I never knew quite what nVidia went through in terms of TWIMTBP. But having read about what they actually do to get that badge on games, I have to say I am impressed. Especially if they lose millions on it in the short term.
it would be interesting if NVIDIA got a X86 license. I would assume that they could come up with a decent CPU pretty quick with all that know how they have. I am also looking forward to the day when VIA starts actually being able to compete in the low-mid range areas for CPUs.
That would make for a nice 4-way CPU arena (and 3-way with Intel's Larabee). Dream on.
I take exception to the 8-series being regarded as great, for the simple reason that the 8800GTX was the only worthwhile card to have. The 8800GTS was average, and the 8600-level cards were worthless. As the article states, I consider the 8800GT to be a different generation (8.5-series maybe?). An indication of just how average the 8-series was is that I hung on to my 7-series until the HD4870, and I didn't feel the upgrade prickles.
The 6-series was incredible after the abomination of the 5(FX)-series. So exciting, the generation performed from the 6800Ultra all the way down to the 6600GT. My 6600GT was the best graphics purchase I had ever made until the HD4870.
I think the RV770 was an incredibly risky move that paid off brilliantly. As a long time nVidia man, I have to just admire the balls of whoever made the strategy decision. It could so easily have been the biggest bonehead move of the decade, but it caught nVidia hopping and Whoop-Assing instead of producing.
Maybe I've become too informed about the graphics industry (I know. How is that possible?), but I've gone off nVidia as a company recently. Is it just me, or does the new nVidia sound like the bragging, school-yard ******s of the industry? I just wish Jen-Hsun would put his cans of Whop-Ass away and get his act together. Oh, and put away the sledgehammers away as well Jen-Hsun, the die sizes are doing you no favours.
As a Linux fan, and someone who wishes ill on Microsoft, I hate to say that people have been writing Microsoft off for quite some time and none of the predictions have come true. I see a disconnect between mobile device OSes and PC systems, and OSes and software written for one won't mean people switching the other as well.
I can't see Tegra making nVidia bags of money either, for simple reason that nVidia can't gouge the OEMs for large margins (Only Apple can do that :) ). These aren't large margin items, and the market, while potentially significant, isn't going to be that huge. Most people still don't want/need HD media capable smart devices. At most, nVidia can hope for a couple of bucks per chip sold. How many iPhone/iPod/Zunes/netbook chips will they have to sell to make up for the $200m dump they took over the dying G86s?
CUDA and GPGPU needs a breakout application to be successful.
Physx hasn't done it. With Direct-X physics it doesn't look like it is going to.
Folding isn't it. Although the cause may be worthy, I don't want to offend participating forum members, but it is little more than an e-peen competition for uber-enthusiasts.
Even multimedia encoding isn't it.
When it does come, I don't think it will come labelled CUDA either. It'll come out labelled Open CL, and nVidia isn't going to have a lock on that are they? It isn't just cross-platform it is cross-processor as well. The same competition nVidia has in the GPU space, they will have in the GPGPU space and more. nVidia has just gotten ahead at this point, which won't matter in the long run because there is nothing yet of substance that they can use to take advantage of their lead.
Don't get me started with nVidia chipsets, which peaked with the nForce 2 and have slowly gone downhill (don't get me started on Soundstorm, I'm still livid over what nVidia did).
nVidia, get a strategy and stick to it! Do you want to be a part of this market?
If the answer is yes, then commit to putting out more than relabelled chipsets that half-heartedly tick feature boxes. Try harder. For the love of Zombie Jesus, put some passion back into it!
If the answer is no, then do something to save that all important Chinese face, and then put a bullet into your chipset team. Maybe you could put them into working on GPUs, or Tegra, or whatever your latest marketing name is this month.
AMD may be broke, pay for their R&D with the change they find in payphones, and wear cricket boxes in their pants so that when Intel comes around and kicks them in the nuts it doesn't hurt so much, but they give the impression they have a plan for the future, the drive and passion to make it happen, and the flexibility to innovate. If only they had more money. Imagine what they could do with Intel's R&D?
Originally Posted by Adnoctum AMD may be broke, pay for their R&D with the change they find in payphones, and wear cricket boxes in their pants so that when Intel comes around and kicks them in the nuts it doesn't hurt so much, but they give the impression they have a plan for the future, the drive and passion to make it happen, and the flexibility to innovate.
great article, the stuff about the way its meant to be played is definately the most interesting part. perhaps another article on what companies do to get their badge up at the beginning of games? would like to see what intel have done to get Runs Best On....
