NVIDIA Kepler speculation thread

[KimB]

AMD+linux=bad times usually(Though, not always). Still, you are correct. The GTX 650 was way over priced when it came out, and I honestly regret buying one instead of, say, a 7750 or 7770 simply due to the price I paid.

I guess it depends. I generally had fairly decent experiences with the open source drivers, but those have the issue that you can't do any remotely serious 3D with them: pretty much the most they're good for is desktop composition, and there's unfortunately no support for WebGL (which allows for significantly better experience in Google Maps these days). But they are nice for fast wake times from sleep mode (~3-5 seconds vs. ~20-30 seconds for proprietary drivers).

In general, though, I strongly prefer nVidia's Linux support for the desktop, though I use Linux on my personal machines much less these days.

 

[mczak]

I generally had fairly decent experiences with the open source drivers, but those have the issue that you can't do any remotely serious 3D with them: pretty much the most they're good for is desktop composition, and there's unfortunately no support for WebGL (which allows for significantly better experience in Google Maps these days).

WebGL should work generally though I think Chrome might still have some of the open-source drivers blacklisted (which can easily be changed). Though I'm not sure how well it works with SI, most likely much much better with bleeding edge drivers than whatever your distro shipped with .

 

[DSC]

http://www.4gamer.net/games/032/G003263/20130207052/

 

[I.S.T.]

It's well known that Agni's Philosophy was using a GTX 680.

 

[DavidGraham]

only a single GTX 680 ? or several ?

 

[DuckThor Evil]

Single.

 

[jimbo75]

Looks like Nvidia has now started combining their GPU segment with Tesla/Quadro for reporting -

Since the first quarter of fiscal year 2011, we have been reporting three primary financial reporting segments – GPU, Professional Solutions Business, and Consumer Products Business. During the fourth
quarter of fiscal 2013, we began reporting two primary financial reporting segments – GPU and Tegra® Processor. This change reflects the way we are nowmanaging our businesses internally.

Revenue was well down on Q3 but they still had a decent quarter overall. Their outlook for Q1 is way down on what the market thought it would be though so there must be a pretty high chance of the stock going south today.

http://phx.corporate-ir.net/External.File?item=UGFyZW50SUQ9MTcxNjMzfENoaWxkSUQ9LTF8VHlwZT0z&t=1

 

[DuckThor Evil]

Must have been a pretty good quarter if that's your worst spin
Where do you get the "market thought" for Q1.

 

[jimbo75]

Must have been a pretty good quarter if that's your worst spin

It was a decent quarter compared to last year but everything I said would happen did. Their entire GPU business (including professional now) was down 7% on Q3. Tegra was down 15% on Q3. I would suggest that their old "GPU" business lost a lot more than 7%, hence the need to hide it by combining with professional all of a sudden. This will prevent some people from asking why the (old) GPU business is cratering in 2013, but will do nothing to prevent it's actual decline.

Where do you get the "market thought" for Q1.

http://www.reuters.com/article/2013/02/13/nvidia-results-idUSL1N0BDFUJ20130213

That is quite a scary dip - almost $170 million down on Q4.

 

[DuckThor Evil]

It was a decent quarter compared to last year but everything I said would happen did. Their entire GPU business (including professional now) was down 7% on Q3. Tegra was down 15% on Q3. I would suggest that their old "GPU" business lost a lot more than 7%, hence the need to hide it by combining with professional all of a sudden. This will prevent some people from asking why the (old) GPU business is cratering in 2013, but will do nothing to prevent it's actual decline.

In the morning when I read the figures I was 100% certain that you'd come up with that angle. In reality it makes sense to group them together. Did you notice that the Intel settlement is no longer counted in the GPU segment? 50% YoY profit increase is "decent"

You've were saying that Q3 was just a one off anomalous good quarter for them, because AMD let them have it, but let's just wait and see what Q4 looks like. That's what you said...

Thanks for the link. It seems nVidia is predicting a similar quarter to year prior.

 

[jimbo75]

Yes that's what I said. I said they had a one-off good quarter and that revenues would be down in Q4, and that's what happened.

Total "new GPU segment" revenue was down 7% even though Quadro revenue was up 10%. We don't know how much Tesla was down but we can still figure that Geforce was down a lot. It might make sense to combine them but Nvidia is full of these kinds of tricks when things aren't looking good.

They've just been cut to hold from buy by Needham - http://www.marketwatch.com/story/nvidia-cut-to-hold-by-needham-2013-02-14 and it won't be the last.

Their discrete GPU business is treading water and Tegra is a millstone around their neck. Faster IGP's, better tablet processors and Intel now vying for their share of the market means Nvidia is a company in very severe trouble. They'll be fine for the next couple of years but things look grim 5 years from now imo if they don't do something about Tegra fast.

