Yesterday we all saw the blog piece from NVIDIA that stated that they were going to start licensing their IP to interested third parties. Obviously, there was a lot of discussion about this particular move. Some were in favor, some were opposed, and others yet thought that NVIDIA is now simply roadkill. I believe that it is an interesting move, but we are not yet sure of the exact details or the repercussions of such a decision on NVIDIA’s part.
The biggest bombshell of the entire post was that NVIDIA would be licensing out their latest architecture to interested clients. The Kepler architecture powers the very latest GTX 700 series of cards and at the top end it is considered one of the fastest and most efficient architectures out there. Seemingly, there is a price for this though. Time to dig a little deeper.
Kepler will be the first technology licensed to third party manufacturers. We will not see full GPUs, these will only be integrated into mobile products.
The very latest Tegra parts from NVIDIA do not feature the Kepler architecture for the graphics portion. Instead, the units featured in Tegra can almost be described as GeForce 7000 series parts. The computational units are split between pixel shaders and vertex shaders. They support a maximum compatibility of D3D 9_3 and OpenGL ES 2.0. This is a far cry from a unified shader architecture and support for the latest D3D 11 and OpenGL ES 3.0 specifications. Other mobile units feature the latest Mali and Adreno series of graphics units which are unified and support DX11 and OpenGL ES 3.0.
So why exactly does the latest Tegras not share the Kepler architecture? Hard to say. It could be a variety of factors that include time to market, available engineering teams, and simulations which could dictate if power and performance can be better served by a less complex unit. Kepler is not simple. A Kepler unit that occupies the same die space could potentially consume more power with any given workload, or conversely it could perform poorly given the same power envelope.
We can look at the desktop side of this argument for some kind of proof. At the top end Kepler is a champ. The GTX 680/770 has outstanding performance and consumes far less power than the competition from AMD. When we move down a notch and see the GTX 660 Ti/HD 7800 series of cards, we see much greater parity in performance and power consumptions. Going to the HD 7790 as compared to the 650 Ti Boost, we see the Boost part have slightly better performance but consumes significantly more power. Then we move down to the 650 and 650 Ti and these parts do not consume any more power than the competing AMD parts, but they also perform much more poorly. I know these are some pretty hefty generalizations and the engineers at NVIDIA could very effectively port Kepler over to mobile applications without significant performance or power penalties. But so far, we have not seen this work.
Power, performance, and die area aside there is also another issue to factor in. NVIDIA just announced that they are doing this. We have no idea how long this effort has been going, but it is very likely that it has only been worked on for the past six months. In that time NVIDIA needs to hammer out how they are going to license the technology, how much manpower they must provide licensees to get those parts up and running, and what kind of fees they are going to charge. There is a lot of work going on there and this is not a simple undertaking.
So let us assume that some three months ago an interested partner such as Rockchip or Samsung comes knocking to NVIDIA’s door. They work out the licensing agreements and this takes several months. Then we start to see the transfer of technology between the companies. Obviously Samsung and Rockchip are not going to apply this graphics architecture to currently shipping products, but will instead bundle it in with a next generation ARM based design. These designs are not spun out overnight. For example, the 64 bit ARMv8 designs have been finalized for around a year, and we do not expect to see initial parts being shipped until late 1H 2014. So any partner that decides to utilize NVIDIA’s Kepler architecture for such an application will not see this part be released until 1H 2015 at the very earliest.
Sheild is still based on a GPU posessing separate pixel and vertex shaders. DX11 and OpenGL ES 3.0? Nope!
If someone decides to license this technology from NVIDIA, it will not be of great concern. The next generation of NVIDIA graphics will already be out by that time, and we could very well be approaching the next iteration for the desktop side. NVIDIA plans on releasing a Kepler based mobile unit in 2014 (Logan), which would be a full year in advance of any competing product. In 2015 NVIDIA is planning on releasing an ARM product based on the Denver CPU and Maxwell GPU. So we can easily see that NVIDIA will only be licensing out an older generation product so it will not face direct competition when it comes to GPUs. NVIDIA obviously is hoping that their GPU tech will still be a step ahead of that of ARM (Mali), Qualcomm (Adreno), and Imagination Technologies (PowerVR).
