Nvidia Tegra roadmap slips a year
Welcome to T35 and Tegra 6
Oct 19, 2011 in analysis, Channel, Finance, Humor, Microprocessors, Rumors
It looks like Nvidia (NASDAQ:NVDA) is keeping to their usual form with Tegra, the schedule just slipped a year. With that we welcome the new chip in the bunch, T35, and the attendant schedule waterfall.
If you are not up to date on the current Tegra roadmap, up until a few weeks ago, it looked like this for Tegra 3, 3.3 and 4, with Tegra 5 and Denver following. Before we move on, a quick update. Insiders tell us that Denver is the T50 core, T50 is the chip family based around it, not the other way around. Or at least it was before everything slipped a few weeks ago. And that brings us back to the current situation, and the ‘new’ chip.
The T30 and T33 speedbump chips are still intact, with chips going on sale to the public at the end of September according to Nvidia’s 2011 Computex announcements. That may not have quite panned out as everyone paying even a bit of attention knows, but the chips should ship by the end of the year. T33 will follow in a bit, it is just a mild reworking/spin of T30, and it’s purpose is to lose less to Qualcomm’s Krait. It won’t, but let’s not shatter any dreams yet, Qualcomm should have an answer to keep the benchmark status quo.
From there we come to the new chip on the block, or at least the newest resurrected chip on the on block, T35. If you recall, T30/33 is a 40nm quad-ish core A9, and T40 was originally meant to be a shrink of that to 28nm. A15 based parts were going to be skipped in order free up engineering resources with an ultimate goal of bringing the A8/ARM-64 ISA based Denver to market first. Nvidia was publicly promising financial analysts that Denver based parts would ship in late Q4/12 long after they knew there was no chance of that being possible. Short term stock bumps seem to matter more than engineering at some Santa Clara based chip companies, legal concerns be damned. Consistency is a an admirable trait.
This was a decent idea, medium-term pain to get long-term gain, but the execution was not close to feasible. Part of the problem was Dear Leader’s constant micro-managing of technical specs, and the attendant wholesale changes. Moving Denver from Kepler to Maxwell cores assured that there was no way it would be out in 2012, and put 2013 as a stretch, but that somehow didn’t make it to the IR briefings. Curious that.
With Denver slip-sliding away, there was a little more breathing room for the beleaguered engineers. Between that and the woefully uncompetitive nature of the A9 cores in the A15 world of 2012, something had to be done. Again, Dear Leader rightfully deemed the A9 cored T40 to be too slow, and single handedly ordered it changed to an A15 design. Fair enough, unless you are working on the chip, then life sucks.
This change made T40 much more competitive on paper, but pushed the schedule out quite a bit. This now extended gap between T30/T33 and T40 is not a good thing, especially in light of the competition. Because those lazy engineers at Nvidia could not meet Dear Leader’s 180 degree change in vision while keeping the original schedule, it slipped more. No shame in that, the fact that there are still employed T40 engineers is testament to their strong will and perseverance.
What does a company do when they are facing a year+ without a competitive core? Resurrect a stopgap, and call it T35. This ‘new’ part is simply the old 28nm shrink of Tegra 3/T30 that was called T40 before the A15ing. Fair enough, and it is a good stopgap, but what do you call it? Tegra 4 of course.
Yes, Tegra 4 is now a 28nm Tegra 3 once again, and the A15 based T40 is now Tegra 5. Because word of massive schedule slips would likely hit management in the portfolio, this change is now deemed to be a silent one. T40 is now Tegra 5, Denver/T50 is now Tegra 6, and T35 was meant to be there all along, honest. Nothing to see here analysts, smile and move along.
In the end, the Tegra roadmap just slipped a year. Nvidia faces a rather painful 2012 and likely early 2013 with a core that is a generation behind, so expect lots of self-congratulatory press releases. In the mean time, Qualcomm is releasing Krait, TI has OMAP5, and Nvidia has hot air. Anyone want to join our pool for when the stock bubble bursts?S|A
81 Responses to “Nvidia Tegra roadmap slips a year”
Trackbacks/Pingbacks
-
-
Oct 20, 2011
[...] Web site SemiAccurate reported Wednesday that the Tegra roadmap “just slipped a year,” citing unnamed [...]
-
-
Oct 20, 2011
[...] says Nvidia Tegra slips Link Categories: FA Notes blog comments powered by Disqus [...]
-
-
Oct 20, 2011
[...] has OMAP5, and Nvidia has hot air. Anyone want to join our pool for when the stock bubble bursts? Nvidia Tegra roadmap slips a year | SemiAccurate [...]
If performance is adequate and the price is right and the right deals are struck it doesn’t matter if you aren’t the absolute fastest. And it’s not like consumers pick the graphics/chips in mobile devices.