 

[silent_guy]

Looks like Nvidia has now started combining their GPU segment with Tesla/Quadro for reporting -

Yup. As I said, it's the only meaningful number.

Revenue was well down on Q3 but they still had a decent quarter overall. Their outlook for Q1 is way down on what the market thought it would be though so there must be a pretty high chance of the stock going south today.

A 0.28% shellacking at this very moment!

 

[Homeles]

Their discrete GPU business is treading water and Tegra is a millstone around their neck. Faster IGP's, better tablet processors and Intel now vying for their share of the market means Nvidia is a company in very severe trouble. They'll be fine for the next couple of years but things look grim 5 years from now imo if they don't do something about Tegra fast.

You are greatly exaggerating things.

 

[Alexko]

Yup. As I said, it's the only meaningful number.

A 0.28% shellacking at this very moment!

The stock took a significant dip yesterday in afterhours trading, but went back up this morning. It's pretty strange, I'm not sure what happened.

You are greatly exaggerating things.

I don't think he is. It's not hard to see that discrete graphics will become far less common in a PC market that, itself, may keep shrinking. Since Tegra has yet to turn any profit, it's pretty reasonable to be concerned about NVIDIA's future.

 

[Ailuros]

You are greatly exaggerating things.

Not really. The PC business is stagnating faster than their Tegra market share is rising.

 

[ams]

Looks like Nvidia has now started combining their GPU segment with Tesla/Quadro for reporting -

This is somewhat logical since GPU technology is shared between Geforce, Quadro, and Tesla.

Revenue was well down on Q3 but they still had a decent quarter overall. Their outlook for Q1 is way down on what the market thought it would be though so there must be a pretty high chance of the stock going south today.

Have you ever heard of a thing called "seasonality"? NVIDIA's Q3 revenue is typically higher than Q4 revenue because NVIDIA's partners need supply in advance of the holiday season. Q1 revenue is typically lower than Q4 revenue because NVIDIA's last generation Tegra products ramp down while next generation Tegra products ramp up, and because less supply is needed by NVIDIA's partners after the holiday season.

They've just been cut to hold from buy by Needham

That makes some sense. With respect to NVIDIA's core businesses, there is very little to announce over the next few months (other than the limited edition Geforce Titan). The proper time to buy NVIDIA stock would be some months from now when NVIDIA is closer to delivering next generation tablets and smartphones, and when NVIDIA is closer to unveiling Project Denver.

Their discrete GPU business is treading water and Tegra is a millstone around their neck. Faster IGP's, better tablet processors and Intel now vying for their share of the market means Nvidia is a company in very severe trouble. They'll be fine for the next couple of years but things look grim 5 years from now imo if they don't do something about Tegra fast.

NVIDIA will be in a much better position in 5 years than today, as the investments made into mobile computing, software-defined baseband radio, and custom low-power CPU/GPU cores will open up new market opportunities for the company. Tegra will be the building block for future Geforce GPU's.

 

[ams]

The stock took a significant dip yesterday in afterhours trading, but went back up this morning. It's pretty strange, I'm not sure what happened.

After-hours trading is notoriously volatile and unreliable due to relatively limited trading volume.

 

[AlphaWolf]

NVIDIA will be in a much better position in 5 years than today, as the investments made into mobile computing, software-defined baseband radio, and custom low-power CPU/GPU cores will open up new market opportunities for the company. Tegra will be the building block for future Geforce GPU's.

This presupposes that others investing in those same areas will be less successful than Nvidia.

 

[Kaotik]

This presupposes that others investing in those same areas will be less successful than Nvidia.

Indeed. (and have they actually said they're making custom CPU cores? So far Tegras have used standard ARM Cortex-Ax cores)

 

[jimbo75]

Have you ever heard of a thing called "seasonality"? NVIDIA's Q3 revenue is typically higher than Q4 revenue because NVIDIA's partners need supply in advance of the holiday season. Q1 revenue is typically lower than Q4 revenue because NVIDIA's last generation Tegra products ramp down while next generation Tegra products ramp up, and because less supply is needed by NVIDIA's partners after the holiday season.

Q4 revenue is not typically lower than Q3 revenue. In 2011 it was lower, in 2010 and 2009 it was higher.

NVIDIA will be in a much better position in 5 years than today, as the investments made into mobile computing, software-defined baseband radio, and custom low-power CPU/GPU cores will open up new market opportunities for the company. Tegra will be the building block for future Geforce GPU's.

How long ago was it they started investing in Tegra? Since then it's just been a money sink and they aren't really close to breaking even. Their OpEx has ballooned as they've hired more and more people into an unprofitable business.