Nvidia says Goodbye to XFX

[FalseChristian]

I like both GPU companies. The only reason I went with nVidia this time is because I like the fact you can tweek any game to your liking and even add profiles and the fact that older OpenGL games work thanks to 'Extension limit'. I'm not sure about ATI, but every ATI card I've had wouldn't run older OpenGL games like Quake/Quake2/Quake3:Arena. I still enjoy them.

 

[SirPauly]

I also think the stories of Nvidia's upcoming demise are exaggerated. But, there are some differences between AMD in the 2xxx and 38xx days vs. Nvidia's 8xxx/9xxx cards.

Back when the 2xxx and 3xxxx were sucking it up, Nvidia practically took a generation off. They felt moving from the fantastic 8xxx cards to the 9xxx cards needed to be nothing more than a sticker change. They gave AMD room. Rumor has it that AMD has already handily outsold Nvidia this generation, and looks like they might have their next gen parts coming out before Nvidia could even get their entire Fermi line up out. AMD isn't rebadging the 5xxx cards and calling it a day. They're moving forward while it seems Nvidia is still treading water.

Nvidia released a chip that's about as big as it can go and about as power hungry as it can go on the current technology. AMD's smaller chip seems to have given them room to move forward on the same process. I'm not saying Nvidia can't do the same, but it seems like they don't have as much flexibility. Combine that with Tegra's lack of sales and Nvidia's chipset business dwindling, there are some reasons for concern.

Again, I don't think they're going anywhere, but they have some real challenges that are different than AMD's situation in the 29xx/38xx days. I think Nvidia knows this, that's why we see them pushing CUDA and the HPC market so hard. While gaming cards may not make them profit like professional parts, it still makes them a lot of revenue, money they use for R&D so they can build these fantastic GPU's. While I don't think they're going out of business, I do think in five years the GPU market is going to look a lot different than it does now.

There are always reasons for concerns. The technology field is as volatile as there is in the world. Look how fast things change with the 58XX series from nVidia compared to the 9XXX series from ATI to the 29XX series from ATI to the 8XXX series from nVidia to what exists today. They're both so talented and they shift back-and-forth from time-to-time.

What some mind-sets seem to enjoy is domination vs Epic fail --- blanket views based on small samplings of data. I see things as positive and steps -- even with non-dominant parts there are a lot of positive steps that need to be made for future products and road-maps.

 

[Meghan54]

Billions of dollars in cash can give you all sorts of flexibility, particularly for a company as relatively small as nV.

In the fantasy world of anti nV rhetoric, it appears that it is indeed a dire situation. In the real world, nVidia has about enough cash to buy 1/2 of AMD outright at the moment(they also have a ~$2Billion dollar edge in market capitalization). People on these forums like to look at benchmark charts and think they auto translate into business stability- the 3dfx loyalists saw their triumph over nV looming until the week they went bankrupt.

If you want to compare how the players are doing as companies, I suggest spending a lot more time combing over financials and a whole lot less time looking at benchmark charts.

I did look at their financials.....AMD's cash & marketable securities assets total $1.896B, NV's same totals $1.772B, not exactly NV outdoing AMD there, eh?

Total cash assets, which includes inventory, accounts receivable, etc., has AMD at $3.313B, NV at $2.656B....again, failing to see how NV could by out half of AMD with that shortfall in assets compared to AMD.


Total assets for both companies: AMD $4.955B, NV $3.731. Still missing your point about looking at the financials to prove your point.....because looking at them doesn't support much of what you've said.

AMD's net profit margin continues to be above 24% while NV's is languishing around 6.36%. Not too good there.


All above taken from latest filed quarterly reports to the SEC.

 

[SlowSpyder]

In the fantasy world of anti nV rhetoric, it appears that it is indeed a dire situation. In the real world, nVidia has about enough cash to buy 1/2 of AMD outright at the moment(they also have a ~$2Billion dollar edge in market capitalization). People on these forums like to look at benchmark charts and think they auto translate into business stability- the 3dfx loyalists saw their triumph over nV looming until the week they went bankrupt.