A nice example being that the original tegra although being much faster than the competition at that point took a year to get adopted by the device manufacturers.
Not to mention how important software still is in the mobile devices, the right software can double the speed to the enduser since it’s still a bit messy at this point.
If android 4.0 was created for dual cores and is optimised for dual cores, will a quad core even be effective? We shall see on november 9th
The best ARM chips are by samsung hence the popularity of the gs2.
Samsung is making another exonus chip for galaxys3
Ill like the poinst from both camps above.
While some might think nvidia is the only one thinking ahead because of their map, I don’t think this is entirely true.
Don’t sleep on samy!
Only 70% of the Android tablet-market is powered by Nvidia – such FAIL!
which if these (http://www.gartner.com/it/page.jsp?id=1800514) numbers are right means they have ~12% market share…..woooo!
So out of nowhere, just like that, a company enters into the ARM market and has scooped up 70% of all the market android market. ppl may poke fun at this but Nvidia has become extremely relevant in the small ARM space that apple cant fill.
This is their 2nd CPU mind you. The many competitors are taking Nvidia very seriously yet there are some who still see tegra as a failure, a chip that isnt selling????
its only been months and nvidia has 12% of the overall ARM market! Thats not far from AMDs x86 market share they have today. This is only their 2nd tegra chip!
spin it as you may, its nothing less than impressive.
“Dear Leader”
Charlie loves being a racist.
Same Charlie that reported that Global foundries 28 nm would be in mass production early 2010, that iPhone would have Thunderbolt and countless other misses.
But you can’t discuss this on the board, since you get banned for trolling.
Semiaccurate: we love discussion as long as everybody think like us.
Charlies articles would be so much better without the rasism. But that may not draw as many clicks.
Love AMD.
Love Radeon.
Love Android.
Hate Intel, Thunderbolt, Nvidia and Apple.
Nater – We have removed your defaming comment submitted to our site. Defamation is an actionable offense. Please stop defaming Charlie. We thank you for providing us both an email address and and IP to track the other similar defamatory comments you might have left on other websites.
For your education:
An example from MN: We hold that a private individual may recover actual damages for a defamatory publication upon proof that the defendant knew or in the exercise of reasonable care should have known that the defamatory statement was false. The conduct of defamation defendants will be judged on whether the conduct was that of a reasonable person under the circumstances.
Jadwin v. Minneapolis Star & Tribune Co., 367 N.W.2d 476, 491 (Minn. 1985). Other cases follow this reasoning. See LeDoux v. Northwest Publications, Inc., 521 N.W.2d 59, 67 (Minn. App. 1994) (“In order for a statement to be defamatory . . . it must be false.”); Janklow v. Newsweek, Inc., 759 F.2d 644, 648 (8th Cir. 1985), cert. den., 479 U.S. 883 (1987) (“Libel, by definition, consists of publication of a false and unprivileged fact.”).
North Dakota law cites much the same types of case law and examines similar requirements for defamation. We will directly inform you, repeating what Charlie has stated in public writing in the past, he has never been fired from a job nor been pushed out.
Excuse me . . . http://www.theinquirer.net
Charlie
Your comments are getting progressive more inane, to the point where they’re almost not even worth reading anymore. You really need to let go of this ridiculous hatred you have against nVidia.
I like how any time AMD screws up, you basically either gloss over it with an “in depth technical review” (ala Bulldozer), or the like, yet any time nVidia so much as makes a single misstep, you rant about how it’s the end of the world over there.
Despite them making some fairly serious faux paus over the past year or two, they still continue to be far and away financially more successful than AMD.
Bottom line is that survival of businesses comes down to fiscal return, not your approximation of their technical prowess.
Even *you*, at the end of the day, need nVidia. Otherwise you’d have little else to make inane comments about. If you know so much about silicon design, why don’t you go get a job at one of these “failing” companies and help them turn things around?
he do that what u say with reason .
when AMD do flop they don’t say wel it must be that way ,, they don’t mask truth (wooden parts in nv580) ,they say this is reason we think lazy programers will learn how to use mp coding but they didnt
so we got what we got ,and even this wasnt that fast like intel version ,it still compete with i7 not i5 or i3
so it’s not great but oki
nv side go like this ,, nv5 gpu flop
well we need new better drivers than all will be oki ,
new drivers fuck-up img resolution to gain speed in tests ,than they say ati do same with 7500 8500 gpu’s
still flop and they hide they don’t say oki we fuck-up and this is reason why we cant have speed of it ,and more technical explanations ala bulodzer
if they did it like this ,ppl will see oki it’s human to make errors ,gl next time
but nooo there beloved leader (someone say north korea :) ) order them to lie
to make false date to cheat to grab last straw if they can to go out in shine
u read that kind tactics before that ha????
yea that is reason why he call him beloved leader ,full of shit and lies .