I've gone off nVidia recently as well , especially hearing about the design flaws which led to the high failure rate of G86 problems.ATI have just put out cards which have proven to be better value for money for the most in the last 12 months , and nVidia still are relying on G92 to counter them (which has done a good job I'll admit).
They really are taking a massive risk with GT300 if it's as big as said in the article ,I'd have thought they learned their lesson but seemingly not...
I run a 4870X2 im my main rig and I do love it. But I still have warmer feelings for my GTX285 in my other rig. TWIMTBP programme is head and shoulders above anything ATI have and CUDA is superior in implementation compared to ATI Stream.
Always a purchaser of ATI, if only because their prices were cheaper.
That said, my 2nd card was a HD4850 and my first was a X1550... But Nvidia's risk taking is perhaps too excessive, it's not revolutionary and it's not going to help gaming performance, which is what most people who buy GPUs want.
wuyanxu ATI releasing a DX11 card (if) at Win7's release wont really do much damage to NVIDIA since there is 0 DX11 games being released and most sane people don't go out and get a brand new $500 dollar card just cause its new (note sane).
Originally Posted by Vergil_117 wuyanxu ATI releasing a DX11 card (if) at Win7's release wont really do much damage to NVIDIA since there is 0 DX11 games being released and most sane people don't go out and get a brand new $500 dollar card just cause its new (note sane).
Ah but having an early footing surely helps.
Look at the 8800GTX. It was the first DX10 card, and there were not many DX10 games at the time, in fact barely. But it won out. Why? Because it came out first, and because it did, the competition would have to work faster and rush their design...so what happened?
I've bounced between ATI and Nvidia for quite few years. My first "real" graphics card was a geforce 2 GTS, then a geforce 3, then a radeon 9500 pro (one of the best bang for buck cards of all time...), then 7800GT, then a 7950GT (yeah... sold that 7800GT pretty quick...), and now 8800GT. At this point if I decide to build another machine I'm probably going to go back to ATI simply because for the money they have the better card. Being that I'm not gonna be spending 295GTX money, everything in the ~$200 range is pretty much a toss up and ATI just happened to land slightly ahead (4870x2, 4890, etc). I do hope both companies stay competitive though because this is the only way we'll see any real innovation keep coming from them.
Originally Posted by Vergil_117 wuyanxu ATI releasing a DX11 card (if) at Win7's release wont really do much damage to NVIDIA since there is 0 DX11 games being released and most sane people don't go out and get a brand new $500 dollar card just cause its new (note sane).
as Elton's reply. with these new technologies, it's more of whoever wins the race. nVidia had been boasting their "first Dx10 card" advertising for a while, right up to 9800GTX.
to expand on my ray tracing part, the reason i think nVidia will win is because their shader pipelines are more general purpose rather than ATI's specific approach. so, it will be easier for nVidia to adopt to a newer kind render technique.
Just sold my trusty HD4870 for 100, and bought a HD2600XT for 30, just to wait for this new HD5870... some say that it'll come out on 10th of Sept... I sure hope so, can't play s*** with that card at 1920x1200...
Well this year I bought my first ATI card. I've become jaded when it comes to Nvidia's cards - too many restickered products.
I've had S3Virge ->MX440->FX5200->6800ge->HD4850. Better performance for the $ and that's the way most people think.
I do believe we need at least a two horse race though. Imagine what would have happened if AMD went under? We'd all be using single core pentiums still :P
After a few years of watching the market I'm going all AMD/ATI for my next build.
After the horrendous issues I have had with my 260s I refuse to use Nvidia now. I have a gtx295 but this will be hitting ebay as soon as I feel a need to let it go. (soon I hope)
Makes me sad to say this since I've always had good luck with Nvidia products, but easily the two predominant thoughts that come to mind anymore when I hear the word Nvidia is "can of whoop-ass" followed closely by "rebranding".
Also another plus I have always found is that Nvidia has had unified drivers for virtually forever. While my old ATI cards have bitten the dust and have crapped out drivers my Geforce 3s are still ticking right along.
Comments 1 to 26 of 48
currently, seeing ATI's Dx11 card will release on time to Win7, it may seem nVidia have lost this round.
but the new era of ray tracing is coming, and i think nVidia will win the first round of ray tracing just like 8800GTX.
it's cycles, it's currently the red team's advantage.
Personally I can't wait for a Tegra-equipped smartphone like the ZuneHD. That looks like a special bit of kit too.
That would make for a nice 4-way CPU arena (and 3-way with Intel's Larabee). Dream on.