If you want to compare how the players are doing as companies, I suggest spending a lot more time combing over financials and a whole lot less time looking at benchmark charts.

I don't think anyone is saying that the 5970 is faster than the GTX480, so that means Nvidia is in trouble. At least not in this thread. I think people are a little alarmed because some things are changing, Nvidia is making their own cards and they lost XFX. All that money in the bank didn't get Fermi out on time and in some ways it sure appeared fairly rushed. All that money and we're still waiting for the 512SP part that was announced in January. At least in the consumer market, Nvidia appears to have lost their dominant position.

What I meant by 'flexibility' is that it seems AMD can more easily make a tweaked and improved part on 40nm. Nvidia may do so as well, but we haven't heard a peep from them (which of course doesn't mean it's not coming, but it at least appears to me that the Radeon 6xxx is going to be positioned against Nvidia's current offering for a while).

As I said, I don't think Nvidia is going anywhere, but I think the landscape is changing.

 

[SlowSpyder]

There are always reasons for concerns. The technology field is as volatile as there is in the world. Look how fast things change with the 58XX series from nVidia compared to the 9XXX series from ATI to the 29XX series from ATI to the 8XXX series from nVidia to what exists today. They're both so talented and they shift back-and-forth from time-to-time.

What some mind-sets seem to enjoy is domination vs Epic fail --- blanket views based on small samplings of data. I see things as positive and steps -- even with non-dominant parts there are a lot of positive steps that need to be made for future products and road-maps.

Of course there's always risks. Even a dominat player like Intel fell behind AMD in many ways for a while. Intel may have a problem on it's hands in the future with ARM. But I think this is a different situation for Nvidia. They are going to get squeezed by both AMD and Intel with APU's, their chipset business is pretty minor these days, Tegra has yet to really make much impact, and in the discreet graphics arena it seems like they are playing second fiddle without any chance or recoving soon (the Radeon 6xxx parts are right around the corner).

Like I've said, I really don't think Nvidia is going away. But I do think that they are going to be a different company in the future... I can see them focusing more on mobile parts and HPC parts. Graphics cards were their bread and butter for so long, they seem to be shifting away from that a bit. The numbers we have seen show that they make a lot more money on professional parts, and with the number of mobile parts that are sold there is a lot of potential there as well. There might come a day where Nvidia decides it doesn't really need graphic cards. Or, maybe their next part will be like a 9700 Pro and 8800GT level of 'wow' and make them tons of money. We'll have to wait and see, but if I was placing a bet I'd say that their focum shifts towards mobile and professional parts.

 

[SirPauly]

Imho,

They have to offer reasons why the GPU is relevant and will continue to work hard in this arena. This is why they invented Cuda, holding GPU technology conferences, trying to enter markets and areas that were strictly CPU territory. In other words if they don't -- the GPU will be just another device integrated like sound and dominated by the two CPU companies.

For all this yelling about proprietary, nVidia has no choice -- if they don't -- they simply will not have the abilities to showcase what the GPU may do for technology. Intel and AMD can dance and try to place pressure -- nVidia can't and must innovate or die.

 

[bryanW1995]

I did look at their financials.....AMD's cash & marketable securities assets total $1.896B, NV's same totals $1.772B, not exactly NV outdoing AMD there, eh?

Total cash assets, which includes inventory, accounts receivable, etc., has AMD at $3.313B, NV at $2.656B....again, failing to see how NV could by out half of AMD with that shortfall in assets compared to AMD.


Total assets for both companies: AMD $4.955B, NV $3.731. Still missing your point about looking at the financials to prove your point.....because looking at them doesn't support much of what you've said.

AMD's net profit margin continues to be above 24% while NV's is languishing around 6.36%. Not too good there.