The “beloved leader” is a skilled technician that has managed to create a company from nothing and make it the biggest and most advanced graphics designer of the planet in 10 years.
This company, lead by the “beloved leader” makes money, and has always demonstrated to be able to see the future a bit before the competitors.
All competitors died but one, which was saved by AMD paying 3 times its value in order to avoid it to become someone else prey.
The “beloved leader” managed to enter (and somehow create) a market which was not part of its core business, like the mobile SoC one, while AMD took 4 years to create an Atom competitor just in time for the market to die out (Atom class devices market is fastly shrinking).
The “beloved leader” has plans to assault x86 castle in a couple of years, something that no one could ever think possible by anyone few years ago (even IBM could not do much against it even with their superior Power technology).
x86 is not only desktop/notebook. Servers and HPC are even more important.
The “beloved leader” has the capacity to take a far away, seemingly impossible to achieve target, and take the path to make it happen.
The path is not always smooth and straight, but he marches on it.
You can say anything about JJH, but not that is not able to reach the point he wants to arrive to. Would AMD have such a leader, today they would not have been in such a situation, late on market with their products both for performance and for design.
P.S: anyone that understand multi-threading knows that not many daily algorithms can be split over different parallel paths. Investing and betting on the development of vastly multi core chips for home/office use is quite stupid.
Multi-threading is good in server and HPC only. BTW, in the end, AMD wasn’t able to come up with an architecture that can compensate the efforts needed to split an algorithm even when it is possible: an 8 core (requiring 8 threads which is an amount of threads useable only for few algorithms), is as fast as a 4 thread architecture which is vastly superior in everyday work where you use 1 or at most 4 threads.
Sorry
Noone. Thanks for a sane posting.
Its about making money. Nvidia makes money.
Nvidia knows that discrete graphics are declining. They at least try to grow a new market with Tesla and Tegra.
Nvidia alos believes that X86 will decline and be replaced by ARM. I think so also. (most users here have not used a computer without X86. Non X86 computers are in a whole different class and ARM will bring this to consumers at 1/10 price that AMD/Intel are over charging)
I am not sure if AMD understands that.
Bulldozer missed clock goals by 30% and is 9 month after roadmap.
ARM market is expanding 50% each year. The X86 market grew under 5%.
Already next year 100 million tablets will be sold. Many of them replace cheap X86. Nvidia will at least have 20-30% of that market.
If tablets are computers, they will already have 20% market share next year.
Do AMD have a backup plan or are they going out guns blazing with discrete graphics and X86 CPUs?
I want AMD to survive. Banking on X86 is not the way to insure survival.
I remember an article on here a while back (2-3 months) in which Charlie was reporting on some developer conference that AMD and ARM were participating in and he was trying to insinuate that the two companies were heading in a processing direction that they hoped Intel couldn’t compete with but… at this point I’m not too confident in AMD’s ability to execute. Even the fab problems that GF is having can’t simply be bad luck for them.
So… AMD’s backup plan is to lose all its market value and be absorbed by ARM so that the two companies can fulfill their vision? And nVidia will still be able to reap benefits from the whole thing either way?
This site is about making money. You do so by having clicks.
Charlie delivers clicks by trolling. A teqnique that many other tech sites use.
Charlie knows that what he writes is not exactly truth. It does not matter since he write for a AMD fan site.
The “hatred” style of writing makes both fans of AMD and Nvidia click on the article.
Charlie’s reporting on Bumbgate was good journalism.
Charlie’s reporting that Global Foundries will mass produce 28nm by early 2010 is not good journalism. It could be good journalism if he tried to find out why they are late instead of banning people on their board for trolling. (when the trolling is quoting Global Foundries own site 60000 wafers per month).
Why did Bulldozer miss clock speeds by 30%?
Why is GloFo so late?
Why do GloFo have so much problem? (when you ask exactly this question on the forum, you get banned for trolling)
If we were to believe Charlie, nVidia as a company should be an utter failure due to engineering incompetence and shitty management.
Yet, somehow, they still managed to make more profit in the last QUARTER than AMD did in the last DECADE.