The 6-series was incredible after the abomination of the 5(FX)-series. So exciting, the generation performed from the 6800Ultra all the way down to the 6600GT. My 6600GT was the best graphics purchase I had ever made until the HD4870.
I think the RV770 was an incredibly risky move that paid off brilliantly. As a long time nVidia man, I have to just admire the balls of whoever made the strategy decision. It could so easily have been the biggest bonehead move of the decade, but it caught nVidia hopping and Whoop-Assing instead of producing.
Maybe I've become too informed about the graphics industry (I know. How is that possible?), but I've gone off nVidia as a company recently. Is it just me, or does the new nVidia sound like the bragging, school-yard ******s of the industry? I just wish Jen-Hsun would put his cans of Whop-Ass away and get his act together. Oh, and put away the sledgehammers away as well Jen-Hsun, the die sizes are doing you no favours.
As a Linux fan, and someone who wishes ill on Microsoft, I hate to say that people have been writing Microsoft off for quite some time and none of the predictions have come true. I see a disconnect between mobile device OSes and PC systems, and OSes and software written for one won't mean people switching the other as well.
I can't see Tegra making nVidia bags of money either, for simple reason that nVidia can't gouge the OEMs for large margins (Only Apple can do that :) ). These aren't large margin items, and the market, while potentially significant, isn't going to be that huge. Most people still don't want/need HD media capable smart devices. At most, nVidia can hope for a couple of bucks per chip sold. How many iPhone/iPod/Zunes/netbook chips will they have to sell to make up for the $200m dump they took over the dying G86s?
CUDA and GPGPU needs a breakout application to be successful.
Physx hasn't done it. With Direct-X physics it doesn't look like it is going to.
Folding isn't it. Although the cause may be worthy, I don't want to offend participating forum members, but it is little more than an e-peen competition for uber-enthusiasts.
Even multimedia encoding isn't it.
When it does come, I don't think it will come labelled CUDA either. It'll come out labelled Open CL, and nVidia isn't going to have a lock on that are they? It isn't just cross-platform it is cross-processor as well. The same competition nVidia has in the GPU space, they will have in the GPGPU space and more. nVidia has just gotten ahead at this point, which won't matter in the long run because there is nothing yet of substance that they can use to take advantage of their lead.
Don't get me started with nVidia chipsets, which peaked with the nForce 2 and have slowly gone downhill (don't get me started on Soundstorm, I'm still livid over what nVidia did).
nVidia, get a strategy and stick to it! Do you want to be a part of this market?
If the answer is yes, then commit to putting out more than relabelled chipsets that half-heartedly tick feature boxes. Try harder. For the love of Zombie Jesus, put some passion back into it!
If the answer is no, then do something to save that all important Chinese face, and then put a bullet into your chipset team. Maybe you could put them into working on GPUs, or Tegra, or whatever your latest marketing name is this month.
AMD may be broke, pay for their R&D with the change they find in payphones, and wear cricket boxes in their pants so that when Intel comes around and kicks them in the nuts it doesn't hurt so much, but they give the impression they have a plan for the future, the drive and passion to make it happen, and the flexibility to innovate. If only they had more money. Imagine what they could do with Intel's R&D?
LOL, awesome description :)
$$$
They really are taking a massive risk with GT300 if it's as big as said in the article ,I'd have thought they learned their lesson but seemingly not...
That said, my 2nd card was a HD4850 and my first was a X1550... But Nvidia's risk taking is perhaps too excessive, it's not revolutionary and it's not going to help gaming performance, which is what most people who buy GPUs want.
Ah but having an early footing surely helps.
Look at the 8800GTX. It was the first DX10 card, and there were not many DX10 games at the time, in fact barely. But it won out. Why? Because it came out first, and because it did, the competition would have to work faster and rush their design...so what happened?
The 2900XT came to being.
to expand on my ray tracing part, the reason i think nVidia will win is because their shader pipelines are more general purpose rather than ATI's specific approach. so, it will be easier for nVidia to adopt to a newer kind render technique.
BTW, great article, really, really good :)
I've had S3Virge ->MX440->FX5200->6800ge->HD4850. Better performance for the $ and that's the way most people think.
I do believe we need at least a two horse race though. Imagine what would have happened if AMD went under? We'd all be using single core pentiums still :P
After a few years of watching the market I'm going all AMD/ATI for my next build.
Weak, Nvidia. Weak.
Also another plus I have always found is that Nvidia has had unified drivers for virtually forever. While my old ATI cards have bitten the dust and have crapped out drivers my Geforce 3s are still ticking right along.