Core Values: What's next for NVIDIA?
By Anand Shimpi 
posted December 4th 2009 3:00PM
posted December 4th 2009 3:00PM
Core Values is our new monthly column from Anand Shimpi, Editor-in-chief of AnandTech. With over a decade of experience poring over the latest in chip developments, he's here to explain how things work and why our tech is the way it is.
The last time NVIDIA was this late to a major DirectX transition was seven years ago, and the company just quietly confirmed we won't see its next-generation GPU, Fermi, until Q1 2010. If AMD's manufacturing partner TSMC weren't having such a terrible time making 40nm chips I'd say that AMD would be gobbling up marketshare like a fat kid. By the time NVIDIA gets its entire stack of DX11 hardware out the gate, AMD will be a quarter away from putting out newly refreshed GPUs.
Things aren't much better on the chipset side either -- for all intents and purposes, the future of NVIDIA's chipset business in the PC space is dead. Not only has NVIDIA recently announced that it won't be pursuing any chipsets for Intel's Core i3, i5. or i7 processors until its various legal disputes with Intel are resolved, It doesn't really make sense to be a third-party chipset vendor anymore. Both AMD and Intel are more than capable of doing chipsets in-house, and the only form of differentiation comes from the integrated graphics core -- so why not just sell cheap discrete GPUs for OEMs to use alongside Intel chipsets instead?
Even Ion is going to be short lived. NVIDIA's planning to mold an updated graphics chip into an updated chipset for the next-gen Atom processor, but Pine Trail brings the memory controller and graphics onto the CPU and leaves NVIDIA out in the cold once again.
Let's see, no competitive GPUs, no future chipset business. This isn't looking good so far -- but the one thing I've learned from writing about these companies for the past 12 years is that the future's never as it seems. Chances are, NVIDIA's going to look a lot different in the future because of two things: Tesla and Tegra.
NVIDIA also made a decision to make its GPU architectures much more flexible, a decision that resulted in the G80 chip at the heart of the GeForce GTX 8800. At the same time, NVIDIA began investing in programming languages to make writing for its flexible GPUs much easier, and also fed HPC feedback into its GPU design cycle: the GT200 was more HPC friendly than the G80 and Fermi is even more HPC friendly than GT200.
NVIDIA believes there's roughly $1b to be made in these HPC markets over the next 24 months, and although it's only made a bit over $10m so far, the company thinks that Fermi is going to be the turning point for Tesla revenue. Let's be realistic, though: at its peak, NVIDIA used to pull in around $1b in a single quarter. Tesla alone won't be enough for NVIDIA, not at those numbers.
Tegra is NVIDIA's SoC brand -- as we talked about last time, the smartphones we love reading about are based on highly integrated SoCs (system on a chip). That's a CPU, GPU, some other specialized processing, memory/storage and maybe even a modem. Tegra contains nearly all NVIDIA-developed technology -- and like everything else in the smartphone space, it's based on ARM, which means NVIDIA won't be dependent on x86 CPUs that will soon have integrated GPUs.
While Tesla depends on NVIDIA's continued development of high end GPUs, Tegra does
If push comes to shove, Tegra has enough upside to let NVIDIA exit the PC business entirely and just make smartphone chips. |
But going from zero to significant market share in the SoC space is difficult. The established players there are companies like Marvell, Samsung and Qualcomm. Even Intel looks like an unlikely underdog in that market.
Although Tegra got a lot of attention with the Zune HD, it's based on an older ARM11 core with the usual general purpose performance shortcomings -- and it doesn't necessarily look so hot compared to other performance oriented SoCs that have moved to Cortex A8. That said, NVIDIA plans on updating its Tegra SoCs once a year, similar to its GPU update cycle. Given the slow level of progress we've seen in the SoC space, there's room for NVIDIA's Tegra approach to do well -- update it annually and it may end up being fast enough to raise a few brows.