Next, the expose on the lack of competence within the professional purchasing employ.
well who talk here money ??
we talk nvidia kick nvidia KO again ..
thay work 3 teams 1 doin what big leader say so
second what ceo say so
third what lead eng. say
and on end u got 3 ideas 3 ways and all time nvidia kick nvidia KO :)
if amd have that kind teams/work we will still wait amd64 to come out
with six’t variation of k6 core ,with name k12
that is point
and to guy up it have NOW 28nm 4core omap
and ms use it to make win 8 arm version not nv
ms and nv have baaaaaaddd blood ,so to nv to have team in ms is nuts .
qval. have in 1 or second q 2013 4core out in 28nm
and it prolly same time
sammy have date u say 6 month 2012
so stop propaganda and read what smart ppl say
and nv stock drop 7% beacose thay (dont)have good sale of tegra3 :)
not beacose charlie say here buhuhuhuu
nv sux
and nv get 1bil on year investments
amd 2bil from ny to build fab i think
in last 10 years
and yet amd have 3x more $ than nv
if nv say we have 300mil$ to share with investors ,u read we need more $$$$
whan amd say we have lost 50 mil$
u read we spend all in r&d
AMD save ATI LOL
ATI save AMD that is better
nv sux in gpu’s from nv4 ,nv5 flop 1bil $
nv6 oki but still not 1bil$ +
nv7 is nv6
nv8 is nv7
u see my point
Don’t say that! Here kids can’t count and can only understand simple words. It doesn’t mean if they are meaningless, they just like hearing anything that makes they believe AMD is better and has a bright future.
P.S: someone here still does not know that there is a version of the huge, hot, broken and unfixable Fermi design that works and kick anything AMD has been able to create so far. Moreover, this flawed design make nvidia gain money, AMD graphics division with their small, efficient philosophy is losing money (again) even though they sells millions of those chips. I wonder how many of them they need to sell in order to break even instead of subsiding their products to keep up with the market share.
Charlie did it again – NVDA down 7%:
http://blogs.barrons.com/techtraderdaily/2011/10/20/nvidia-off-7-semi-accurate-sees-delay-in-tegra-roadmap/
Charlie is the Puppet Master of Wall Street.
Seriously, dude – you have more power making crazy money with random predictions than any analyst I’ve ever seen. You should hook up with JP Morgan or Goldman Sachs and get a sweet deal going.
That would have nothing to do with Greece’s impending default, and the domino-effect to come, eh?
It works hard to catch the competitors. That’s what companies like Qualcomm and TI have done when Tegra 2 was released. They haven’t Cortex-A9 in their plans yet, nor a chip beyond 1 GHz and they came out with a competing product more than a year later. Did they go bankruptcy? No.
What you have not stated, dear Charlie, is that all competitor products to Tegra3 will come out when 28nm will be available for mass production, that’s in the best case second half of 2012 (TI OMAP was scheduled for 4Q 2012 not project Denver which was set for some thing vague as 2013-2014 together to 64-bit architecture which you have probably forgot). That leaves nvidia 9 months from now to get another SoC at 28nm after Tegra3.
Will be it a simple T3 shrink o a slightly revised one? Still have to see where competitor will release quad cores SoCs. And it will be a year from now when we’ll ever see any product adopt those SoC while Tegra3 is shipping now. With a 12-month advantage.
Remember that Win8 has as a reference design Tegra3, not a Krait or anything else. And we have still to see how much advantage a Krait will have over a simple A9 design. It may be that a 200MHz difference may fill the gap. And it is not said that the simple A9 architecture will consume more at higher clock.
Independently from what are you rants here, nvidia was the one that started the ARM SoC race to better performance with a tight developing cycle. Competitors have never shown before to have the capacities to keep up with fast developing rounds (and infact they have reacted with a 12-month delay). nvidia used this same method to throw off the market all old competitors in the graphics market (including ATI that was saved by AMD just in time).
They are doing the same with ARM SoC. First with dual core Cortex-A9, first with a quad core (and a interesting solution with the added low poer companion core that all others will copy in their next projects for sure), probably first with a Cortex-A15, and surely the only one that will be have a 64-bit version for servers and high end solutions.
They aer the only company with a long plan for the future and not simply stating when they will implement new ARM design as they are created by ARM themselves when the production process is good enough.
Who will survive this long race I do not know, but is is sure that someone won’t be able to keep up.
They need that low power companion core to get a tolerable standby time in phones or tablets.
Just wait for some battery life numbers when those 4 huge cores are under load…
Geez Palliot, you’re still saying some of the same stuff? nVidia didn’t have a 12 month lead with Tegra 2. The first Tegra 2 product was in August 2010, but the first ones worth buying didn’t hit until around October. Playbook hit in April. If you want to count AC100, an awful product in low volume that got quickly discontinued, you’d may as well count Pandaboard for OMAP4. Of course you should also count A5 as a dual-core A9 SoC.
What’s becoming clearer now with Tegra 3 is that nVidia has gained some of its time to market by using 40nm G cores with TSMC. These are likely ARM hard-macros instead of customized/optimized cores. The story continues with Tegra 3, hence why nVidia can only get 500MHz on an LP Cortex-A9 core while TI, Samsung, Apple, and Qualcomm can use LP and get much higher speed with customized designs.
