Nvidia: Remember who got you where you are today
Posted on 12th Jan 2010 at 13:08 by Richard Swinburne with 43 comments
Literally one minute was devoted to mentioning that its upcoming Fermi products are "in production", but Nvidia didn't go so far as to confirm an actual arrival date. There was a working Fermi card showed off on its stand running the DirectX 11 Unigen engine, but Nvidia's biggest and most complex chip to date didn't even make the main stage.
With the entire presentation dedicated to its Tegra 2 products, this tips in the face of the original Tegra launch at Computex 2008, where the launch was shoved in mid-week and in the middle of the afternoon - hardly prime time.
For two quarters without volume production for anything over the
Clearly the business motives are "move with Nvidia's whims or you'll be left in the cold" and we don't think "loyal friend" even features in Nvidia's vocabulary. Other reports from Nvidia's press conference stated that it was a "strong showing from Nvidia and its partners". It's new, and profitable partners, yes.
I don't doubt that I'd love an Audi equipped with Tegra 2, and I'm interested in the upcoming Tablet PCs, but the closest thing we, its oldest fans and long time supporters - you know, the little guys - got to a new PC product was the GT240 - a 40nm, DirectX 10.1, GDDR5-enabled 9600GT that had a quieter release than a mouse with its butt sewn up farting.
For a company that's juggling ARM CPUs, workstation products, and the odd lol-worthy chipset and 3D-everything, Nvidia seems incapable of doing two things at once within one of those product markets. For example, we've given up waiting for the 40nm shrink and cost down of the GT200 architecture for the midrange.
I will certainly acknowledge Nvidia has moved at the right time with its ARM investment and massively successful marketing machine. With the explosion of smartphones and impending bomb drop of tablet PCs, it will do very well to bring us new products in many segments that previously may have been uninteresting.
So the future looks bright for Nvidia, but I doubt it will be long until we're not invited to most of their press events and launches, and you'll start to see the company favouring more mainstream tech publications and national media. It's once dominant consumer graphics line has already become a sideline to mobile and workstation computing will the company only do "enough" to keep up in the race with ATI, but how long until that no longer becomes financially viable? How long until the other dollars overflow the investors pockets and the strain between long standing techies who have devoted entire careers to gaming and performance graphics, stretches the company internally? I hope those engineers remind the big wigs who paid for those Ferrari's.






43 Comments
Discuss in the forums ReplyI do think that tegra is gonna be massive of course, but Nvidia are forgetting the very succesful Graphics business, and thats not gonna do them any good, afterall, they have many fanboys who only buy Nvidia GPUs, but if they're forced to buy ATI, then they'll slowly stop being such an Nvidia fanboy, and that'll lose them alot of their free marketing....
so thats alot of bad then
Ahhh... my bad, I thought I read somewhere that jen became the CEO in 2005 :?
Exactly! And a point that possibly should have been addressed. There are other shows and plenty of other means to demonstrate new GPUs such as the mythical fermi so I wouldnt get too down about it. It took ATI a year and a bit to release anything that could tackle the 8800gts never mind the 8800gtx. Itll all arive soon. With every company there is a yo-yo effect and this time ati happens to be at the top with their legs dangling.
Andy
Andy
+1
Myself being personally not very favorable towards Nvidia, I must say my zune HD with tegra can really browse a web page fast. I dont doubt they have a great processor going for them there. But in all reality - When has Nvidia ever done anything that did not have the highest possible fiscal reward for them? They are a business and as such they are perfectly able to see that the performance hungry games of former days are no longer existent on the PC. Heck they clearly have known for a long time that their GPU from 2005 can still run pretty much every game out there at the average joe 1280x1024 resolution.
So while we may want Nvidia to keep pushing the envelope and dumping millions into R&D in an epic battle with AMD, thats just not going to happen. Fiscally it makes no sense.
On the other hand our culture is seeing the rise of the casual gamer and they are a powerful foirce - people, and I say that classifying the vast majority, want to play Farmville (:(). Nvidia knows this, one needs only to look at the profits made by Zynga on facebook apps this year to know where they can make their money at.
Another huge indicator of where the market has shifted to is the PSP and Wii. On these platforms (missing citation) 45% of the gamers who play more than 1 hour a week of these two consoles are Female. Thats huge, where you once had a vastly male population of gamers you've now given yourself an outlet to the other half of the people in the world, the girls.
In conclusion: As much as we would like nvidia to stick around and play the ponies in the high end GPU market, they are going for the mutual fund and the emerging trends that will make them the money they so greatly lust for.
I was about to write something very simlar but nowhere as detailed.
I wouldn't put myself into the hardcore hardcore gamer envelope like a lot of people here. But I would say I'm a lot more clued up than the majority of people who play games these days.
