Q1 2013 discrete GPU market share?

[SirPauly]

Always enjoyed this question in 2009 and the CEO was nice enough to answer for the community:

Q: With AMD's acquisition of ATI and Intel becoming more involved in graphics, what will NVIDIA do to remain competitive in the years to come?

The central question is whether computer graphics is maturing or entering a period of rapid innovation. If you believe computer graphics is maturing, then slowing investment and integration is the right strategy. But if you believe graphics can still experience revolutionary advancement, then innovation and specialization is the best strategy.



We believe we are in the midst of a giant leap in computer graphics, and that the GPU will revolutionize computing by making parallel computing mainstream. This is the time to innovate, not integrate.



The last discontinuity in our field occurred eight years ago with the introduction of programmable shading and led to the transformation of the GPU from a fixed-pipeline ASIC to a programmable processor. This required GPU design methodology to include the best of general-purpose processors and special-purpose accelerators. Graphics drivers added the complexity of shader compilers for Cg, HLSL, and GLSL shading languages.



We are now in the midst of a major discontinuity that started three years ago with the introduction of CUDA. We call this the era of GPU computing. We will advance graphics beyond programmable shading to add even more artistic flexibility and ever more power to simulate photo-realistic worlds. Combining highly specialize graphics pipelines, programmable shading, and GPU computing, computational graphics will make possible stunning new looks with ray tracing, global illumination, and other computational techniques that look incredible. Computational graphics requires the GPU to have two personalities, one that is highly specialized for graphics, and the other a completely general purpose parallel processor with massive computational power.



While the parallel processing architecture can simulate light rays and photons, it is also great at physics simulation. Our vision is to enable games that can simulate the interaction between game characters and the physical world, and then render the images with film-like realism. This is surely in the future since films like Harry Potter and Transformers already use GPUs to simulate many of the special effects. Games will once again be surprising and magical, in a way that is simply not possible with pre-canned art.



To enable game developers to create the next generation of amazing games, we've created compilers for CUDA, OpenCL, and DirectCompute so that developers can choose any GPU computing approach. We've created a tool platform called Nexus, which integrates into Visual Studio and is the world's first unified programming environment for a heterogeneous computing architecture with the CPU and GPU in a eco-processing configuration. And we've encapsulated our algorithm expertise into engines, such as the Optix ray-tracing engine and the PhysX physics engine, so that developers can easily integrate these capabilities into their applications. And finally, we have a team of 300 world class graphics and parallel computing experts in our Content Technology whose passion is to inspire and collaborate with developers to make their games and applications better.



Some have argued that diversifying from visual computing is a growth strategy. I happen to believe that focusing on the right thing is the best growth strategy.



NVIDIA's growth strategy is simple and singular: be the absolute best in the world in visual computing and to expand the reach of GPUs to transform our computing experience. We believe that the GPU will be incorporated into all kinds of computing platforms beyond PCs. By focusing our significant R&D budget to advance visual computing, we are creating breakthrough solutions to address some of the most important challenges in computing today. We build Geforce for gamers and enthusiasts; Quadro for digital designers and artists; Tesla for researchers and engineers needing supercomputing performance; and Tegra for mobile user who want a great computing experience anywhere. A simple view of our business is that we build Geforce for PCs, Quadro for workstations, Tesla for servers and cloud computing, and Tegra for mobile devices. Each of these target different users, and thus each require a very different solution, but all are visual computing focused.



For all of the gamers, there should be no doubt: You can count on the thousands of visual computing engineers at NVIDIA to create the absolute graphics technology for you. Because of their passion, focus, and craftsmanship, the NVIDIA GPU will be state-of-the-art and exquisitely engineered. And you should be delighted to know that the GPU, a technology that was created for you, is also able to help discover new sources of clean energy and help detect cancer early, or to just make your computer interaction lively. It surely gives me great joy to know what started out as the essential gear of gamers for universal domination is now off to really save the world.