So nVidia would be able to get SoCs out quicker due to not spending the design effort, but that puts them at a disadvantage in idle vs performance, hence what you get with Tegra 3. I doubt all the other companies will be rushing to copy nVidia’s approach if their Cortex-A9 cores are not substantially higher leakage than Tegra 3′s low-power core to begin with. Just lower the clock speed and voltage on one core (or more than one) and power gate everything else.
Tegra 2 took off in tablets but OMAP4 is getting more design wins in phones. Maybe that trend will repeat with Tegra 3, but it’s not like nVidia is able to reach its projected claims (tablets out in August.. really??), who knows if something like i.MX6 won’t show up in the mean time.. But this is all operating under the assumption that quad-core will actually be worth anything in the mobile space. In most cases you’ll probably prefer to have a higher clocked dual core, and OMAP4470 may come out around the same time too..
Also, forgot to mention.. Tegra 2 got a huge advantage on tablets due to being the Honeycomb reference platform, everyone agrees with this. They’re not going to have the reference for ICS so they lose that, even if new tablets don’t need the update as direly as they did with Honeycomb.
And just because we’ve seen Windows 8 on Tegra 3 doesn’t make it “the reference”, not least of all because we’ve seen it on x86 too.. but I really doubt Windows 8 is going to just be an immediate hit with tablets. The software ecosystem is just not there.
Tegra 2 was ready at the beginning of 2010.
They had to wait for a product able to exploit it. There was nothing on the market that was remotely able to use its powers.
In fact they needed Android to grow to version 3 and it was not a surprise that Tegra 2 was used as the reference design. As said all other competitors were more than 12 months late.
nvidia has so many advantage over the competitors that they can delay their plans of few months in order to suck as much as possible from the current Tegra 2 products. Beside the fact that the SW market is still behind, as Android 4 is still not available, and that is what makes differences for the user.
As things are now, nvidia has the capacity to create 3 different SoCs in the same time frame competitors can do 2. Of course different release date makes these products different and may give advantages to a competitor for a brief time (see OMAP that is now lightly faster than Tegra2 without really be an over killer, and that’s because nvidia scrapped 1.2GHz version Tegra2).
We’ve been through this a thousand times dude, it doesn’t matter when it was “ready.” Maybe you think Tegra 3 has been ready too and is just slipping nVidia’s projections because nothing can take advantage of it. And hey, you might have a point, since not an awful lot in the tablet space is going to put four cores to good use. Yet you think it needs Android 4.. why?
I see you ignore what I said about the competitors taking longer to do better cores at LP, forcing nVidia to do tricks like muxing an LP core with a GP to get good enough battery consumption at near-idle.
Unfortunately for nVidia that technique isn’t going to work with Cortex-A15 because of its tightly coupled L2 cache. nVidia will also have to adapt its GPU to use ARM’s coherency link with A15 instead of streaming through L2 cache… I guess it’s pretty fortunate that ARM is offering the big.LITTLE configuration, nVidia is probably going to have to take advantage of A7s at LP and A15s at GP where others will use both at LP.
Tell whatever you want, bur nvidia is revolutionizing the ARM market which has been in the back for such a long time.
I repeat, before Tegra there was not an ARM market as visible as it is now nor the chance to grew at x86 market expenses.
nvidia is leading the market with their production and design speed, while the competitors arrive on the market much later. With a better design? Maybe, but they are already a generation back. There’s not a request of quad core for smartphones? Maybe. Today. There was not a request for dual core + power GPU smarphone before Tegra2. Tegra was first.
What do you not understand is that nvidia is guiding the market where they want. With a powerful quad core phone many other device are possible (see ASUS padphone o Motorola Atrix). See a new generation of tablets like Asus Transformer 2 which with the righr price will makes all the other tablets on the market appear as toys. See new netbooks that can now compete with x86 based ones. See tons of other devices that can be built (HTPC anyone?) that are able to replace current obsolete and cumbersome technology. nvidia is guiding ARM technology exiting unknown market of embeded devices and this allows ARM to leave the power consumption limitations that have always been needed in that market. With noew and more powerful design for devices that are not necessary battery powered, extreme power efficiency is less important but performance are.
When concurrents will be able to make a quad core ARM, nvidia will already a step forward with A15 and a new GPU (they already have planned to change Tegra GPU technology with a unified shader architecture, so will not be a surprise that a new architecture that will couple A15 will be revealed). When the concurrents will arrive at A15 designs, nvidia will have quad core A15. And so on.
And this is only for the CPU part. For the GPU, PowerVR and such have to keep on nvidia capacity to develop faster GPUs every year. And that will not be so easy for them as architecture complexity increases and some thousands of millions of transistor will have to be handled. Remember that nvidia already designs the most complex integrated devices ever created on earth up to now, and the fact that companies like PowerVR and Imageon have been always limited to the embedded market makes clear what are their limitations in this field.