This is an easy and relatively painfree way to test the waters that ARM, Imagination Technologies, and AMD are already treading. ARM only licenses IP and have shown the world that it can not only succeed at it, but thrive. Imagination Tech used to produce their own chips much like NVIDIA does, but they changed direction and continue to be profitable. AMD recently opened up about their semi-custom design group that will design specific products for customers and then license those designs out. I do not think this is a desperation move by NVIDIA, but it certainly is one that probably is a little late in coming. The mobile market is exploding, and we are approaching a time where nearly every electricity based item will have some kind of logic included in it, billions of chips a year will be sold. NVIDIA obviously wants a piece of that market. Even a small piece of “billions” is going to be significant to the bottom line.
Give me an APU type chip with
Give me an APU type chip with at least 24 ARM 64 bit CPUs and as many Nivida graphics SMs as possible, this could compete with the Xeon pi, as a poor man’s ray tracing and rendering farm on a chip! Giving it good OpenCL, OpenGL support is also a selling point, while also providing CUDA,
but a CUDA only Platform will not sell well, as the trend is towards more open API standards! Provide better support for Linux(Android,Chrome OS), as you will need to get your product in as many chromebooks as possible! Better get a good APU designed, beacuse AMD is going to be making APUs around ARM, so give Your IP partners the same support That ARM gives to their IP partners! I would love to have an ARM based laptop, sans the WINTEL, and APUs will rule!
actually this is interesting.
actually this is interesting. say so if other ARM licensee using GeForce IP for their graphic portion then it is possible for other ARM cpu to get Tegra like graphical enhancement?
Even though they’ll never
Even though they’ll never admit it, I can’t help but read in a little regret related to this upcoming console cycle.
There are lots of things to
There are lots of things to consider beyond the licensing of GPU parts to other chip vendors.
Kelper was made on TSMC’s 28nm HP (High Performance node), if a interested 3rd party try to move those to a 28nm HPM (High Performance Mobile) or future 20nm, 16nm FinFET, they will find out that those parts will not yield. Or not yield as good as they like to have.
Moving a gpu from one process to another require quite a bit of work. Not to mention also trying to combining a ARM core to it.
That’s why AMD use semi custom design, where they know which process to use and how it will affect yields. This will greatly mitigate the risk of low yields for clients such as Sony, MS. It’s a bit easier to have their client order what they want, and then make those parts for them.
Similar to ordering sushi from a professional Sushi Shop(AMD), and making your own sushi(NVDA).
isn’t that nvidia already
isn’t that nvidia already integrating kepler architecture for their next year T5? that’s mean it is possible to move the tech into mobile. if not nvidia wouldn’t be dare to say kepler will be inside their T5. it is not about making current kepler to fit into mobile device. it is to integrate kepler tech into mobile.
Problem with Tegra 5 is they
Problem with Tegra 5 is they have no solid launch date. With the first Tegra 4i design phone set to emerge around Q1 of 2014. The schedule on the nVidia road maps are in serious risk of slipping further by unforeseen delays. NVDA will have to make some sort of announcement regarding Tegra 5 on one of their upcoming quarterly reports(IMHO likely Q4).
Adding to NVDA’s pain, numerous phone makes: LG have announce the use of Snapdragon 800 in their upcoming phones. Others, Samsung, Sony(Xperia ZU), HTC, Asus will very likely follow suit. This segment is where Tegra 4i should be competing, but delays almost guaranty no design wins.
The mobile segment is dominated by GIANTS!
Tegra 4i will not competing
Tegra 4i will not competing with Snapdragon 800. the best maybe nvidia will aim Snapdragon 600 with it. about T5 maybe nvidia will wait for next die shrink. but if they want to bring it faster maybe they might pull similar move they did with T3.
Just my take.
NVidia is a
Just my take.
NVidia is a little too late to do this and it feels really desperate. By the time the tech is integrated into chips, there should be a shift in the process tech to 20nm and this would make the existing 28nm GPU tech essentially useless/outdated. You have already alluded to it in the article and it feels like the company is having too many issues with basic planning. They have been shooting themselves in the foot for too long and all this will get them into more trouble at least in the short/medium term.