Keep in touch.



Jensen

The key is how off or accurate his vision is from 2009 to today?

 

[raghu78]

The key is how off or accurate his vision is from 2009 to today?

Nvidia's biggest challenges come over the next 5 years with dGPU market shrinking into more of a niche market. The biggest volume segments which are entry to mid level will get hit the most. Intel muscled Nvidia out of the chipset market and the next casualty is the entry-mid level dGPU.

Intel and AMD are going to provide the GPU performance of the USD 100 GPU making it redundant. Kaveri brings HD 7750 - HD 7770 performance and GDDR5 memory support essentially eliminating need for a entry level graphics card on AMD notebooks/desktops. Haswell Crystalwell (Iris Pro with 128 MB EDRAM) on 22nm and Broadwell on 14nm are clearly looking to take more of the bill of materials of the laptops by eliminating the entry level dGPU which might be Nvidia or AMD based. Furthermore each and every generation Intel / AMD will devote more transistors towards the iGPU. the iGPU is going to get a disproportionate amount of transistor increase at every node.

Nvidia's other biggest challenge is from the fact that the ARM market is increasingly being dominated by Qualcomm, Apple and Samsung. With Intel Baytrail looking to take the fight to the ARM market, the ARM market will be under even more pressure. Long term Intel, Qualcomm, Apple and Samsung might survive. The latter two because they have a duopoly of the smartphone market with the rest being marginal players.

In the next 5 - 7 years we will either see Nvidia overcome these challenges admirably or eventually be gobbled by Intel.

http://seekingalpha.com/article/1423521-nvidia-stumbles-on-the-next-generation-tegra

"Given the trends for both its GPU and Tegra segments, I believe Nvidia badly needs a partner. Probably, the most logical company to purchase Nvidia would be Intel, just as AMD bought ATI. Then Intel could begin a process similar to AMD's of incorporating Nvidia GPU designs into Intel processors. Nvidia's GPU ventures into servers are interesting, but they would be all the more compelling in the form of Intel/Nvidia SOCs.
Nvidia's GPUs would also benefit from Intel's excess foundry capacity. In addition to packaging Nvidia GPUs inside future (post-Haswell) processors, Intel could also offer motherboards with discrete Nvidia GPUs included. I don't look for this in the near future. Probably Nvidia will have to suffer, and its investors along with it, a good deal more before relief comes in the form of a suitor."

 

[blastingcap]

I do not own any NVDA stock anymore due to factors that raghu78 talked about, but also because I think we're in for a general downturn any moment now, so I also got rid of most of the rest of my stocks too, including INTC. Intel has a long road ahead of it to try to catch up in smartphones, and it's at a major disadvantage. Even if it somehow manages to get a big share of the market, margins are likely to be poor relative to what they get from desktop and server parts.

 

[Imouto]

ASDF ASDF.

Delete this plox.

 

[ams23]

Nvidia's biggest challenges come over the next 5 years with dGPU market shrinking into more of a niche market. The biggest volume segments which are entry to mid level will get hit the most. Intel muscled Nvidia out of the chipset market and the next casualty is the entry-mid level dGPU.

The so-called "dGPU" is evolving just as the "iGPU" is evolving. At some point in the near future, CPU cores will be integrated on GPU die. So integration is happening everywhere. The important thing for NVIDIA is that the GPU not be commoditized and that the GPU continues to be more and more relevant to the future of computing. Anyway, improved Intel integrated graphics may cannibalize low end NVIDIA discrete graphics, but on the flip side, NVIDIA's Tegra integrated processors will be a disruptive force to Intel (and AMD) integrated graphics processors in markets where NVIDIA is not a player today. This disruption is happening as we speak (with upcoming products such as HP Slatebook x2), and will become more forceful as Android and Windows on ARM ecosystems continue to grow and evolve over time.

