news report 55 nm graphics chips are in short supply

Report: 55-nm graphics chips are in short supply

Shopping for a DirectX 10 graphics card? You might have to look harder than usual—and potentially pay more, as well. DigiTimes brings word that 55-nm graphics processors are in tight supply, yet both AMD and Nvidia are too concerned with their 40-nm, DirectX 11 products to do anything about it.

The site quotes Taiwan’s Commercial Times, which itself got the information from “sources in the retail channel,” as saying the shortage of 55-nm graphics chips may last all the way through the first quarter of next year.

Nvidia, for one, reportedly wants to avoid an inventory surplus of current-gen products, so DigiTimes says it “currently does not plan to increase its GT200 series GPU supply.” (The 55-nm GT200b chip powers GeForce GTX 260 and 275 cards on the desktop.) However, the company’s DirectX 11 graphics processor may not arrive until next year.

We reported in our Radeon HD 5870 review that AMD plans to introduce $100-200 DirectX 11 cards based on the new Juniper GPU before the end of the year. Throw in Hemlock, a dual-GPU monster also scheduled for this quarter, and AMD should have a reasonably meaty DirectX 11 lineup in time for Christmas. We’ll have to see whether Juniper-based cards are widely available soon after launch, though.

0 responses to “Report: 55-nm graphics chips are in short supply

  1. lol indeed, albeit, usually not from those trees though, we tend to use white spruce for paper up here in Canada, but I do get the paper launch joke if thats what you intended 🙂

  2. Well, Nvidia is trying to grow the market for GPUs, not just trade one market (gaming) for another (compute), and not only do I respect that strategy but it also means that they are not going to turn their back on gamers. However, the real “wait and see” issue is what sort of price points they will be able to hit with the GT 300 family.

    I’m pretty sure they won’t be able to get much more mileage out of renaming the G92 any more times, so they will actually have to introduce some new chips for the mainstream and value markets with the GT300 architecture, but it is too soon to tell if they will be doing that sooner or later.

    If Nvidia can’t launch mainstream stuff along with their halo GT300 card, or very soon thereafter, we could end up seeing an unprecedented market share shift to ATI.

  3. GTX285 at anything more than 200$ is a no go for me, I looked at them but compared to the 5850 they were not appealing.
    1) I have a 8800GT and PhysX has not shown me anything spectacular.
    2) GTX285 with mature, optimized drivers is decent compared to a 5850 with immature drivers. (wait and see)
    3) DX 11 is will run better look better on a DX11 card vs a DX10 (duh) and that means future proof for some time.
    4) 5850 early batches has a higher chance of getting 5870 underclocked for channel supply reasons only (hence overclockable)
    5) Nvidia is obvious trying to save themselves (long term) and is prioritizing Compute over Gfx which means if they are showing they don’t belive in gfx as much as compute then why should I believe in them. (Yes you can argue this but I will take a wait and see approach and Im sure u have read initial Fermi architecture details.)

  4. Well,

    Direct Compute 11 is a programming language and also a series of tools which can tap into a GPU as a General Purpose Compute Device. What you would need is for a Physics Library (like Havok or other open source variants) to be written in Direct Compute 11.

    OpenCL is much the same.

    So it’s not really apples and oranges per-say. PhysX are a Novolagic based set of libraries written in CUDA (C, C++ or Fortran with CUDA extensions) in order to take advantage of a CUDA architecture based GPU (nVIDIA 8 series and higher).

  5. Again, not to worry about the games, they will come. As it is, I don’t own any games that use PhysX, and it has been out for a while now. The games are out there, just not many of them. I’ll take a yet-to-ripen orange over a rotten apple, since you now seem comfortable comparing those two.

    At present, physics in games is not very meaningful. There are a precious few examples where I think it makes any difference at all. Mirror’s Edge is one, but it is a game I am only peripherally interested in. More leaves blowing around in Sacred 2? Whatever, and that’s besides the point that I’m not even interested in the game. How does physics make Batman any better? Just better explosions? Sorry, not good enough to meaningfully change the gaming experience, not to the point that I’d choose it over a DX 11 card or pay substantially more for a video card that supports it.

    The concept is great though. But it needs to be built on an open or at least a neutral standard.

    But you’re not interested in that. You’re interested in getting people to buy Nvidia’s cards, for whatever reason.

    At the end of the day, a 4870 for $126 is a killer deal; a GTX 285 for $250 – fail, wait for 5850; at GTX 260 for $150 – yeah, that’s a killer deal, so why don’t you try to sell that card for Nvidia instead.

  6. Well now that someone has finally showed you google, use it to look up how many PhysX games you can buy vs how many DirectX 11 games you can buy.