All above taken from latest filed quarterly reports to the SEC.

yes, but most of amd's money has to be spent treading water/getting smoked by intel just to stay in the game. nvidia has a lot more flexibility. if you just looked at the $$ available for the graphics divisions nvidia has an enormous advantage still b/c their other main source of revenue is also graphics. this might change over the next 2-3-4 years as the apu becomes the wave of the future, but for now NV still clearly has a dominant rather than weak position. Remember, we have seen ati ascendant in the past but they typically don't get much over 50% discreet while nv has been as high as 70% and is still holding at around 50% after 2 years of intense competition and one year of basically nothing.

 

[n0x1ous]

yes, but most of amd's money has to be spent treading water/getting smoked by intel just to stay in the game. nvidia has a lot more flexibility. if you just looked at the $$ available for the graphics divisions nvidia has an enormous advantage still b/c their other main source of revenue is also graphics. this might change over the next 2-3-4 years as the apu becomes the wave of the future, but for now NV still clearly has a dominant rather than weak position. Remember, we have seen ati ascendant in the past but they typically don't get much over 50% discreet while nv has been as high as 70% and is still holding at around 50% after 2 years of intense competition and one year of basically nothing.

exactly

 

[RussianSensation]

If you want to compare how the players are doing as companies, I suggest spending a lot more time combing over financials and a whole lot less time looking at benchmark charts.

:thumbsup: Not even the info of NV's product cash flow segmentation can stop the "NV's days are numbered" comments.

At least in the consumer market, Nvidia appears to have lost their dominant position.

Consumer market is split into Desktop and Notebook. Within each of those segments, you have sub-segments such as <$100, $100-200, $200-300, >$300. Between the sub-segments, <$100 and $100-200 desktop segments account for probably 70&#37;+ of all desktop videocards sold between the 2 companies. So you are saying that NV has lost dominant market positioning in the <$200 desktop segment then? Your statement is too general imo. We'd need a marketshare segmentation to conclude that NV's dominant market position has been lost in the most important <$200 segments, unless you mean "overall desktop+notebook share" has fallen, which is true. Also, market share between segments doesn't tell us cash flow per segment either. There is too much missing info.

I still think Fermi was largely criticised as a result of launching late. On top of that, it arrived with not-so-great launch drivers and negative initial reviews (probably due to being late, high launch prices, and higher power consumption, etc.) As videocards, GTX460/470/480 are certainly better than what AMD had with HD2900 or all of 3800 series. If Fermi launched Sept-Dec of 2009, I doubt we'd see this much criticism. The biggest problem was poor execution and showing up to the fight 6+ months late. Now that the dust has settled, as a technological architecture, Fermi is actually pretty good (AMD 5000 series can't compete in DX11 games). Looking at current HD5000 series pricing, the HD5770 is the only noteable standout. Yet the negative perception of Fermi will probably never stop, regardless of pricing.

 

[Madcatatlas]

:thumbsup: Not even the info of NV's product cash flow segmentation can stop the "NV's days are numbered" comments.



Consumer market is split into Desktop and Notebook. Within each of those segments, you have sub-segments such as <$100, $100-200, $200-300, >$300. Between the sub-segments, <$100 and $100-200 desktop segments account for probably 70%+ of all desktop videocards sold between the 2 companies. So you are saying that NV has lost dominant market positioning in the <$200 desktop segment then? Your statement is too general imo. We'd need a marketshare segmentation to conclude that NV's dominant market position has been lost in the most important <$200 segments, unless you mean "overall desktop+notebook share" has fallen, which is true. Also, market share between segments doesn't tell us cash flow per segment either. There is too much missing info.

I still think Fermi was largely plauged with poor launch drivers, negative initial reviews due to high launch prices and lack of proper fan profiles, etc. As a videocards, GTX460/470/480 are certainly better than what AMD had with HD2900 or all of 3800 series. If Fermi launched Sept-Dec of 2009, I doubt it'd see this much criticism. The biggest problem was poor execution. The cards are decent.