I can see that it makes no sense for them to shove a load into R&D for new GPUs for PCs just now. There just isn't the market for it with very few PC games pushing the boundaries now and it looks like there won't be a new generation of consoles for a few years for them to try and get some product into either.
The casual gamer is where its at for making money just now. Unfortunately I think this means that AMD/ATI won't be pushing as hard to get cutting edge tech on the table either. However it may mean the end of "The way it's meant to be played" with games and GPUs being more universal.
Next, do some research into GPU sales to see where most sales occur. The casual gamer is not even the demographic where the dollars go. What they are doing is trying to add value to the gaming market by adding depth and breadth to the gaming experience with robust features such as 3d Stereo and improved physics. I realize some people say 3d is a gimmick - and then they admit they have not even tried Nvidia's 3d when I question them deeper. I demo 3d Stereo to many people at many events (PAX where I worked the Nvidia booth and many LANS where I bring my own setup including a 60" DLP that works with their 3d). People love it. Period. Heck, at Intel I had tons of Intel engineers with their mouths gaped.
Where do you think every GT200b in the last few months has gone? Tesla cards. Or, so I hear from a few industry peeps.
Nvidia Sterescopic 3D is a gimmick. It encourages people to use poor TN panels with poor quality and gets people onboard with the novelty effect. You can't game for hours in it, it gives everyone that has tried in the office a headache. There is a Stereoscopic 3D kit sitting in the lab and no one ever uses :p
I'm not saying 3D has a general rule of thumb is arse - people love Avatar for example, it's just I strongly don't believe Nvidia's locked down propitiatory solution is right for our "open" PC market.
AMD's lazier approach is not the right answer either, but whatever happened to interoperability?
SNiiPE_DoGG: I hear you and I agree, but what I'm saying is that the change is unceremonious and feels somewhat underhanded, which is a shame. I feel sorry for the companies that have grown up around Nvidia to support them.
I meant there are two clear paths here:
TEGRA: spend a lot on R&D --> manufacture for relatively small cost --> sell in unit volume of tens of millions for good profit.
Highend consumer GPU: Spend a lot on R&D --> Manufacture for high cost --> Sell in unit volume less than 5 million for slim profit margin.
It's pretty clear from this simple comparison that the money is not in in high end VGA at all anymore. Not to mention the points I listed in my previous post.
EDIT: Bindi - yes I know it is really a shame that they hang them to dry like that :\ especially after so many years of passing all of the CS and RMA off for them to handle the idiot masses
Oh I agree totally. But by this argument how long before performance PC gaming is in the *******?
In the end we will have a scenario where Nvidia could end up making an architecture for consoles, then ship the same derivative for the PC market and keep it until a new console arrives. Microsoft/Sony pay for the development, then the PC market is farmed off as a second thought that will always buy something?
Also, I do believe I successfully provoked a discussion today: Win me B)
Bindibadgi: I probably have come across many more people than you in terms of using this technology. I will go with my experiences on this one. The factor holding back the technology has been a diversity of 3d displays. That is changing this year.
Fair enough mate, but I still don't agree that the fundamental 3D technology will change things for many more years in the general public. TVs for movies, maybe, but gaming, I strongly doubt it.
Personally I find it a fad an would invest in a quality, larger panel, however I realise that more people like the "idea" that something is better, like a cheap HDTV, when it actually isn't. Marketing, not good quality sells products at the end of the day.
I've yet to really play a "3D" title that looks like it's anything more than layered 2D that has one trick: to throw things out the screen.
Also: will it be a case that we stop evaluating PC components on their own and in comparison to one another given the push to propitiatory standards? We already have an opening scenario where you must purchase one type of graphics card to play one game in full: Batman for PhysX, Avatar for Stereoscopic 3D, and in some respects, STALKER for DX11 - although that will change soon.
For all Intel's, or AMDs wrongs, at least they got it right with things like PCI, AGP, PCI-Express and HyperTransport..
Also: Tegra is taking products from GeForce line, not the other way around ;)
I love competition and have done projects for Intel and AMD/ATI - I support all three companies.
I think it does not really matter which direction the technology flows in the vertical Nvidia chain so long as it does flow in whatever direction benefits the end user.
I agree 3d is not the end all, it is a value added feature. All the monitor vendors and even DirectTV back me up on that assessment ;)
3D TVs using IPS displays still have the same issues as the TN panels. Namely, causing severe headaches and even nausea. I was speaking off the record to a number of execs at TV manufacturers during CES and the vast majority agreed with me that 3D is a fad, but it's here and so they've got to push it. They're just hoping that, with a united push, consumers will buy into it because it'll increase ASPs for both hardware and content (I'm mainly talking TVs/3D Blu-ray here rather than games).
Every seasoned journalist I speak to says the same and it's not as if I'm new to this game either.