 

[notty22]

Here is some data :
JPR: Nvidia GPU shipments are up despite turbulent PC market

Intel’s GPU shipments went down 5.3% quarter to quarter, largely due to a decline on the overall PC market where the company’s integrated offerings are quite popular. AMD’s were mostly flat, with a 0.3% drop during this period, while Nvidia bucked the trend going up 3.6%. On a year to year basis things looked a little grimmer for AMD, which was down a full 29.4%, with Intel down 8.8% and Nvidia up once again 3.6%.


Other findings published on JPR’s report include the following:

• AMD’s quarter-to-quarter total shipments of desktop heterogeneous GPU/CPUs, i.e., APUs jumped 30% from Q4 and declined 7.3% in notebooks. The company’s overall PC graphics shipments slipped 0.3%.
• Intel’s quarter-to-quarter desktop processor-graphics EPG shipments decreased from last quarter by 3%, and Notebooks fell by 6.3%. The company’s overall PC graphics shipments dropped 5.3%.
• Nvidia’s quarter-to-quarter desktop discrete shipments were flat from last quarter; and, the company’s mobile discrete shipments increased 7.6%. The company’s overall PC graphics shipments increase 3.6%.
• Total discrete GPUs (desktop and notebook) were up 1.1% from the last quarter and were down 11% from last year for the same quarter due to the same problems plaguing the overall PC industry. Overall the trend for discrete GPUs is up with a CAGR to 2016 of 2.6%.
• Ninety nine percent of Intel’s non-server processors have graphics, and over 67% of AMD’s non-server processors contain integrated graphics; AMD still ships IGPs

 

[raghu78]

The so-called "dGPU" is evolving just as the "iGPU" is evolving. At some point in the near future, CPU cores will be integrated on GPU die. So integration is happening everywhere. The important thing for NVIDIA is that the GPU not be commoditized and that the GPU continues to be more and more relevant to the future of computing. Anyway, improved Intel integrated graphics may cannibalize low end NVIDIA discrete graphics, but on the flip side, NVIDIA's Tegra integrated processors will be a disruptive force to Intel (and AMD) integrated graphics processors in markets where NVIDIA is not a player today. This disruption is happening as we speak (with upcoming products such as HP Slatebook x2), and will become more forceful as Android and Windows on ARM ecosystems continue to grow and evolve over time.

Nvidia's tegra business is finding the going tough with revenue down 50% QoQ and 22% YoY. When a supposedly growing division shows lower revenue than the same quarter a year ago those are not good signs. Qualcomm has replaced Nvidia from the next gen Google Nexus and Surface RT. Intel with 22nm Baytrail and a completely new out of order architecture is going to make things even more difficult for the ARM crowd as 32 nm clovertrail is already faster than arm a15.

the entry level dGPU upto USD 100 will be eaten up by the iGPU. the reasons are simple. Both Intel and AMD are going to be able to allocate a significant portion (say 100 sq mm) die size for iGPUs. Intel has a process node lead so they will make it even more difficult for entry level Nvidia/AMD dGPU with their high end Iris Pro iGPUs. Intel's mobile efforts is picking full speed and in 2 - 3 years they will take significant market share. the smaller players will eventually quit - Nvidia, Freescale, Marvell. the bigger players will survive - Qualcomm, Apple, Samsung. In fact Intel might even get Apple to switch to Atom by the time 14nm FINFET products arrive in late 2014 / early 2015.

 

[ams23]

Nvidia's tegra business is finding the going tough with revenue down 50% QoQ and 22% YoY.

Next time, try to understand the numbers before blindly quoting them. The dip in overall Tegra revenue in Q1 2014 (fiscal date) relative to Q4 2013 (fiscal date) [ie. QoQ] is because T3 production is quickly ramping down in anticipation of T4 production ramping up. Since T4 was delayed by at least three months, T4 mass production will not start until Q2-Q3 2014 (fiscal date), so that is why Q1 2014 (fiscal date) revenue is down relative to Q1 2013 (fiscal date) [ie. YoY]. The same trend will continue next fiscal quarter too. The important thing is that, even with a three month delay to T4, overall Tegra revenue for fiscal year 2014 is projected to be in line with overall Tegra revenue for fiscal year 2013, which means that Q3-Q4 2014 (fiscal date) will show strong growth relative to prior quarters this fiscal year. Note that the large impact on overall Tegra revenue involved with ramping down prior gen Tegra will be less severe in the future with the introduction of new products such as T4i.