Fermi as in GTX480 512 cuda cores= is not possible
Fermi as in GTX480 is BAD in every metric except performance. If you want to live with it, go ahead, noone is saying "you cant buy that".

Also, Meghans post raises some questions. Care to answer them RS or Benskywalker?
You seem oh so happy to reurgitate Nvidias economy at every opportunity, can you shed some light on the questions raised by Meghans post?

 

[golem]

Also, Meghans post raises some questions. Care to answer them RS or Benskywalker?
You seem oh so happy to reurgitate Nvidias economy at every opportunity, can you shed some light on the questions raised by Meghans post?

Meghan's post is somewhat misleading, it only listed the asset side. If you also look at the liabilities, AMD has 2.58B in debt while Nvidia has 23.73M in debt. AMD has more assets than Nvida but also MUCH MUCH more debt.

Take a look at the total debt to equity ratios. Nvidias is at .86 while AMD is at 343.09.

 

[klansek]

In the fantasy world of anti nV rhetoric, it appears that it is indeed a dire situation. In the real world, nVidia has about enough cash to buy 1/2 of AMD outright at the moment(they also have a ~$2Billion dollar edge in market capitalization).

wait, last time I checked M. CAP for AMD was 4,6B and for NVDA 4.14B, so I do not know form where did you pull your +2B edge.

 

[golem]

wait, last time I checked M. CAP for AMD was 4,6B and for NVDA 4.14B, so I do not know form where did you pull your +2B edge.

Where are you looking? According to Yahoo financials Nvidia has a market cap of 6.16 Billion to AMD's 4.62B.

 

[WelshBloke]

wait, last time I checked M. CAP for AMD was 4,6B and for NVDA 4.14B, so I do not know form where did you pull your +2B edge.

Nvidias market cap is a little over 6B

AMD is 4.6B according to Yahoo.


DOH! too slow

EDIT AGAIN: what are the analysts bit to the bottom right on yahoo Nvidias look horrendous

EDIT YET AGAIN. no they dont. Yahoo uses negative debt instead of profit . They actually look quite good.

 

[RussianSensation]

I did look at their financials.....AMD's cash & marketable securities assets total $1.896B, NV's same totals $1.772B, not exactly NV outdoing AMD there, eh?

Total cash assets, which includes inventory, accounts receivable, etc., has AMD at $3.313B, NV at $2.656B....again, failing to see how NV could by out half of AMD with that shortfall in assets compared to AMD.

Still missing your point about looking at the financials to prove your point.....because looking at them doesn't support much of what you've said.

What are you talking about? That's not how one goes about evaluating a company's financial strength. You don't compare one line to the other....or one total to the other total :wub: Finance 101.

Here is the proper financial way of comparing 2 companies (very high level) - I took the latest B/S quarterly data available:

Liquidity Ratios provide information about the firm's ability to meet its short-term obligations. They are of particular interest to those extending short-term credit to the firm, or to gauge the company's ability to repay its short-term liabilities should repayment be demanded immediately.

Current Ratio
AMD = 2.03
NV = 3.24

Working Capital (Current Assets - Current Liabilities)
AMD = 1.7B
NV = 1.8B

Quick Ratio (Current Assets - Inventory / Current Liabilities)
AMD = 1.68
NV = 2.71

Cash Ratio (Cash + Marketable Securities / Current Liabilities)
AMD = 1.16
NV = 2.16

NV has better short-term liquidity.

Financial Leverage Ratios provide an indication of long-term solvency of the firm.

Debt Ratio (Total Debt / Total Assets)
AMD = 0.85
NV = 0.26

Debt-to-Equity Ratio (Total Debt / Total Equity)
AMD = 3.43
NV = 0.009

Did you forget that NV has 2.4B of Long-Term Debt? :whiste:

AMD's net profit margin continues to be above 24&#37; while NV's is languishing around 6.36%. Not too good there.

Profitability Ratios measure the success of a firm in generating profits.

NV took a massive write-down of $480 Million dollars related to GPU/chipset issues of 2008 and a lawsuit settlement related to the same issue.