  7. Yes, the phrase is common; and it is commonly used by people when they are unable or unwilling to come to terms with a given situation and so they just want the situation to go away.

    Also, I am not so sure you understand the difference between PhysX and DX11. Are you aware of the fact that they are unrelated? Have you done any study of the two? If you have not and you are unable to find some good material to read I can probably google some reading material for you, but a google search is not that hard to do and you might be able to understand how to do it. Go to §[<<]§ and enter a search term. If you still have trouble, let me know, I'd be glad to assist.

  8. So I was right you are unaware of this very common phrase and what it means.

    You also don’t know the difference between PhysX and DX11.

    Also I did notice AMD putting out a lot of paper lately, still no games though. Such a shame.

  9. Apples and Oranges can be compared on many levels; I thought you might be aware of this. They are both fruit, they both grow on trees, they are both round, they are of comparative size, both are high in fiber, both contain naturally occurring sugars, and so on. So, if nothing else, you can be excited to learn that you can, in fact, compare apples and oranges.

    As for DX11 and physics – you almost seem to imply that the two are mutually exclusive, though they are not. Direct 3D can actually allow for physics calculations to be done with compute shaders. AMD is working fleshing out this option, as well as for physics through OpenCL. You can find more here:


  10. I’m guessing from your response you have never heard the term “Apples and Oranges”. Look it up sometime, it’s a rather common phrase.

    As for DX11 vs PhysX. You do know they don’t compete? Of course you don’t.

    Please do some research next time.

  11. Not to worry about the games, they will come. As it is, I don’t own any games that use PhysX, and it has been out for a while now. The games are out there, just not many of them. I’ll take a yet-to-ripen orange over a rotten apple.

  12. DirectX 11 has no physics libraries. No games either. Apples and Oranges.

    Very suspect that you’d demand Dx11 but dismiss OpenGL.

  13. No PhysX is win, IMO. I don’t want to support some a proprietary standard if I don’t have to. Very suspect that you’d demand PhysX but dismiss DX11.

  14. The is a good business decision of shifted their inventory policies, by both nVidia and ATI.

    §[<<]§ Looks like the OEMs can no longer play the "attempt to make large purchases and demanded huge discounts once a quarter, virtually forcing the chip designer to sell its GPUs at lower prices since otherwise its quarterly results would be affected and in some cases it would be forced to write off older inventory in order to improve sales of newer products" game. Nvidia, explained that the company’s goal is to have inventory for 30 days of sales, which allows the firm to respond to competitive challenges in a timely manner. ATI’s goal seems to be generally similar: to keep as few chips in stock as possible.

  15. You’ll be…BLINDED BY THE LIGHT, WRAPPED UP LIKE A <blank> when you see the power of our new GPU!

  16. So you are going to compare those couple of games that are optimized for Nvidia to 5850 majority? 5850 without a doubt is faster than GTX285 in all resolutions and AA settings and can be just as fast 5870 when it’s overclocked.

    §[<<]§ 5850 uses the same memory as 5870. 1200mhz is what they are rated for. Majority of these 5850 overclocks great. With a right voltage it can leap to 1000mhz in this review. §[<<]§ For $10 difference it's a no brainer. Not to mention the link for $249 GTX was refurbished while 5850 is retail price. You just get more with 5850 than GTX285. Eyefinity, SSAA, angle dependent AF, DX11, much better power consumption.

  17. I saw that it was an outlet item, but does that absolutely mean refurbished? Dunno, maybe it does. If it does, that place needs to be more clear about stating it.

  18. only on one of the two cards they had. the other 5850 overclocked much worse. also at stock clocks the 5850 isnt much better in some cases than a gtx285. in fact the gtx285 actually beats it in a couple of games. that being said the 5850 is certainly the much better card overall with lower power consumption and better features including DX11

  19. The spectacle of Nvidia fanbois on the run…….keep praying for Fermi and hope Hemlock doesn’t squash it like an over ripe grape 🙂

  20. Why would you want GTX285 for $249? When you can get dx11 5850 with more features for $259 that is much faster and has more potential to be faster once it’s overclocked. Firingsquad got it to 5870 speeds without a problem.

  21. This just puts AMD in a good position. Nvidia while don’t even have a competing 40nm part yet. People should sell some Nvidia stocks now.

  22. Switching to some form of carbon from silicon is far more likely to happen first. Transistors, and I’m talking broadly here not just semiconductor ICs, were made out of germanium long ago, then silicon, fill in the pattern and you get carbon.

  23. I am a phd student in physics, concentrating on spintronics. It’s a pipe dream, I can assure you. The only reason you’ve heard of it is because in order to survive in science you have to misrepresent what you do as being more grandiose than what it really is.