Also do not underestimate the capacity to create drivers needed for new class of applications and games (see the problems with Intel and PowerVR DX11 suport for the new Atom).
nvidia is not the best possible for sure. It is the fastest on market with more than decent solutions. The time needed for the competitors to reach nvidia performance levels, allows nvidia to be a step ahead. They have technology, design a support advantage in all sectors over the competitors which probably only have to offer a cheaper price. nvidia does not aim at selling cheap obsolete low value technology made at 65nm to have better market place. They develop for the top market and aim to gain enough for the next step (still for the top market) and thus leading the market where they want to.
You should have a look at what nvidia did when they arrived on the market with RIVA and what they did to technologically kill all the competitors that were present on the graphics market (were not GPU yet) at that time.
In short: performances, new features, support, first with everything to decide where the market had to go (GPU, they invented that, it was not ATI or anyone else). With fast developing cycles no other one could sustain (they made a new graphics design every 6 months). Even SGI gave up at the end (and they were the leader in performances for workstations and professional market). They are trying to do the same here. There are not many companies that can bring 2 billions of dollars on the table every year to design 3 generations of SoC at the same time. Mostly because many of that do not know where to go 3 years ahead. nvidia has a long term plan which goes beyond smartphones and tablets. None else has it. Those that cannot or those that will fail just a generation will die. No mercy. No pity. The rule of real competition and innovation.
I only read the last third of your last post, because frankly, I think you have outed yourself as a paid nVidia rep, but here are my thoughts.
First of all nVidia didn’t invent the GPU. Amiga did.
http://en.wikipedia.org/wiki/Original_Chip_Set
http://en.wikipedia.org/wiki/Commodore_Amiga
http://en.wikipedia.org/wiki/Graphics_processing_unit#History
Secondly: the last time nvidia to quote you “technologically kill all the competitors” was during the DX8 age with the GeForce 4 Ti, which they they bought from 3dfx, much the same way ATI bought the R300 (9500/9700) from artx. One could argue that nVidia was also winning when they had the 6XXX and 7XXX series, but never by as much as the GF4 was winning.
thirdly: nVidia misses it’s time to market as often as everyone else, this is primarily because engineers are optimists that believe highly in their skills. They believe all problems will be solved quickly and easily.
Fourthly: Saying such and such a tablet will make all other look like toys is like saying that such and such a car is superior because bosche made the ECU. There is SOOOO much more to a tablet than the processor. I’m sure that you can build a great tablet with an nvidia cpu, just like you can build a crappy tablet with an nVidia CPU.
I reply to myself to avoid the column to become too narrow.
Amiga’s GPU, which I know quite well having worked with it, is not the same ad the first GeForce nvidia proposed. GeForce 1 was a revolution on the graphics market, something the competition take a bit of time to understand and copy. In fact it was a killer for many competitors that cannot go over a certain design complexity. They died.
nvidia has been leading the market since the introduction of the first GeForce. Yes, not all their products were perfect or astonishing, but they never have never lost the leadership of the graphics development. With leadership I do not mind performance only, but the direction of the development. In fact, they have been leading the professional market for quite a bit of time now due to the capacity to be ahead of the competition. And as it is now, competition just come late with the same feature set, just a bit faster.
In consumer market customer look at the price more than technology, and that was what has been killing competitors: they could not produce as good chips as nvidia at the same price. No, ATI couldn’t as well, as they were almost on bankruptcy when AMD bought them. So far so good for a company that many thought to be on par with nvidia at technological level.
nvidia didn’t build their GF4 on 3DFx architectures. They bought SLI technology from them, and then in fact they were the first to propose dual GPU solutions. 3DFx has nothing to do with GF4 being good or not. GF4 was an evolutionary development of the GF3, 2, and 1.
nvidia has missed its time to market many times for sure. But if you look at their products, when on market they have always been quite ahead of the competition. And there’s not only performance in competition, but support too.
Unfortunately there isn’t much more in a tablet than 3 factors: OS, display quality and the performances. iPad apart, you can see than all the other competitors have been offering almost the same things with no one being able to make a real difference with respect to the others.
In fact they were all based on the same chipset or feature set: dual core, fast enough GPU and displays that are not much different one from another and quite distant from iPad quality. Despite the initial launch prize, that the end they all came to the same low common price = not enough differentiation.
You can’t deny that a quad core Tablet with a powerful GPU, a new OS and hopefully a good display will be a different thing than whatever is on the market today. And will be a different thing till someone will introduce a different quad core SoC or new features (Lightpeak? Satellite connection? Mind reading?). Until then, only nvidia will be the choice to make the difference and allow producers to sell innovative products and not only “obsolete & recycled” technology just a bit faster than the predecessors.