Gross Profit Margin = (Sales - COGS / Sales)
Look at last 3 quarters of GPM. 45.5% ending May 2010, 44.7% ending January 2010, 43.4% ending October 2009....

Seriously people, before one can discuss financial standing of a company, it's important to learn the basics of finance/accounting and take all relevant information into account.

Also the comment about NV buying AMD with a "short-fall of its assets" was taken out of context. I am sure what Ben really meant was that NV could borrow $$$ against its total assets. The long-term loan + Cash/Short-term investments on-hand would allow NV to make a reasonable offer for AMD if they wanted to (but we have seen what this strategy did to AMD's buyout for ATI - still ongoing debt payments).

NV is a better managed company at the moment. I am sure Ben could easily more.

 

[bryanW1995]

Fermi as in GTX480 512 cuda cores= is not possible
Fermi as in GTX480 is BAD in every metric except performance. If you want to live with it, go ahead, noone is saying "you cant buy that".

Also, Meghans post raises some questions. Care to answer them RS or Benskywalker?
You seem oh so happy to reurgitate Nvidias economy at every opportunity, can you shed some light on the questions raised by Meghans post?

I already responded to them, but I'll do so again just for you.

amd's core business is cpu, with 10-20&#37; max in graphics. they bought ati to enable them to do more with their cpu (mission accomplished), but it's unlikely that they'll ever be primarily a gpu/graphics card company. nvidia, otoh, has no chipset business, no cpu business, and in fact is basically a gpu/graphics company that writes kickass software as well. they are pretty close to 100% gpu related, their tesla lineup, cuda, etc are all geared towards this. so if nvidia has ~ $2 billion in the bank to fight daamit and daamit has ~ $2 billion to fight nvidia AND intel, nvidia has a dominant position.

none of this changes that nvidia dropped the ball this gen, and it appears that amd is not letting their foot off the gas for next gen either. just keep in mind that it took nvidia a LONG time to achieve their favorable market position, and nvidia will need to continue to struggle at 28nm as well probably before we'd really have a sea change in market perception re consumer graphics. Also, as mentioned, nvidia is still doing many things on the software side to keep them in the game, and gtx 460 is a clear winner. they will swallow their pride for now as intel did in the a64/p4 days, and use their enormous cash advantage to regroup and refocus for the next round. this sort of thing happens in business/sports/life all the time, one team gains an advantage and it forces the competitor(s) to re-evaluate their (failing) strategy. It's not a guarantee that NV won't continue with the fermi model of larger/hotter/slightly faster, but I do think that they'll end up getting a 2nd vendor for 28nm (if not GF then someone else) and they'll probably be quicker to market going forward with a more consumer-friendly gpu like gtx 460.


edit: keep in mind, this is a game that amd is unlikely to win in the long term. if nv gets too cheap then intel buys them. if amd gets too dominant in the consumer gpu field, intel can cooperate more aggressively with NV as a counter to amd's rise. For years it has been intel >>>nvidia>>>amd. now it might be more like nvidia>amd but intel is still killing both of them. jhh bit off more than he could chew with his "whoop ass" comment, but as mentioned earlier re xfx & nvidia hurt feelings can quickly get forgotten when there is profit to be made or a competitor to be thwarted. and no, i don't think that i've ever seen thwarted used in a sentence bofore.

 

[T2k]

Weren't BFG and XFX two of the biggest manufacturers for NV, along with EVGA?

This is looking worse and worse for NV. I really hope they don't go under. I suppose their IP will not be lost regardless, but we need competition in the discreet GPU market. It probably wouldn't take intel too long to release *something*, but it would take them years to come up with something competitive at the high end.

Wasn't BFG just a label on some custom OEM-mfr'd card?

 

[RobertR1]

It would take another 2 poor product cycles for nvidia to feel the strain. The thing is, you can be sitting cash pretty at one point and start collapsing quickly. The cookie can crumble fast at times, esp when your market is limited and facing stiff competition.

