AMD's "Small-Die" Strategy

We outlined AMD's "new" GPU strategy in our Radeon HD 4850 preview article, but in short AMD has committed to designing GPUs for the mainstream $199 - $299 segment and simply using CrossFire (multi-GPU) to address higher end markets. NVIDIA on the other hand will continue to make very large monolithic GPUs in order to continue to push the industry forward. Both approaches are appreciated and necessary, they simply target different markets.

In our GT200 review we highlighted the fact that NVIDIA had built an extremely large, highly parallel, microprocessor. With 1.4 billion transistors and a die size of around 576 mm^2, NVIDIA's GT200 is nothing short of huge.

The table on the previous page shows that AMD's RV770, despite being aimed at mainstream gamer price points ($199 - $299), is also very large. At 956M transistors, the RV770 has 44% more transistors than RV670 and 68% the transistor count of NVIDIA's GT200. We threw the RV770 into NVIDIA's die size comparison just for kicks:


Based on what we know of NVIDIA's die size, this should be to scale

Even AMD's die, although designed to be svelte and affordable, is big - especially for being fabbed at TSMC. NVIDIA still holds the crown for largest die fabbed at TSMC, but AMD shows us that even a more mainstream approach still requires tons of transistors. As we mentioned in our 4850 preview:

"A pair of RV770s, AMD's new GPU, end up consuming more power than a single GT200 - despite being built on a smaller 55nm process.

A pair of these RV770s only costs $400 compared to $650 for a single GT200, but I suspect that part of that is due to differences in manufacturing process. If NVIDIA hadn't been so risk averse with the GT200 and built it on 55nm (not that I'm advocating it, simply posing a hypothetical), the cost differences would be smaller - if not in favor of NVIDIA since GT200 is built on a single card.

When the smoke clears, AMD's strategy is to simply build a GPU for the masses and attempt to scale it up and down. While NVIDIA is still building its GPUs the same way it has for decades, starting very large and scaling down.

AMD isn't taking a radically different approach to building and designing GPUs than NVIDIA, it's simply building one market segment lower."

We've got a lot of discussion on efficiency between AMD and NVIDIA coming up in this article, although AMD's die is noticeably smaller than NVIDIA's - as you've already seen with the Radeon HD 4850 - there are many areas where RV770 can go toe-to-toe with NVIDIA's mammoth GT200.

Index Building a RV770
POST A COMMENT

173 Comments

View All Comments

  • jay401 - Wednesday, June 25, 2008 - link

    Good but I just wish AMD would give it a full 512-bit memory bus bandwidth. Tired of 256-bit. It's so dated and it shows in the overall bandwidth compared to NVidia's cards with 512-bit bus widths. All that fancy GDDR4/5 and it doesn't actually shoot them way ahead of NVidia's cards in memory bandwidth because they halve the bus width by going with 256-bit instead of 512-bit. When they offer 512-bit the cards will REALLY shine. Reply
  • Spoelie - Thursday, June 26, 2008 - link

    Except that when R600 had a 512bit bus, it didn't show any advantage over RV670 with a 256bit bus. And that was with GDDR3 vs GDDR3, not GDDR5 like in RV770 case. Reply
  • JarredWalton - Thursday, June 26, 2008 - link

    R600 was 512-bit ring bus with 256-bit memory interface (four 64-bit interfaces). http://www.anandtech.com/showdoc.aspx?i=2552&p...">Read about it here for a refresh. Besides being more costly to implement, it used a lot of power and didn't actually end up providing provably better performance. I think it was an interesting approach that turned out to be less than perfect... just like NetBurst was an interesting design that turned out to have serious power limitations. Reply
  • Spoelie - Thursday, June 26, 2008 - link

    Except that it was not, that was R520 ;) and R580 is the X19x0 series. That second one proved to be the superior solution over time.

    R600 is the x2900xt, and it had a 1024bit ring bus with 512bit memory interface.
    Reply
  • DerekWilson - Sunday, June 29, 2008 - link

    yeah, r600 was 512-bit

    http://www.anandtech.com/showdoc.aspx?i=2988&p...">http://www.anandtech.com/showdoc.aspx?i=2988&p...

    looking at external bus width is an interesting challenge ... and gddr5 makes things a little more crazy in that clock speed and bus width can be so low with such high data rates ...

    but the 4870 does have 16 memory modules on it ... so that's a bit of a barrier to higher bit width busses ...
    Reply
  • JarredWalton - Wednesday, June 25, 2008 - link

    I'd argue that the 512-bit memory interface on NVIDIA's cards is at least partly to blame for their high pricing. All things being equal, a 512-bit interface costs a lot more to implement than a 256-bit interface. GDDR5 at 900MHz is effectively the same as GDDR3 at 1800MHz... except no one is able to make 1800MHz GDDR3. Latencies might favor one or the other solution, but latencies are usually covered by caching and other design decisions in the GPU world. Reply
  • geok1ng - Wednesday, June 25, 2008 - link

    The tests showed what i feared: my 8800GT is getting old to pump my Apple at 2560x1600 even without AA! But the tests also showed that the 512MB of DDR5 on the 4870 justifies the higher price tag over the 4850, something that the 3870/3850 pair failed to demonstrate. It remains the question: will 1GB of DDR5 detrone NVIDIA and rule the 30 inches realm of single GPU solutions? Reply
  • IKeelU - Wednesday, June 25, 2008 - link

    "It is as if AMD and NVIDIA just started pulling out hardware and throwing it at eachother"

    This makes me crack up...I just imagine two bruised and sweaty middle-aged CEO's flinging PCBs at each other, like children in a snowball fight.
    Reply
  • Thorsson - Wednesday, June 25, 2008 - link

    The heat is worrying. I'd like to see how aftermarket coolers work with a 4870. Reply
  • Final Destination II - Wednesday, June 25, 2008 - link

    http://www.techpowerup.com/reviews/Powercolor/HD_4...">http://www.techpowerup.com/reviews/Powercolor/HD_4...

    Look! Compare the Powercolor vs. the MSI.
    Somehow MSI seems to have done a better job with 4dB less.
    Reply

Log in

Don't have an account? Sign up now