40nm Supply Redux

If you have seen our Radeon 5800 supply article, then you know that AMD is currently trying to come to terms with a significant shortage of Cypress dice. Since the 5800 series launched in September, TSMC’s yields have taken a hit as the company ramps up 40nm production. And while this is resulting in more usable chips per week than when AMD started, it’s lower than it was supposed to be. Compounding matters is high demand for these cards thanks to their performance, features, and a lack of significant competition from NVIDIA at this time.

So when we were briefed about the 5970, we asked AMD point-blank whether it was a good idea to be launching another Cypress based card so soon, and at a time when they already don’t have enough chips to go around. Their answer was equally straightforward: why not?

The design is done and AMD is already capable of building the 5970. For AMD, there is no benefit in waiting; no matter what they do, anything with a Cypress chip in it today is going to sell out. Holding back may be slightly more egalitarian, but as the 5970 is a luxury part, it’s not a high-volume part anyhow, so its introduction isn’t going to significantly disrupt 5800 shipments even if it does use 2 GPUs per card. Ultimately I don’t think we would even be having this discussion unless the profit margin on the 5970 is higher than the 5870, so at some point this comes down to AMD doing what is most profitable for them.

At $600, AMD isn’t going to sell a ton of 5970s, and the launch numbers reflect this. While the 5800 series cards launched with tens-of-thousands of cards, the 5970 launch will simply be with thousands of cards. Even as a low-volume part, we’re expecting the 5970 to sell out just as fast as any 5800 card did. But depending on what AMD does with future chip shipments though and what TSMC’s yields do, this may be the first product line where demand finally gets met in the near future.

We also had a chance to talk to AMD about the overall 40nm supply situation. AMD of course isn’t very pleased with the situation, but this is something they’ve apparently planned for, after their first 40nm test chips came back as being less impressive than their 55nm and 65nm test chips were. Besides TSMC’s subpar yields, AMD is unable to get as many wafer starts as they’d like, which is compounding the issue.

Finally, we’re told that the TSMC situation is continuing to improve, and that AMD currently expects the Cypress chip supply to pick up in December. To what level of production “pick up” goes with we’re not sure, but it’s likely less than demand. In talking to AMD, they didn’t seem confident in being able to keep any Cypress-based products in stock through Christmas. Supplies will improve through the end of the year, but it sounds like it’s going to be 2010 before supply and demand finally balance out.

Meet The 5970 The Card They Beg You to Overclock
POST A COMMENT

114 Comments

View All Comments

  • Zool - Wednesday, November 18, 2009 - link

    So yes. The answer is that the gpu is doing less work with vsync than without it.(dam still no edit button) Reply
  • Yojimbo - Wednesday, November 18, 2009 - link

    The plural for a casting/shaping instrument "die" is "dies" not "dice." Reply
  • Lennie - Wednesday, November 18, 2009 - link

    I am going to post this again here. Thought it may not get noticed since I posted it first as a reply in previous pages. Hope I wont hurt anyones feelings :)

    I thought everyone knew about Furmark and ATi by now. It used to be like this on 4870 series too.

    It went like this, at first there were few reports of 4870(X2) cards dying when running Furmak. Further investigation showed that it was indeed Furmark causing VRM's to heat up to insane levels and eventually killing them. Word reached ATi from that point on ATi intentionally throttles their card when detecting Furmark to prevent the damage.

    Yeah in fact the amount of heat load Furmak puts on VRMs is unrealistic and no game is able to heat up the VRMs to the level Furmark does. OCCT used the same method (or maybe even integrated Furmark) to test for stability (in their own opinion ofc)

    So beware about Furmark and OCCT if you have HD4K or 5K.

    The term "Hardware Virus" is rightfully applicable to Furmark when it comes to HD4K (and 5K perhaps)
    Reply
  • Lennie - Wednesday, November 18, 2009 - link

    I want to add that VRM overheating is quite tricky, since normally people only check on GPU temps.

    When you run Furmark you would notice that GPU temps are in acceptable range while at the same time your VRM's are cooking without you knowing about it.

    So remember to always check your VRM temps when running graphics stability tests like Furmark or OCCT's graphics test specially when you're overclocking the card.

    I use Riva or Everest to check on VRM temps.
    Reply
  • Rajinder Gill - Thursday, November 19, 2009 - link

    The temp reading that is displayed by Everest is an averaged figure. The junction temp of the slaves is the critical issue. Even though the average temp may appear to be within bounds, there is the possibility that one of the slaves may be running abnormally. Volterra keep their specifications under NDA. What I do know is that the general configuration if one slave shuts down is that the remaining slaves take the load. The result is not usually pretty. I think ATI may have implemented throttling to prevent the kind of burnouts users experienced running OCCT GPU tests on the last gen.

    Personally, I think the 3 phase Volterra solution used on the 5970's is right on the hilt for current draw (circa 135 amps per GPU). I'd wait for non reference solutions with enhanced power delivery if you plan on overclocking this card long term or plan to stress it heavily when OC'd).

    Later
    Raja

    Reply
  • Rajinder Gill - Thursday, November 19, 2009 - link

    I should add that I'm assuming ATI used the biggest Volterra slaves rated at 45 amps each and not the 35/40 amp varieties. Reply
  • thebeastie - Wednesday, November 18, 2009 - link

    Any one that has a clue is buying a proper case like the Storm Sniper Black Edition and fitting this with heaps of space to spare.
    Also I recommend a case with positive or at least neutral are flow.
    The storm sniper has a dust filter on its 20cm side fan to push more air in and aid in the GPU fans air flow that will go out the back of the cards vent holes at the DVI ports.
    Reply
  • at80eighty - Wednesday, November 18, 2009 - link

    Plan on getting one of these in another 7-8 months - The way I see it, despite being bleeding edge - ATI has a deadlock winner in this card and will produce only limited quantities so I'm kind of 'worried' about the availability & price down the line Reply
  • Silverforce11 - Wednesday, November 18, 2009 - link

    Wait til the end of December. Apparently the yeilds of cyrpress are going to be improved a lot then, so the prices will either remain the same or be lower a bit.

    However, since nV has nothing real for a long time, i dont foresee a drop in prices on ATI parts. Given the estimates, NV will have fermi out in april 2010, but not in significant quantities for a while after that. Im gonna grab a 5970 around xmas. :)
    Reply
  • at80eighty - Wednesday, November 18, 2009 - link

    Dude I hate you already :p I just bought a 5750 - I dont have the necessary components that would not bottleneck the 5970, so I'm going to have to wait a while. plus the whole 3 new monitors thing for the Eyefinity Reply

Log in

Don't have an account? Sign up now