Intel with 22nm Baytrail and a completely new out of order architecture is going to make things even more difficult for the ARM crowd as 32 nm clovertrail is already faster than arm a15.

This is incorrect. [Dual-core] Clovertrail SoC's are most certainly not faster than [Quad-core] A15 SoC's such as T4. As for Baytrail, it will not be available until end of 2013. T5 "Logan" will be sampling by then, and will be in production early 2014.

the entry level dGPU upto USD 100 will be eaten up by the iGPU.

Obviously improved integrated graphics can cannibalize low end discrete GPU's (as I said earlier), but that impact will be more than offset by Tegra moving up the food chain in areas where NVIDIA is not a player today. Within one year's time, T5 is expected to be available with ~ 400 GFLOPS GPU throughput, with support for OpenGL 4.3 and DX11.

With respect to fabrication process technology, yes, Intel generally has the lead, but other foundries such as TSMC will benefit moving forward from ongoing investments made by large companies such as Apple and Qualcomm (in addition to ongoing investments made by smaller companies such as NVIDIA).

 

[raghu78]

Next time, try to understand the numbers before blindly quoting them. The dip in overall Tegra revenue in Q1 2014 (fiscal date) relative to Q4 2013 (fiscal date) [ie. QoQ] is because T3 production is quickly ramping down in anticipation of T4 production ramping up. Since T4 was delayed by at least three months, T4 mass production will not start until Q2-Q3 2014 (fiscal date), so that is why Q1 2014 (fiscal date) revenue is down relative to Q1 2013 (fiscal date) [ie. YoY]. The same trend will continue next fiscal quarter too. The important thing is that, even with a three month delay to T4, overall Tegra revenue for fiscal year 2014 is projected to be in line with overall Tegra revenue for fiscal year 2013, which means that Q3-Q4 2014 (fiscal date) will show strong growth relative to prior quarters this fiscal year. Note that the large impact on overall Tegra revenue involved with ramping down prior gen Tegra will be less severe in the future with the introduction of new products such as T4i.

you really eat up whatever crap explanation Nvidia throws at the market. Why was Tegra 4 production delayed ? Nvidia did not push Tegra 4 out. They just don't have any high profile / high volume designs wins for Tegra 4. So now they are proritizing T4i to make up for the loss of revenue from low number of T4 design wins. If revenue this year is going to be the same as last year with a new T4i (Nvidia's first integrated LTE chip) product which is a cheaper and higher volume product then it also means Tegra 4 is not expected to sell as well as Tegra 3. This has a clear relation to the lack of design wins for Tegra 4.

http://www.digitimes.com/news/a20130318PD225.html
http://www.digitimes.com/news/a20130201PD209.html
http://www.fudzilla.com/home/item/31138-nvidia-scores-sole-tegra-4-design-win-courtesy-of-zte

This is incorrect. [Dual-core] Clovertrail SoC's are most certainly not faster than [Quad-core] A15 SoC's such as T4. As for Baytrail, it will not be available until end of 2013. T5 "Logan" will be sampling by then, and will be in production early 2014.