Looking at it this way, the biggest threat to NVIDIA today doesn't come from Intel or AMD, but rather Imagination Technologies, whose graphics cores are heavily used by Apple and Samsung in smartphones -- and NVIDIA believes its strengths as a GPU maker on the PC side will give it the advantage here. I'm willing to give NVIDIA the benefit of doubt, but we had better see a big splash in 2010 with Tegra and Tegra 2.
So. Will NVIDIA remain a high end GPU maker for PCs, will it see success in HPC, or will it move entirely to the application processor/SoC space? That future is at least a few years away, and as AMD has already shown us, a lot can happen in a few years. A lot more than I could predict at least.
[Zune HD image courtesy of iFixit]
Anand Shimpi is CEO and Editor-in-chief of AnandTech. Contact him at anand AT anandtech DOT com or on Twitter at @anandshimpi. Views expressed here are his own.
@kleptophobiac
intel isn't very good about doing supports for drivers unlike nvidia, you get updated drivers almost monthly or quarterly and the drivers trickle down to older cards unlike amd/ati or intel i740???
anyone that thinks Nvidia is going anywhere is a fool. Nvidia owns the majority of the discrete graphics card market. AMD has an uphill battle
Well, I'm an ATI/AMD fanboy; but I don't want anything of this sort. My roommate is an nVidia fanboy (a shameless one) and I really don't want to hear him crying has he install a 5870 into his computer. I would have to hit him.
And of course the market needs competition. Same way that if AMD disappeared Intel would slow down research and be able to raise prices, ATI/AMD would surely do the same if nVidia were to say goodbye to GPUs for PCs. That would be terrible.
I'm quite confused with Nvidia and intel stuff....
So currently Nvidia won't be making any new GPU for intel i3-i5-i7-i9 chips until the lawsuits end??
so it means that those intel CPUs wont be able to use Nvidia GPU??
but why one of my friend is using an intel i7 cpu with nvidia GPU??
and why does manufacturers like Apple dont include Nvidia GPU into its product that uses i7 core?? (they changed to ATI)
@touchaponk
GPUs are fine, nvidia can't make intel compatible chipsets due to licensing bs.
http://en.wikipedia.org/wiki/Chipset
@touchaponk
Chipset, they won't be making chipsets. GPU's they will make a ton of.
I'm still waiting to see what P.A. Semi does.
What a great FRACKING article...
Anand rules. Glad to see he's pulling down an Engadget size audience.
Great article, too. A little short for what I'm used to reading from Anand, but I'm assuming it was made to order. :)
Considering that near 70% of the world's enthusiast computers contain nVidia GPUs I would be beyond stunned to hear that they were even remotely considering dropping GPUs in favor of SoCs.
nVidia has many times upped the bar for what GPUs can do and what they offer and I would be truly surprised to see them fail at maintaining that trend. To suggest nVidia stop making GPUs is like suggesting Intel would stop making CPUs. They're the significant market share holder and even though they have other products than CPUs or GPUs, they're not going to drop their flagship line in favor of the products they produce that have marginal (at best) market penetration.
NVidia chipsets have always weighted in with relatively high power consumption. Good luck transitioning to SoC, where power is of the utmost concern.
I still cant believe AMD sold their relatively successful 3d System-on-Chip (SoC) intellectual property (IP) to Qualcomm. They were one of two vendors, themself and PowerVR, and now its just PowerVR selling 3d IP.
@rektide
that's because AMD was hemorrhaging for money... if they didn't they probably would have kept it.
Intel still can't make a decent GPU. If they push nVidia out at this point, we consumers looking for decent performing netbooks that can do 1080p or 3D will lose.
I hope that NVIDIA steps back up to the plate and improves. I used to prefer their cards to ATI's, but the last three NVIDIAs I've owned all burned out on me within weeks.
It must be nice to be filthy rich!