What I’d like to see is the company sitting back and making sure that the 20nm process technology gets into Tegra before the Kepler stuff and make a concentrated effort at making the Tegra platform the go-to chipset for the high performance mobile devices. If it takes 2 years then so be it. Why try to show up with hobbled parts that are embarrassing compared to what is available from other vendors. It is not like the smart-device era will disappear in the next couple of years. If they are able to develop a good GPU that is able to outperform the other IGPs (at the cost of some extra power consumption) on the other ARM designs, then integrate them into newer versions of the Shield, I can certainly see NVidia coming back stronger than ever. It could create the necessary tools that would enable independent developers to produce high quality games for the SoC.
But I don’t think the company has the time or the money to sustain themselves for this long a duration without putting out any products (no matter how undesirable). AMD is in a much better position today and with the introduction of Kaveri, it will only go from strength to strength after this year. I am being extremely negative on the outlook for this company but I am making these comments only after looking at how its ARM products have been welcomed. I don’t want it to disappear and it will not any time soon as it has pretty much locked up the dGPU business on the PC/Laptop platform.
Definitely desperation. Follow the data points.
1. Laura Croft leaves Nvidia hanging until release.
2. EA decides to work primarily with AMD to optimize for AMD hardware.
3. AMD’s recently announced United Gaming Strategy.
4. Kaveri releasing concurrent with PS4 and Xbox One.
Kaveri was developed on the same timeline as the PS4 and Xbox One APUs. Perfect opportunity for AMD to tweak and optimize Kaveri for maximum compatibility with those console chips, thus making it very easy for game developers to port to and optimize for Kaveri.
I would expect the EA optimization agreement is the tip of the iceberg and AMD is working with all the major developers and with all the next gen game engines to streamline that optimization.
It’s obvious the developers know what’s up – AMD is holding a handful of aces and is providing all the bottom line advantages and it’s to everyone involved interests (except Nvidia’s) to focus on AMD in general, they are, after all, already deeply involved with AMD engineering teams on the consoles, and their upcoming full HSA APUs and GPUs in particular, which almost perfectly dovetail with the console APUs (and coding) to provide a FAR bigger gaming experience ‘bang for the buck’ than is possible with an Intel/Nvidia CPU/GPU solution. That translates into substantial more game sales to the low and mid level PC market where the majority of machines are sold.
Everyone wins here except Nvidia and gamers who stick with Intel/Nvidia solution for next gen gaming.
I suspect AMD wants a handful of the hottest AAA next gen games highly optimized for Kaveri and the 8xxx GPUs which they will make sure are in the hands of reviewers when Kaveri is released with benchmarks that will destroy Intel/Nvidia solutions anywhere near the price range.
That will propel even more Kaveri optimizations amoung more developers who want on that bandwagon. And why not if if costs practically nothing to do so?
Keep in mind Kaveri graphics will be ADDITIVE to 8xxx graphics. When purchasing a Kaveri, which WILL be the absolute hottest gaming must have when it releases, an Nvidia AIB will be at a MASSIVE disadvantage vs. an AMD AIB. All that delicious HSA optimization and additive graphics power will *puff* disappear into thin air.
THAT IS WHAT NVIDIA IS LOOKING AT. Intel won’t be happy either, but it’s a pretty minor market segment for them.
Add to that the near certainty the professional market will be embracing HSA (and AMD HSA APUs) with open arms and Nvidia’s problems compound.
Nvidia is reacting to FUTURE sales projections where they see first their consumer GPU AIB market start taking a massive hit by the end of this year when Kaveri and the 8xxx GPUs release and over the next few years start seeing their professional GPU market follow suit.
That is why their newly announced licensing strategy is a desperation move. What other future strategy options do they have.
One more item – I see releasing Shield as a confirmation Valve decided in the end to go with an AMD solution for Steam Box. Really, considering the above, what choice do they have? The only way Shield made sense was a part of an Nvidia solution to Gabe’s stated desire for the Steam Box being able to ‘stream several games simultaneously’. Nvidia’s last desperate attempt to keep SOME slightest foothold in x86 gaming. When Valve ended up going with AMD Project Shield was no longer needed – so Nvidia in another desperate attempt to remain relevant in any gaming market anywhere (Tegra 4 is looking to be DOA) is trying to leverage it as an Android device and PC with ‘certain Nvidia GPUs) streaming devices.
If you hold Nvidia stock, now would be the time to sell it and invest in AMD.