Make a Mercedes class car and sell it at Volkswagen class C prices, and you have a killer product, even though competitors can make slight faster than bicycle can at Skoda prices. ANd do not forget about netbooks and other similar devices that can be bought only with a fast enough CPU.
Try with understanding the concepts of sufficient and needed: to make a good device using nvidia new SoC is not sufficient, to make a top tier product, you need it.
Once a tablet with Tegra 3 is on the market, no other competitor can claim to have the fastest, best, ultra innovative whatever marketing words you can come up with as that would not simply be true and clearly that won’t to anyone that compares the two devices, one with whatever good new dual core and one with nvidia new quad core.
And do not forget Win8 support. It will take more than a year to competitors to have a sufficiently fast enough products able to cope with it.
One out of the super mobile market, nvidia shows all its advantages.
I am not paid to say these things (unfortunately). They are just a description of how the situation is now and how it will progress based on what nvidia did in the past.
You may like it or not, but that’s surely shows a development plan that gos beyond the useless rants Charlie writes here to no one benefit on a 2-month time slip vs a multi year investment and market forging strategy.
All you JHH lovers better get checked for STDs. You can catch them by giving him blowjobs, ya know.
well things go like this ..
nvidia after tegra 2 flop ,and loosing over 100mil $ from r&d
decide ,and show to press tegra 3 28nm part 4core..
but thay cant make 28nm (and others can ,that’s wierd) so thay make 40nm tegra 3 .
now tegra 4 will be so called pre arm64bit
soc to combine new g3/4+ ic and arm soc in one chip but in 28nm this time..
and denver so called tegra 5 will be true arm 64bit soc hi-perf. server kind monster
with gpgpu parts in it and 28nm leading to 22nm and thay say it will be and of 2012 to
go with win 8 server variation for arm cpu’s
but thay got hot batery eater in tegra3
and apple ti qval. will go with 28nm 4core soon (1-2q 2012) and tegra 3 will eat another flop ala terga 1/2 ,and that mean problems for nv and arm division.
so thay did what thay know best (8800->9800->220->etc) thay shrink tegra 3 in 28nm and called her new generation of tegra ..
so in same time arm true 64bit soc have time now to work up and glue/combine with
gpgpu segments (ala amd apu) and make true server cpu to work with tesla cards.
thay started with true 8core arm 64bit cpu
with some smart new gen (kepler??)gpgpu segments ,but we all know how nvidia work
one team do one thing other something else
so how i got info ,it’s now 16/10 core
(16 arm 12 gpgpu)
and thay will not finish this till 1-2q 2013 ,part will be 22nm (that was inline with TSMC 22nm path)
so charlie is right tegra 4 will be only 28nm tegra 3 ,and is good soc will be less hot/power hungry than tegra3
and we will have faster pad’s (thay say 20% to up 50% faster)
and tegra 4 is now tegra 5
started like switch with arm 32bit and arm 64bit soc on same chip with integrated g2,g3,g4 net segment thay buy this year,
to god knows what will be on end .
but prolly 22nm tegra4x2 if thay are smart
better to put all smart lady’s to work on super chip denver and leave tegra5 to be tegra 4 on 22nm with 2 T40 cores ala intel/amd first 2 core cpu’s glued together..
my 0.02$
Oh crap, Drashek is this your offspring or something?
BTW, on the public finance meeting, Nvidia only gave a ball park schedule of Denver. No one promised it’s shipped on Q4/2012.
Hi, Charlie, the audio record of finance meeting is online, everybody could hear it. Or maybe you should check it first.
I don’t know why Charlie keeps saying “Denver is on the Tegra roadmap”, “Denver is T50″, “Denver will be Tegra 6″, “The original Denver fails, then Nvidia has to change their plan, blablabla….”
From my solid source, Denver is designed to a high performance 64bit arch ARM chip off the Tegra roadmap from the very start point.
Sometimes, it’s fun to see a person hates a company in such a twisted way.
I don’t think what you just said contradicts what he said.
charlie this is plasticity
tranee’ gate to Jenson Gate, Name of Revelations will soon be in Book of Tom. Gone With Calling Wind. From Tuna to Hard PCB, Charlie Herd dem All.
WHY DO PEOPLE WANT COSTLY STUFF, FOR FIRST PERSON SHOOTER, ARE WARPED aNDREW PROBLEM. WHY NOT SOMETHING SIMPLE:gITmO, wITH lOVE.
better yet, nvidia card with razors, to cut up fingers dring insertion. first person blood. nice RED Effect.
drashek typeO gates….
Pure ranting… even for Charlie…
So it’s not true? Where’s your tech-news/blog site so I can read the truth?
Show one piece of evidence, anywhere, that supports Charlie’s ranting…
Wow, you’re playing that card, are you? OK, I will play along…
So… show evidence to refute Charlie’s claims, or STFU and GTFO.