    None of the things you have rattled off will be viable technologies within the next 10 – 20 years.

  24. If spending $250 is so inconsequential of a thing to you that you cannot wait a week or two for the better deal then by all means, order the lesser card now, or, better yet, step up to a 5870. To me, nothing is worse than ordering the inferior product and being stuck with it for the next couple of years because I couldn’t wait a couple of weeks. Yes, I have done this before.

  25. Not that I’d recommend spending $250 on a last gen card right now.

    But where can you get a 5850 for $10 more, or at all for that matter, the only listings I see are ‘out of stock’ or ‘4 week pre order’.

  26. Great “deal” if you want to pay near the same price as a 5850 for a card with slower performance, higher power consumption, and is a full Direct X generation behind. Yeah, what a deal.

  27. Also: quantum computers, spintronics, atomtronics, and computation using photons. There are a lot of promising replacements for CMOS electronics.

  28. I still like the idea of organic computers, but I doubt they’ll come to fruition any time soon if at all.

  29. Silicon-based metal-oxide-semiconductor transistors do have a fundamental limit to how small they can be made. Current theoretical estimates place this limit at around 1nm (gate oxide thickness), though we may well reach an economic limit before then (as it becomes more and more expensive to combat leakage, interconnect speed limits, tracing, and other higher-level limits).

    We will eventually need to abandon silicon-based semiconductor design for something more capable of working on a true “nano-scale” (likely candidates include quantum transistors and molecular electronics).

  30. uh? People have still been developing new technologies that have continually gone down. As we get into smaller nanometers, we will just move to picometers afterwards.

    This “it has to stop” thing is just all shenanigans. It also has no bearing on the performance increases that we can achieve either. Even if we stayed at 55nm for 10 years, you bet processor manufacturers would still find a way to up performance. I don’t know, maybe they’d add more cores? duh.

    Look at a graphics card. They have hundreds of small cheap cores, which sum up tons more floating output at the cost of less versatility. The same concept is experimentally being applied to processors as well.

  31. First one was a term I just coined, maybe quotes would have helped?

    Second one was a typo. I tried to type ‘smaller’ and ‘lower’ at the same time I think. =P

  32. Well then maybe you should learn to read because he never said they are making processors slower or taking away the functions of current chips.

  33. What are you talking about?

    The AMD HD 4890, 4870, 4850, 4730, 46×0, 45×0, and 43×0 cards, as well as pretty much the entire non-mobile line of current Nvidia cards, are *[

  34. “nano-race?”

    “brick wall where we can’t go any slower”

    Do you understand what you’re talking about? In general, making something smaller and retaining its functions is a *[

  35. I heard 90nm was in short supply as well.

    Seriously though, I don’t think it’s too much of a shock. AMD’s 3 series used 55nm and that was, what, at least 2 years ago? It’s time to move forward and this looks like a smart move for both companies.

    On a semi-related note: I know a decrease in transistor size is great for margins, but I hope that this doesn’t turn into a nano-race. Eventually we’re going to reach a brick wall where we can’t go any smaller. What will we concentrate on then I wonder? I heard of semi-conductors made of graphene will allow us to reach 1000GHz so maybe we’ll have another “hertz” race. =P

Cyril Kowaliski

Latest News

pancakeswap (CAKE)
Blog, Cryptocurrency, Investments, News, Price Prediction

PancakeSwap Price Is Up 33% in 30 Days as February Begins – Here’s Why You Need to Buy Now!

eco crypto
Blog, Cryptocurrency, Investments

The 9 Best New Cryptocurrencies YOU NEED TO BUY in 2023

With hundreds of cryptocurrencies created daily, it can be tough to differentiate the projects that won’t last longer than a few months from those  that will explode in value. Of...

3.1 m memag
blockchain, Blog, Cryptocurrency, Gaming, Innovation, News

Meta Masters Guild (MEMAG) Hits 3.4 Million – Buy Now as Phase 6 Begins!

News about the first mobile-based Web3 gaming guild has spread quickly among the crypto community and the latest numbers back up the hype. Even though the presale started just a...

crypto reddit
Blog, Cryptocurrency, Investments, News, Price Prediction

7 Best Cryptocurrencies On Reddit To Invest In 2023

Blog, Cryptocurrency, Investments, Sustainability

The Clock is Ticking to Buy this Eco-Crypto – 7 Reasons C+Charge Will Explode Post-Presale!

savannah marshall fght
Blog, Cryptocurrency, News

Boxing World Champ, Savannah Marshall Partners with FGHT as presale approaches to $4 million!

Twitter Files

Elon Musk’s Twitter Pledges Free Write-Only API Access To Bots Creating “Good” Content