22nm FINFET Baytrail will be fighting TSMC 28nm ARM A15 chips. 20nm is not going to be available in any meaningful volume till H2 2014. the volume of 20nm in Q2 2014 is going to 2% of TSMC production. read TSMC Q1 2013 earnings call transcript at seekingalpha for the above info. even then the graphics companies will be the first to use TSMC 20nm. Also 22nm FINFET is vastly superior to even 20nm process of foundries. thats the reason TSMC, GF and Samsung are running to 16nm FINFET and 14nm FINFET a year after 20nm.

so next year its about 2nd gen 28nm ARM A15 designs against the 22nm FINFET Baytrail SOC. the Silvermont Atom core is a completely new out of order microacrchitecture on a low power optimized Intel process. its going to be a one way traffic with Intel decimating the ARM designs on perf, perf/watt

Obviously improved integrated graphics can cannibalize low end discrete GPU's (as I said earlier), but that impact will be more than offset by Tegra moving up the food chain in areas where NVIDIA is not a player today. Within one year's time, T5 is expected to be available with ~ 400 GFLOPS GPU throughput, with support for OpenGL 4.3 and DX11.

With respect to fabrication process technology, yes, Intel generally has the lead, but other foundries such as TSMC will benefit moving forward from ongoing investments made by large companies such as Apple and Qualcomm (in addition to ongoing investments made by smaller companies such as NVIDIA).

rubbish. the T5 Logan is expected to have around 200 GFLOPS which in itself is not easy at 5W TDP for a quad core ARM A15 SOC, especially on 28nm.

http://www.bit-tech.net/news/hardware/2013/03/29/tegra-5-faster-xbox-360-ps3/1

With 22nm FINFET Atom due late this year and 14nm FINFET Atom scheduled for late 2014 volume production the next 2 - 3 years will show how much Intel's process lead is helping it to displace the ARM guys from the mobile market.

 

[sontin]

you really eat up whatever crap explanation Nvidia throws at the market. Why was Tegra 4 production delayed ? Nvidia did not push Tegra 4 out. They just don't have any high profile / high volume designs wins for Tegra 4.

Nonsense. nVidia explained why they delayed Tegra 4. And that's the reason why there is no "high profile" device out with Tegra 4 right now.
You have no clue what will come in the second half of this year.
For example: There is a rumour from Digitimes that Microsoft is working on a Tegra 4 8" RT tablet.

If revenue this year is going to be the same as last year with a new T4i (Nvidia's first integrated LTE chip) product which is a cheaper and higher volume product then it also means Tegra 4 is not expected to sell as well as Tegra 3. This has a clear relation to the lack of design wins for Tegra 4.

http://www.digitimes.com/news/a20130318PD225.html
http://www.digitimes.com/news/a20130201PD209.html
http://www.fudzilla.com/home/item/31138-nvidia-scores-sole-tegra-4-design-win-courtesy-of-zte

Tegra 4i is expected to ship in late FQ Q4 2014 for them. It attributes nearly nothing to the revenue of the Tegra segment for FY 2014.

Get your facts right before making false statements. :|

 

[raghu78]

Nonsense. nVidia explained why they delayed Tegra 4. And that's the reason why there is no "high profile" device out with Tegra 4 right now. You have no clue what will come in the second half of this year. For example: There is a rumour from Digitimes that Microsoft is working on a Tegra 4 8" RT tablet.

what was the explanation you are talking about. whatever it was not many are buying it.

http://seekingalpha.com/article/1423521-nvidia-stumbles-on-the-next-generation-tegra

"Revenue for the Tegra segment declined dramatically from $208.4 million in fiscal Q4 to $103.1 million in Q1, a greater than 50% decline sequentially and a 22% decline from the year ago quarter.
Nvidia management offered up a rather ludicrous "we meant to do that" defense. Nvidia management claimed that the successor to the in-production Tegra 3 quad-core processor, the Tegra 4, had been deliberately delayed as the schedule for a similar Tegra 4i processor was moved up. "

 

[sontin]

I don't care if other people buy it. They explained their reasons for the delay.

 

[ams23]

Why was Tegra 4 production delayed ? Nvidia did not push Tegra 4 out. They just don't have any high profile / high volume designs wins for Tegra 4.