 

[T2k]

I already responded to them, but I'll do so again just for you.

amd's core business is cpu, with 10-20&#37; max in graphics. they bought ati to enable them to do more with their cpu (mission accomplished), but it's unlikely that they'll ever be primarily a gpu/graphics card company. nvidia, otoh, has no chipset business, no cpu business, and in fact is basically a gpu/graphics company that writes kickass software as well. they are pretty close to 100% gpu related, their tesla lineup, cuda, etc are all geared towards this. so if nvidia has ~ $2 billion in the bank to fight daamit and daamit has ~ $2 billion to fight nvidia AND intel, nvidia has a dominant position.

none of this changes that nvidia dropped the ball this gen, and it appears that amd is not letting their foot off the gas for next gen either. just keep in mind that it took nvidia a LONG time to achieve their favorable market position, and nvidia will need to continue to struggle at 28nm as well probably before we'd really have a sea change in market perception re consumer graphics. Also, as mentioned, nvidia is still doing many things on the software side to keep them in the game, and gtx 460 is a clear winner. they will swallow their pride for now as intel did in the a64/p4 days, and use their enormous cash advantage to regroup and refocus for the next round. this sort of thing happens in business/sports/life all the time, one team gains an advantage and it forces the competitor(s) to re-evaluate their (failing) strategy. It's not a guarantee that NV won't continue with the fermi model of larger/hotter/slightly faster, but I do think that they'll end up getting a 2nd vendor for 28nm (if not GF then someone else) and they'll probably be quicker to market going forward with a more consumer-friendly gpu like gtx 460.


edit: keep in mind, this is a game that amd is unlikely to win in the long term. if nv gets too cheap then intel buys them. if amd gets too dominant in the consumer gpu field, intel can cooperate more aggressively with NV as a counter to amd's rise. For years it has been intel >>>nvidia>>>amd. now it might be more like nvidia>amd but intel is still killing both of them. jhh bit off more than he could chew with his "whoop ass" comment, but as mentioned earlier re xfx & nvidia hurt feelings can quickly get forgotten when there is profit to be made or a competitor to be thwarted. and no, i don't think that i've ever seen thwarted used in a sentence bofore.

This whole post is sooo wrong...


...HINT: soon there will be no "primarily a gpu/graphics card company" (sic) anymore as both AMD and INtel slowly taking over everything except performance/high-end desktop and professional graphics card market with their upcoming combined CPU-GPU chips.
SHort on x86 license, short on chipset license and lacking a viable mobile product lineup NV is completely doomed unless they can move to a CUDA/GPGPU-based business model (let's ignore their Great Leader's silly lies about being a software company, he's famous about regularly deceiving or even lying to investors.)

 

[nemesismk2]

I think the stories of Nvidia's upcoming demise are exaggerated and wishful thinking for some. No my comment is not because I have just purchased an Nvidia video card although I have!

 

[RussianSensation]

This whole post is sooo wrong...


...HINT: soon there will be no "primarily a gpu/graphics card company" (sic) anymore as both AMD and INtel slowly taking over everything except performance/high-end desktop and professional graphics card market with their upcoming combined CPU-GPU chips.

You think NV and AMD will stand still in the discrete space?? Both firms will continue to increase performance, even in the <$200 space. Just like notebook graphics will always be 1 step behind desktops graphics due to space/TDP limitations, embedded CPU graphics will also be constrained.

Also define "soon"? People who didn't care for gaming performance were already buying Intel onboards graphics for ages. Those same people won't care fore faster graphics either. For everyone that does care about graphics, the embedded CPU graphics aren't going to change anything either because they are still too slow.

 

[Dark Shroud]

You think NV and AMD will stand still in the discrete space?? Both firms will continue to increase performance, even in the <$200 space. Just like notebook graphics will always be 1 step behind desktops graphics due to space/TDP limitations, embedded CPU graphics will also be constrained.

