ATI Radeon HD 5870: DirectX 11, Eyefinity, And Serious Speed

Double Or Nothing

It’s Good To Set Goals

ATI says it approached this design with five different goals. First on the list, naturally, was incorporating DirectX 11 support to coincide with the launch of Windows 7. The timing there couldn’t have been much better, as Micrososft’s next-gen operating system is RTM and on the verge of retail availability.

Second, it wanted to improve performance in DirectX 9, 10, and 10.1 titles. Because DirectX 11 games aren’t shipping yet, the company knew its “legacy” capabilities would be the benchmark by which it’d be measured for many months after launch.

Third, the company had an eye on stream computing. This is an area Nvidia’s CUDA architecture has outright dominated since inception. With OpenCL 1.0 and DirectCompute now standardizing the way developers handle GPGPU functionality, this is ATI’s first chance to really step out.

Fourth, it shot for two times the processing power of its previous generation in a comparable power envelope. According to ATI’s own measurement, it achieved that goal. And while maximum TDP is actually higher this time around, idle power is significantly lower.

Finally, ATI’s architects sought innovation, achieved through Cypress’ display output configuration and certain image quality enhancements.  

How Do You Double Performance?

Perhaps the easiest way to double the processing power of a GPU is by doubling the resources most likely to affect performance. The result is 2.7 TeraFLOPS single-precision and 544 GigaFLOPS double-precision performance.


Radeon HD 5870
Radeon HD 4870
Die Size
334 square millimeters
263 square millimeters
Transistors
2.15 billion
.956 billion
Memory Bandwidth
153 GB/s
115 GB/s
AA Resolve
128
64
Z/Stencil
128
64
Texture Units
80
40
Shader (ALUs)
1,600
800
Idle Board Power
27W
90W
Active Board Power
188W
160W


Whereas the RV770 had 10 SIMD cores, Cypress sports 20. As before, each core contains 16 stream processor units. And each stream processor boasts five ALUs, which ATI calls stream cores. Multiply those out and you get 1,600 total stream cores or shaders. Sixteen hundred shaders times 850 MHz times two FLOPS gives you that 2.7 TFLOPS measurement, all else being perfect.

Block diagram of CypressBlock diagram of Cypress

As with the generation prior, texture units are tied to the SIMD arrays—four per engine. With 20 arrays, that’s 80 total texture units. Of course, RV770 featured 40.

And though they also look fairly similar on a full-size die shot, Cypress’ render back-ends are also significantly improved. This part of the chip was a concern back when ATI first introduced us to its RV770 architecture. But GDDR5 memory helped mitigate the effects of stepping down to an aggregate 256-bit memory bus. Moreover, improvements to anti-aliasing performance and Z/stencil rate demonstrated that ATI had fixed much of what was “broken” on RV670.

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
214 comments
    Your comment
    Top Comments
  • hispeed120
    I'm. So. Excited.
    26
  • cangelini
    viper666why didn't they thest it against a GTX 295 rather than 280??? its far superior...


    Ran it against a GTX 295 and a 285 and 285s in SLI :)
    23
  • megamanx00
    O M F G!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    Just wish the darn thing wasn't so big, but man, what a card! Now I'm thinking about a bigger case :D
    22
  • Other Comments
  • hispeed120
    I'm. So. Excited.
    26
  • Anonymous
    Can't wait
    9
  • crosko42
    So it looks like 1 is enough for me.. Dont plan on getting a 30 inch monitor any time soon.
    21
  • jezza333
    Looks like the NDA lifted at 11:00PM, as there's a load of reviews now just out. Once again it shows that AMD can produce a seriously killer card...

    Crysis 2 on an x2 of this is exactly what I'm waiting for.
    20
  • woostar88
    This is incredible at the price point.
    8
  • LORD_ORION
    Err... I thought I was going to see more for the price. Regardless, I think ATI missed the mark here. I am interested in playing games on my HDTV since me and my monitor don't care about these higher resolutions. Fail cakes... Nivida is undoubtedly going to rape ATI in performance with the 300 series. This is good news for mainstream prices however.... you can ptobably upgrade to a current DX10 board soon for a very good price, and then buy a 5850 for $100 in a year from now. Result? Don't but a 5000 series card yet until the price comes down? Heh, I bet the cards will be $100 less in December if the 300 series launches.

    This is not to say I am an Nvidia fan, just undoubtedly you would do well for yourself to hold off for a bit if you want to buy a 5000 series... as the price will come down for a good price/performance ratio soon enough.
    -41
  • tipmen
    wait, wait, before I look can it play cry... HOLY SHIT?!
    20
  • viper666
    why didn't they thest it against a GTX 295 rather than 280??? its far superior...
    -35
  • cangelini
    viper666why didn't they thest it against a GTX 295 rather than 280??? its far superior...


    Ran it against a GTX 295 and a 285 and 285s in SLI :)
    23
  • Annisman
    I refuse to buy until the 2GB versions come out, not to mention newegg letting you buy more than 1 at a time, paper launch ftl.
    2
  • jasperjones
    Thanks for the timely review. I have to say though, some of the technical details are beyond me. It'd be useful if you explained terms such as "VLIW architecture" or "tessellation engine"
    15
  • viper666
    oh my bad... didn't see the rest of the pages :)
    -14
  • megamanx00
    O M F G!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    Just wish the darn thing wasn't so big, but man, what a card! Now I'm thinking about a bigger case :D
    22
  • Annisman
    Oops, who am I kidding ? I just ordered 2 5870's. One Sapphire, and one HIS, seeing as how they limit you to one per customer.
    17
  • falchard
    I think most of this review has to do with how many games are optimized for nVidia. The Crytek Engine 2.0 and Source Engine are well known for heavily favoring nVidia architecture yet compose the bulk of the benchmarks. I think the fact ATI can do best in these engines when they have a detect ATI instant nerf its performance speaks measures for the actual card.
    18
  • charlesxuma
    I WANT MY MOOOMMYYYYYY !!!!!!!
    -10
  • tipmen
    Another thing is that the 5800x2 isn't out yet, now think of two of those bad boys in Crossfire.
    14
  • blackbyron
    Not bad for Crysis benchmark. I really want 5870 for my christmas present, but damn I also need to buy a new PSU.
    6
  • blackbyron
    In addition, I am impressed that the 5870 has a better power consumption and better gaming performance compare to DX10 cards. If the card is affordable I'd definite buy one.
    12
  • cangelini
    jasperjonesThanks for the timely review. I have to say though, some of the technical details are beyond me. It'd be useful if you explained terms such as "VLIW architecture" or "tessellation engine"


    Jasper,
    TBH, the architectural details are secondary to how the card performs. However, if you'd like a better idea of what tessellation can do for you, check out the picture of the Alien on page six!
    3