DESPERATE !!!!
LMAO !!!
yeah I’ll bet you are literally “laughing (your) ass off.” That’s as believable as anything I read around here
“you are literally “laughing (your) ass off.”
correct
“That’s as believable as anything I read around here”
correct
Bob, you’re a tool, a complete and total knee-biter. Charlie has people within nVidia who talk to him. Who talks to you? The voices in your head?
“Bob, you’re a tool, a complete and total knee-biter.”
Probably so…
“Charlie has people within nVidia who talk to him. ”
Sure, according to Charlie, one wonders how NVidia ever makes a dime…
I wonder if the only reason Charlie can put theses awesome rants about Nvidia is just because he ended having better moles in that company, as opposed to AMD which the bulldozer rant only came after the release, even though it seems management knew about the problems beforehand, before Charlie ever being put under NDA.
Kinda weird.
NO… NVidia has a very arrogant marketing dept that planted the seeds for Charlie’s anit-NVidia rants dating back over a decade…
You should know Charlie has a mental illness and is not allowed near sane people like mister Huang or his engineering team
Charlie lie about Tegra before and look, Tegra now is a big success, now he is lying about Tegra again, what an idiot
BTW you can go and buy yourself a nice galaxy smartphone with Tegra II but what about AMD? can you buy a nice AMD smartphone? of course not AMD only make cheap X86 clone with old technology
Oh, Guy Wong, your consistency is refreshing, Like a sunrise, we can always count on you to remind us that:
everything charlie writes is wrong, he’s a bad person. reputable companies like Intel and nVidia never lie and are always working in the best interest of me, the consumer.
thank you for not letting trivial things, like facts, influence you.
Charlie is never under NDA from nVIdia, what info he gets he can print whenever and however he wants. The relationship with AMD is more cordial, some info is obtained under NDA, some info he may prefer to delay as to not damage that relationship.
Wow, this is pathetic.
I really don’t like the writing of this article. Using Dear Leader obviously want you to think of the North Korean leader and it’s frankly a disgusting comparison.
No, it isn’t. It is an accurate depiction of life under the illogical, nonsensical, micromanagement style of JHH and his drooling minions.
You could just as well be talking about Steve Jobs, he didn’t make sense to his minions but he was correct nevertheless.
No making sense and not making sense are sometime two completely different things.
Especially if one is about not understanding the strategy and other about constantly changing it …
omg sooo true… Jhh is batshit crazy
I it’s not what you make, it’s how you present it :) Can AMD finally learn it from nVIDIA an iNTEL?
That is what you do charlie when AMD screws up , you rush to your favorite rant topic , Nvidia , ironically you ignore the incompetence of AMD engineers that failed them twice in one year , once with the mighty HD 6000 series that failed to beat GTX 500 , and the second time with the disaster that is called Bulldozer !
Charlie go hide in a corner and cry !
umm did you miss the article in which he faults AMD entirely? or did you not learn to read past titles of articles little boy?
Charlie still whines like the overweight bitch he is whenever Nvidia does anything, its the good with the bad of this site. I actually enjoy reading the rants he posts and the impending backlash from the comments.
So, in a nutshell, red camp is taking a beating right now, green camp is in for some pain in the mid-term, and blue camp is smirking while quietly sweep up a few little mistakes as they try to miraculously transform a CPU into a GPU. An ocean of ARM companies, meanwhile, are eating the 3 giants from the bottom up like piranha fish. And like him or not, we lost the man who pushed the industry to where it is now.
Seems like a pretty depressing time in the tech world. How about an article on SI or Kepler, to cheer us up? Preferably one that doesn’t involve XDR or the (in)competence of fabs.
whoops didnt mean to reply
“And like him or not, we lost the man who pushed the industry to where it is now.”
Sadly, this does not seem to be meant as a joke.
I have never seen a bigger tool on an internet blog….
So, Frank, are you saying that the Nvidia Tegra schedule did not just slip a year? Or that you just don’t like the way it was presented?
Shifting from core A9 to A15 should be a good move, but the wait for it is going to be a killer.
I’m Frank. He’s frank.
He’s right, tool or not.
And … that is why it hurts.
What did nvda do to you?
You should wonder why you let nVidia keep doing things to you instead.
LOL!!! indeed…. like the 8800 turned into a 9800 turned into a gt250….
nvidia is the master of the shell game and portfolio lies…
nvda shattered dreams.
He wrote several articles on what they did, frank.
Do you have financial ties to Nvidia?
Or maybe you’re whining because you have first-hand knowledge that what Charlie’s saying can’t possibly be true?
Or, is Nvidia doing what every other company is doing and hiring PR firms to troll the internet on their behalf?