Trying using some common sense for a minute. It is not relevant to know exactly why T4 was delayed (NVIDIA's explanation is that they prioritized T4i and T5 to speed up time to market, which is certainly a plausible explanation given that resources are not unlimited and given that there is stiffer competition at end of year 2013 than at beginning of year 2013), but it is relevant to know that T4 production was delayed and it is relevant to know that T3 production was ramping down when analyzing Tegra financial results, especially relative to prior qtr and prior year. As for the myth about no quality T4 design wins, that has already been debunked by NVIDIA and their partners (such as HP, Toshiba, Vizio, etc.), but you lack the patience to wait for designs to come to market such as this one: http://www8.hp.com/us/en/ads/x2/slatebook-x2.html

22nm FINFET Baytrail will be fighting TSMC 28nm ARM A15 chips. 20nm is not going to be available in any meaningful volume till H2 2014.

By the time that Baytrail is available, A15-based SoC's such as T4 will have been available for many months, and next gen SoC's such as T5 will be right around the corner. The mobile industry moves incredibly fast, and ARM licensees are not standing still. Process tech is only one part of the equation when it comes to ultra mobile SoC design. For instance, Apple has yet to make use of bleeding edge fabrication process technology in their ultra mobile SoC's (although that may change in the future), and they appear to be competing just fine with respect to perf and perf/w. In general, I do believe that you are overstating Intel's fabrication process tech advantage in the ultra mobile space. At the moment, Intel's ultra mobile fabrication process tech is actually in some ways slightly behind TSMC. Intel will move to 22nm ultra mobile at the end of 2013 and early 2014, and TSMC will move to 20nm ultra mobile by early to middle of 2014. Intel will move to 14nm ultra mobile by end of 2014 and early 2015, but they will stay at 14nm for about two years (until end of 2016 and early 2017), and TSMC will move to 16nm FinFET ultra mobile by early to middle of 2016.

rubbish. the T5 Logan is expected to have around 200 GFLOPS which in itself is not easy at 5W TDP for a quad core ARM A15 SOC, especially on 28nm.

No, not rubbish. T5 is expected to have up to ~ 400 GFLOPS throughput based on commentary that it will be similar in performance to "Kayla" and based on commentary that it will be faster than PS3/Xbox360. 200 GFLOPS throughput would not qualify as "faster than" PS3/Xbox360, nor would it qualify as being anywhere close to Kayla. TDP is unknown, but NVIDIA did state that T5 will not require any fan/heatsink.

Anyway, improved Intel integrated graphics cannibalizing low end NVIDIA discrete graphics is not an issue because Tegra will become good enough that it will end up cannibalizing low end NVIDIA discrete graphics. In fact, at some point well into the future, each and every product that NVIDIA sells will be a Tegra.

 

[raghu78]

Trying using some common sense for a minute. It is not relevant to know exactly why T4 was delayed (NVIDIA's explanation is that they prioritized T4i and T5 to speed up time to market, which is certainly a plausible explanation given that resources are not unlimited and given that there is stiffer competition at end of year 2013 than at beginning of year 2013), but it is relevant to know that T4 production was delayed and it is relevant to know that T3 production was ramping down when analyzing Tegra financial results, especially relative to prior qtr and prior year. As for the myth about no quality T4 design wins, that has already been debunked by NVIDIA and their partners (such as HP, Toshiba, Vizio, etc.), but you lack the patience to wait for designs to come to market such as this one: http://www8.hp.com/us/en/ads/x2/slatebook-x2.html

In fact you should try taking your own advice. why would Nvidia delay their flagship Tegra SOC if they had a large number of high volume design wins. Given these chips have a very short lifecycle of 1 year it does not make any sense. You don't spend millions of dollars designing a state of the art SOC chip and taping out and then delay production ramp. you don't do such a thing unless you really don't have the volume of design wins and would rather utilize your resources in another segment where you have more design wins lined up and more potential customers. T4i caters to low cost smartphones with integrated LTE. it has a quad core a9 which should be pretty outdated for 2014 and primarily for low cost / high volume smartphones.

