh rates chose another solution. ATI removed the limitation with the X1800 series as it required no Master card and there wasn’t a compositing chip.

With the advent of the X1800 series the Crossfire dongle was no longer required. What you needed to achieve this was a CrossFire Xpress 3200 motherboard and two ATI RADEON cards that communicated over the PCI Express bus removing the compositing chip from the card. Today’s implementation of Crossfire on all ATI HD 2xxx series and above has implemented a bridge system much like NVIDIA’s SLI implementation. Two cards or more (Up to four), assuming the motherboard supports four video cards, and Crossfire, can be daisy-chained to each other improving performance over a single video card.

ATI uses three methods to render with CrossfireX: Scissors, Checker board mode and Alternate Frame Rendering Mode. Scissors mode splits the frame into half with one video card rendering the top half of the image and the other rendering the bottom half of the image with three cards splitting the screen into three and four into quarters. Checker board has the image split into smaller squares which are assigned to different cards. Alternate Frame Rendering has the first video card render one whole frame, the second the next frame, etc. In most cases Scissors mode is the most inefficient as some parts of a rendered scene are more complex than others, meaning that one card may render part of a scene faster than another with the card waiting on the second card to render its part of the scene. AFR is most useful because the image usually doesn’t change much from one frame to the next, meaning the workload is mostly similar so having one card render an image then the next card render a slightly different image is more efficient than having one card render part of a scene as parts of the scene can have much more geometry or pixels than others.