Nvidia is not going anywhere. It's the same story like every other GPU generation. One company releases a card first, it trashes everything out, then another company releases their competing card, which destroys the former's releases. they been battling like this for as long as I could remember. What's really halting Nvidia efforts is intel, which seems bent on sole power of the x86 architecture...I say we force intel to make the architecture open, or at least force em to licenses it unbiasly. That way we'll probably see a lot better competition in the gpu cpu market.
in nintendo put a tegra chip in the next gen ds, nvidia will be just fine.
So... when is a basement GPU company going to start up to sue AMD for having a monopoly?
@NinjaMonkey
why amd? intel makes up like 50% of the graphics market.
Tegra has been a failure. Why? Few reasons.. it's based on Arm11. Newer and more efficient SoCs based on Arm CorTex A-8 (Omap3, etc.) are already being used on so many smart phones (palm pre, droid, iphone3gs, omniaHD, n900). And then there's CorTex A-9, which is what the OMAP4 will use -- completely blows the Tegra out of the water. If the Tegra actually had some decent OEM's on board (HTC, Nokia, Samsung) and they actually had devices out that are using it right now, then it would be different. But there aren't. Even the latest Tegra netbooks (mobinova) are laughable compared to Atom alternatives. 2nd reason was kind of mentioned -- it's taken so god damn long to come out. It was announced what, almost 2 and 1\2 years ago? And there's only ONE decent device using it -- the ZuneHD. Compare that to the Omap3..which is used in tons of other popular devices, uses less power, and is faster. For more info, check out an article called "Myths of the ZuneHD" -- which basically debunks the Tegra more than it does the ZuneHD.
Anyways -- NVidia needs some new management, that's for sure.
The Tesla is nice though..I'll give them that.
Mature, insightful, very well informed... Engadget needs more articles of this caliber...nuff said.
@Generic Dude
Unfortunately not every writer on engadget has an engineering degree...
Maybe tegra and tesla can make it to the desktop at some point? I don't really care weather I use Nvidia or AMD although the 9400m was a pretty good bit of kit for laptops like my MB Pro.
But I do care about the fact you need 2 competing manufacturers to move technology on and keep prices low.
Be interesting to see what happens but I agree, they are finished for the moment, in 4 or 3 years who knows?
The problem I have with Intel is their GPUs suck. So the next atom will still suck without some discrete GPU frosting.
I'm a video guy. By that I mean broadcast, cable, satellite, etc. And on this front Intel sucks too. Case in point: on windows DXVA has the ability to set the dynamic range of the video allowing for the selection of PC levels (0-255) and TV levels (16-235). On intel it only does 0-255, and all calls to set TV levels are ignored. This makes Intel GPUs USELESS for the living room. Intel will only fix this if it is a WHQL requirement. IE: unless MSFT holds a gun to their head they just don't give a shit. That's also why on HD content you only get primitive BOB deinterlace. Film detection for HD? LOL.
Intel moves their shitty integrated GPUs via bogus pricing constructs that freeze out folks like nvida ION.
Atom exists to allow Intel to continue to ship their old obsolete chipset solutions. Nothing more.
@(Unverified)
Agreed on all points. It's a shame that Intel allows this, I mean, why not improve their GPUs? They have enough money and resources to develop them more, but they don't...they just pass off their inadequate offerings on to us and expect us to like them.
@(Unverified) If people would actually stop buying their offerings then they would step it up. But people appear content with garbage, so that's what they keep giving us, in terms of integrated graphics that is.
@(Unverified)
intel's point of bad GPUs is to offload work to CPUs so that they can selll more CPUs. Otherwise consumers will claim that their CPU is powerful enough and have no need to upgrade...
It would be a sad day in history if Nvidia would stop producing GPU's. Ati/Amd would have no competition to speak of and prices would go up. Right now they kept prices conservative because there were Nvidia cards that performed slightly better albeit at a steeper price. Now with the 5970 there is nothing that can beat it, result ... well I've seen cards go up to well over 600 euro's over here.
Even if you don't like them, we need healthy competition.
I bought an Nvidia sound card and it died after a year. They need to get these right first.