By the time that Baytrail is available, A15-based SoC's such as T4 will have been available for many months, and next gen SoC's such as T5 will be right around the corner. The mobile industry moves incredibly fast, and ARM licensees are not standing still. Process tech is only one part of the equation when it comes to ultra mobile SoC design. For instance, Apple has yet to make use of bleeding edge fabrication process technology in their ultra mobile SoC's (although that may change in the future), and they appear to be competing just fine with respect to perf and perf/w. In general, I do believe that you are overstating Intel's fabrication process tech advantage in the ultra mobile space. At the moment, Intel's ultra mobile fabrication process tech is actually in some ways slightly behind TSMC. Intel will move to 22nm ultra mobile at the end of 2013 and early 2014, and TSMC will move to 20nm ultra mobile by early to middle of 2014. Intel will move to 14nm ultra mobile by end of 2014 and early 2015, but they will stay at 14nm for about two years (until end of 2016 and early 2017), and TSMC will move to 16nm FinFET ultra mobile by early to middle of 2016.

How did you even come up with a ridiculous suggestion that TSMC is ahead in terms of process tech compared to Intel. the transistor performance and leakage characteristics of Intel's 32nm process is unmatched by any foundry 28nm process. TSMC's 20nm planar transistor process does not stand a chance against Intel's 22nm FINFET. Also TSMC's 16nm FINFET is based on 20nm design rules. so there will be no area reduction compared to 20nm only power and performance improvements due to the FINFET transistor device. Intel's 14nm FINFET is a true full node with close to 50% area reduction compared to 22nm and performance/power improvements.

http://semimd.com/blog/2012/10/16/tsmc-accelerates-finfet-roadmap/

In an interview after the keynote, Liu said TSMC will take a “modular fin” approach in finFETs. TSMC will marry a 16nm fin with a 20nm backend. “It has 20nm design rules,” he said.

Even that 16nm FINFET production is scheduled for volume production in H2 2015 with major volume only in 2016.

http://seekingalpha.com/article/1340341-intel-s-manufacturing-advantage-is-bigger-than-you-think

"According to TSMC's latest conference call, there will be minimal 16nm volume during 2015, and it is really 2016 when we will see a high volume ramp up of products on the process."

 

[sontin]

In fact you should try taking your own advice. why would Nvidia delay their flagship Tegra SOC if they had a large number of high volume design wins. Given these chips have a very short lifecycle of 1 year it does not make any sense.

They explained it. Is it really so hard to use google to get all the informations? :|

You don't spend millions of dollars designing a state of the art SOC chip and taping out and then delay production ramp.

They delayed the tape-out and not the production ramp.

you don't do such a thing unless you really don't have the volume of design wins and would rather utilize your resources in another segment where you have more design wins lined up and more potential customers.

Makes sense IF they had another 28m Tegra chip on the market right now.

T4i caters to low cost smartphones with integrated LTE. it has a quad core a9 which should be pretty outdated for 2014 and primarily for low cost / high volume smartphones.

There will be no faster 1W SoC on the market at the beginning of 2014. And a 2,3GHz A9r4 will be faster than Krait200 and Krait300 which Qualcomm is using with their Snapdragon 400 and 600 series.

 

[raghu78]

They delayed the tape-out and not the production ramp.

wrong. tapeout has already happened. Nvidia had Tegra 4 silicon running at CES 2013. In fact you are clueless if you didn't know that. :whiste: tapeout to volume production is 9 - 12 months. For a Q3 2013 release Tegra 4 tapeout would be early Q4 2012. The pushed out the production ramp of Tegra 4 and that was clearly stated in their Q1 earnings call.