The difference is that AMD can use the same R&D for discrete cards and APU GPU cores. The same goes for funding the R&D, for every dollar they spend it can be applied twice or three times over if you count FirePro.

Nvidia will has to be able to fund R&D for Tesla & Quadro. With APUs from Intel & AMD eating up the desktop & mobile (Netbook/Notebook) low & midrange markets that's going to hurt Nvidia's cash flow.

Also define "soon"? People who didn't care for gaming performance were already buying Intel onboards graphics for ages. Those same people won't care fore faster graphics either. For everyone that does care about graphics, the embedded CPU graphics aren't going to change anything either because they are still too slow.

By the end of 2011 all Intel desktop & mobile will be Sandy Bride based. The same can be said for AMD. So while older products will still use previous tech all new products sold will be APUs on the low end & mid level consumer market. You can also bet as businesses need to replace work stations they'll movie to devices running cheaper more efficient APUs.

 

[exar333]

This should bode well for PC games again though. Once computers come with a console-equivalent embedded GPU, thats a HUGE market for gaming. Thats a huge positive here.

It will almost be like back 15 years ago when you didn't need a 3D card!!!!!

 

[RussianSensation]

By the end of 2011 all Intel desktop & mobile will be Sandy Bride based. The same can be said for AMD.

So what? Can you buy H55/57 mobo with Clarkdale and have onboard graphics? Yes. Can you buy AMD mobo with integrated GPU? Yes. Did either of these initiatives kill <$200 graphics? No. Both firms are introducing APUs because it's cheaper to manufacture a CPU+GPU vs. a CPU + integrated GPU in a chipset package. Who needs discrete graphics? People who want to play 3D games. The vast majority of users who buy laptops wouldn't care if the graphics is 10x faster because they just don't play 3D games on them. We have already seen iPad eat into netbook/laptop sales because so many people only use laptops for Internet browsing and basic work.

I think you guys are vastly overestimating the performance of the first generation of APUs. Maybe AMD will bring something decent, but by the looks of it we are looking at 5600 performance level. SB GPU is a joke though. At best it will be as fast as G310M from NV/5450 (which can't really play any modern games in 2010, nevermind in 2011). By the time AMD ships Llano, we'll have even more advanced modern games. Sure Intel and AMD will have APUs, but they will be pretty useless for those who are buying discrete in the first place. Basically the increase in performance will be there, but it won't be enough just yet. The lowest GT410M from NV is already 2x faster than G310M graphics. So before SB-GPU even launched, it's already obsolete.

AMD has had faster integrated graphics compared to X3100/X4500HD for years. That doesn't seem to matter since people predominantly purchase Intel-based laptops. This is because most consumers care about CPU performance and battery life in the mobile space, where GPU performance is secondary. So the fact that AMD will have integrated APU in the mobile space is hardly relevant until they can make a decent CPU that can be fast and offer good power consumption/management. We'll still see SB with Optimus technology for those who buy discrete graphics imo. SB isn't going to change that unless you just want to play 5 year old games at 1024x768. But this is just my opinion.

 

[golem]

The difference is that AMD can use the same R&D for discrete cards and APU GPU cores. The same goes for funding the R&D, for every dollar they spend it can be applied twice or three times over if you count FirePro.

Nvidia will has to be able to fund R&D for Tesla & Quadro. With APUs from Intel & AMD eating up the desktop & mobile (Netbook/Notebook) low & midrange markets that's going to hurt Nvidia's cash flow.

If the low end discrete market goes away for Nvidia, it also goes away for AMD right? If this happens it hurts Nvidia a LOT, but AMD is crippled too. They lose all those discrete graphics sales and don't necessarily gain any APU sales.

A lot of people seem to think APUs are disaster for Nvidia (which it might be) and a boon for AMD (which most likely is not). An APU sale doesn't equal a CPU + GPU sale, not unless they gain market share against Intel or can charge a lot more for an APU vs a CPU alone.

The likely only really winner in this is Intel, they don't have discrete video cards sales to cannibalize.