3D works in the cinema because the big screen fills your peripheral vision . A 50in TV does not, quite simply, and if you combine that with the glasses that refresh at 60Hz, it's a recipe for unpleasantness. It's like using a 60Hz monitor all day, only worse because the sense of depth around the edge of the screen is completely fubared and you end up getting a headache, feeling disorientated and looking like a tw*t. ;)
You hit the nail on the headache.
DirecTV will say it's a value-add because they can (and probably will) charge more for their subscription service in 3D.
Because they've got to sell something now LCD prices are plunging and you get 24" monitors free in cereal packets these days.
I didn't see this the first time:
You are financially invested in it, so yes, I think you will try to push it as a benefit ;):)
When I can watch 3DTV without having to sit there with a headache or stupid, expensive glasses - I'll be the first in the queue.
Nuff said.
On one hand, this is quite possibly a genius move from the company's perspective, for several reasons it definately makes sense. With the rise of GPGPU applications predicted (albeit by nvidia themselves), the users already demanding high performance / server computing could switch to nVidia's Tegra platform if they can engineer a high performing product. I beleive this is a rather large market.
Also, If nVidia continue to develop GPGPU programming, ala CUDA, they could expand this market into other fields, possibly breaking into general, everyday areas of computing. This would create significant advantages for nVidia over competitors if they own the rights to the technology that becomes mainstream.
Obviously they are also targeting the more mainstream mass market (casual gamers) too, look at the high volume of GT250 GPU's as outlined in the article. Am I right in thinking these are the cheapest CUDA enabled GPU's too? This could work well for nVidia if they take advantage of the above theory. Don't forget about consoles either, the next-gen consoles will be capable of far more than gaming, my opinion is that they will turn into HTPC's with gaming as an add on.
On the other hand...it creates a distortion in 'our' market. Enthusiasts.
If nVidia, quite rightly, go where the real money is and leave AMD/ATI to gaming graphics, where is the incentive for progress? We may never play Crysis at 60FPS people! Competiton is clearly an issue for companies, but for consumers, it is highly beneficial. Innovation, progress and ultimately lower prices are driven by competiton. If nVidia reduce their presence in the market, or even if ATI follow suit, where does this leave us? We may have few sub-par, expensive gaming cards to play with.
well, that's my two cents anyway, I'm off to bed.
Ughh :Shudders: I need to go lay down now.
We are already there with software/games, hardware is just a matter of time.
What if the future of gaming is not high res moddable games that require the 3,000 dollar computer to play? What about a unified resolution between consoles and PC? - 1920x1080 for now but of course it could be raised in the future (2560x1440).
I can see quite a few benefits to this model (for arguments sake I wont put in the drawbacks), It could make out machines draw less power, run cooler, and ultimately be cheaper as we reach that threshold of performance (we are already there for this gen), developers might put more games out on the PC, PC gaming could become more accessible not necessarily to idiots but to lower income brackets.
I see a few benefits that would be good, not that I think its the outcome I want but IMO its better than the death of gaming computers all together.
If we could go your medium, I wouldn't mind at all seeing as everyone wins, well except people who like 16:10.
Aren't you based in London? Tell ya what, I'll swing by at Easter when I get back, you can give it to me, and I'll try it for a couple of months. I'll even
indemnify you against claims for damageswrite you a review at the end of it.Whaddya say? :D
I did laugh when nvidia pushed their propietary 3d format as a standard. It made me think about the other times they have pushed 'standards' which only benefit themselves and hurt consumers ie CUDA, PhysX. In the case of PhysX they even blocked using an nvidia card with an ATi primary (which was probably the only time i was going to give money to nvidia).
This also made me laugh:
some cash cow it is, MS have sold millions of Zune HDs after all ;)
Nvidia are looking more and more likely to pull out of our beloved enthusiast market, and personally i dont trust ATi as a single-horse after the HD5000 pricing (although to be fair it wasnt all their fault).
Someone doesn't remember the X1800 series...or the 8800 Ultra...
Hell back in '05 the X1600XT was $300!! No such thing as a midrange back then.
Oh and the X850XT PE selling for about $900 wasn't ludicrous at all right? :D
if you look at the advantages of the 5000 series one of the big ones is it has a small die area, so it should you should be getting a high number of cores per wafer. ie prices should have been lower (btw i dont mean the RRP, i mean the inflated prices that suppliers charged when there was no supply). This wasnt due to ATI directly, it was TSMC having crap yields and driving the cost of each core up (supply and demand 101).
Any company in a single-horse situation will charge what they think they can get away with because their first loyalty is to their shareholders, then to us.
What was that, a little bird on me shoulder says "Nvidia hasnt had top card since the 4870x2 came out!" That a looooooooooong time to be in 2nd place.
I will give nvidia credit for the 8800gtx which was a great card, but we are talking ancient history now.
They burnt any good will with me when they screwed around with games only made for their inferior equipment and silly ideas on how to implement physics.
Game over nvidia, try again next century!