Makes sense IF they had another 28m Tegra chip on the market right now.

maybe to you. it does not make sense especially when Nvidia stated that Tegra 3 chip is ramping down production and the revenues are down 22% compared with the same quarter year ago.

http://seekingalpha.com/article/142...ses-f1q-2014-results-earnings-call-transcript

"Sales volume of Tegra 3 processors declined as customers began to ramp down production of Tegra 3 base mark phones and tablets. We expect this to continue in to the next quarter as customers start to announce Tegra 4 design with further new designs and phone ramp starting in the second half of the year."

 

[sontin]

wrong. tapeout has already happened. Nvidia had Tegra 4 silicon running at CES 2013. In fact you are clueless if you didn't know that. :whiste: tapeout to volume production is 9 - 12 months. For a Q3 2013 release Tegra 4 tapeout would be early Q4 2012. The pushed out the production ramp of Tegra 4 and that was clearly stated in their Q1 earnings call.

So how many months are between october/november and july/august?
Oh 9-10 months.
So where is the delay in the production ramp when they will even hit your timeframe?

maybe to you. it does not make sense especially when Nvidia stated that Tegra 3 chip is ramping down production and the revenues are down 22% compared with the same quarter year ago.

http://seekingalpha.com/article/142...ses-f1q-2014-results-earnings-call-transcript

"Sales volume of Tegra 3 processors declined as customers began to ramp down production of Tegra 3 base mark phones and tablets. We expect this to continue in to the next quarter as customers start to announce Tegra 4 design with further new designs and phone ramp starting in the second half of the year."

Ah, oh?
You wrote:

you don't do such a thing unless you really don't have the volume of design wins and would rather utilize your resources in another segment where you have more design wins lined up and more potential customers.

So what is the other SoC in the other segment? :hmm:

 

[raghu78]

So how many months are between october/november and july/august?
Oh 9-10 months.
So where is the delay in the production ramp when they will even hit your timeframe?

so now that you are proven wrong on the tapeout you have jumped to say there is no delay in production ramp. my dates were estimates. Nvidia has publicly stated they have pushed out the production ramp. please read the earnings call transcript. don't be so desperate to defend your favourite company. :whiste:

 

[SirPauly]

at some point well into the future, each and every product that NVIDIA sells will be a Tegra.

Indeed and agree with this.

 

[sontin]

so now that you are proven wrong on the tapeout you have jumped to say there is no delay in production ramp. my dates were estimates. Nvidia has publicly stated they have pushed out the production ramp. please read the earnings call transcript. don't be so desperate to defend your favourite company. :whiste:

That was mine:

They delayed the tape-out and not the production ramp.

They didn't state that they have pushed out the production ramp. They are only talking about the overall delay of Tegra 4.

 

[ams23]

why would Nvidia delay their flagship Tegra SOC if they had a large number of high volume design wins.

Read what I wrote more carefully. NVIDIA claims that they shifted more resources towards T4i and T5 in order to improve time to market for those respective SoC's by several months, at the expense of time to market for T4. That is called a tradeoff. Even with the delay, most T4-powered devices will still be available in time for back-to-school sales and for holiday season sales, so this is arguably a good tradeoff.

How did you even come up with a ridiculous suggestion that TSMC is ahead in terms of process tech compared to Intel.

Again, try to read more carefully. With respect to ultra mobile fabrication process tech, there is no doubt that TSMC-fabricated ultra mobile SoC's [at 28nm LP] are in some ways using a more advanced fabrication process right now than Intel-fabricated ultra mobile SoC's [at 32nm]. Anandtech actually confirmed this in writing and in a graph where they measured CPU long idle power for Krait vs. Atom: http://images.anandtech.com/reviews/SoC/Intel/CTvKrait/idle-krait-cpu.png . Obviously this will change soon, but there is no denying that TSMC's 28nm LP fabrication process will result in lower idle power and higher transistor density compared to Intel's 32nm fabrication process. At the end of the day, due in part to the significant ongoing investments made from large players such as Apple and Qualcomm, with continuing ongoing investments from smaller players such as NVIDIA, new fabrication process tech will be pushed forward at TSMC even as Intel assumes a lead in that area.