What ever happened to the GT380 or GTX380 or the GPU that was supposed to compete with the 5870. I realize they had some fabrication issues but really - still no card even going out for review?
With Nvidia spending almost $900,000,000 on R&D last year, I think they'll figure something out.
"Let's see, no competitive GPUs, no future chipset business" -- Anand, your logic and reasoning skills need some work.
I want to know when Microsoft is going to release the Zune HD 2 supposedly with Tegra 2
A lot of Nvidia fans and enthusiast has been waiting on what the company has been cooking to fight with the recent world's fastest video card
The 5970 Radeon: http://bit.ly/worlds-fastest-video-card-made-by-radeon-very-remarkable
Little did we know, that Nvidia is infusing a 'nearlike' tesla-trooping technology to boost the power of the new gpx card on an insane level.
Kudus! Hopefully the price of the new card, won't tesla-toast us though...
I read on Anandtech once that Geforce FX was a massive failure, even compared to Fermi (so far), but Nvidia survived even that. I don't think unless Fermi is utterly horrible nvidia will have quit the GPU space - they've survived worse.
The chipset business is dead, certainly, and Ion is dead in the long run. Why aren't more devices using Tegra?
@YpoCaramel
soon... nintendo ds and tegra,,,
Did a double take when I saw Mr. Shimpi on here!
Started visiting his site when I was a wee boy of 12.
Taught me stuff that was re-confirmed for me my second year computer science courses. No word of a lie.
Glad to see he is branching out though! Wonder if this competes with dailyTech though? I thought that was a sister site of Anandtech? Seeing as how it was started by Kris, a long-time writer at Anandtech himself?
Anyways, yes, Nvidia, and doomsday bells because they're late to on graphics cycle?
From what Anandtech has taught me about the video card industry, I understand that the teams work in parallel, so the team designing the successor to Fermi should be on track, or only slightly behind due to the Fermi delay, no?
And come on, this is the company that suffered the FX series and Radeon 9700 in the same year and still survived.
I usually cheer for the under-dog which until recently was AMD/Ati, and I believe it still is. But honestly, Nvidia is not going anywhere from the GPU business, take this armchair experts word for it! :)
It is soooo nice to see a well-written and edited article that employs decently substantive analysis... are you listening Engadget "editors"?
So this is what Nvidia has been up to lately, Not to say that the Tegra opportunity is a needle on a hay stack, but I don't like the idea that Nvidia
will stop manufacturing PC GPU cards in the future.
Detailed point of view about Tesla and Tegra: http://bit.ly/tesla-and-tegra-what-are-they
A very nice howdy writeup from Anandtech CEO
Yes, the current Tegra is one generation behind the competition, but the new tegra (tegra 2) will be based on ARM Cortex A9, so they will be on par with the other ARM SOC manufacturers when they come out with their next gen products in 2010.
They may even get ahead if they manage to put a very good graphics GPU in Tegra 2 and put at least 2 A9 core in the SOC.
What worries me is that they likely won't be able to use GlobalFoundries' 28nm process which seems to be a killer when it comes to producing Cortex A9 SOCs.
i cant help but be a fanboy for the losing side, because that side always has the most potential to wow us, its such thrill to see what comes out the doors of a desperate company, so now what will nvidia do?
Hey Remember a few months ago when Nvidia's CEO was saying how much he loved Apple? Maybe we will see Tegra powered iPhones? I think you could say that would be pretty huge for them. No?
Anand Lai Shrimpi!!!!
Take notice! Engadget finally has someone credible to do silicon commentary! Anand and Anandtech rule.
ATI Threw in the Towel by releasing a more powerful GPU. Right....
This all smells like "Plan B" to me. Or maybe even Plan C. Nvidia's core business is graphics, anyone who claims otherwise is a liar, and they've failed at that.
Now they're working on their last resort, and if they fail at it, they're done for.
@FuturePastNow
any good business with depth should have back-up plans if they want to survive...
I loved this article! Well written and very clear. Much better than that Entelligence garbage I've read.
not sure if I'm ready to buy