what's most interesting is that nearly half their sales aren't from GPUs! they have really diversified and protected themselves incase PC gaming goes south, but that isn't about to happen any time soon. good on them!
Maybe by "GPUs" he meant the "Gaming" segment of their "Revenue by Market Platform" graph. That's what immediately occurred to me but I asked him because I don't really know, and "GPU" covers more than just gaming.
Automotive revenue was $113 million of the $160 million for Tegra, up 47% year-over-year.
This is amazing. That means cars are getting a lot smarter, and $113M is a lot of money, ~10% for a company that is primarily known for GPUs. So its a great diversification step and one no one 5 years ago thought would be a market with significant money in it.
Its cool how companies leverage their strength in one area and apply it to others.
It's cool how companies *that aren't run by incompetent morons* leverage their strength in one area and apply it to others. I'm sure if AMD's board weren't absolutely clueless and didn't change every year, their balance sheet would be looking a whole lot better.
That's a huge baseless assumption made by someone without any marketing, financial, or managing experience, let alone, an opinion made by someone who isn't intimately aware of the decisions being made by AMD's board of presidents.
AMD's not doing as well now due to Intel's much stronger leverage over the CPU market, and Nvidia's aggressive marketing. They're caught between a rock (Nvidia) and a hard place (Intel), with very little they can do to push out without sacrificing limited resources in either CPU or GPU advancements.
AMD sold their mobile GPU to Qualcomm. They failed to focus their remaining tech on the mobile sector. Their Kaveri part, which powers the PS4, benches around Atom class, which is also about ARM v8, while consuming Core levels of power.
Their Radeon part isn't much better, meaning they can't ship a handheld PS4 or XB1 part because their CPU and GPU are incapable.
ARM, Tegra, Pascal, and Atom, and Core M all have the ability to hit those power levels however.
companies are free to pick their calendar year start and finish times. the fact that it is q1 2017 is because nvidia chose their reporting calendar as such. their q1 xxxx starts in [feb xxxx - 1] year, q2 starts in [may xxxx - 1] year, ...
Any yet again every time Nvidia posts an earning report someone who doesn't understand FISCAL years (FY) posts their uninformed rant when it is themselves who show they lack even basic knowledge.
What is a 'Fiscal Year - FY'
A fiscal year (FY) is a period that a company or government uses for accounting purposes and preparing financial statements. The fiscal year may or may not be the same as a calendar year.
The company was incorporated in February 1998. This means their fiscal year runs from February to February. Their first quarter involves February, March and April.
The Tegra division makes a loss every quarter. The automotive sector does not generate enough revenue to make the group profitable. They needed smartphone design wins to make that happen and they failed it that respect. So, the comment about the "good call" in leaving the mobile market is bullshit.
The automotive sector is growing and should soon generate enough revenue to make the group profitable. The mobile market is under pricing pressure, I believe, but I am sure NVIDIA would rather be involved in it than not. That doesn't mean it wasn't a good call to leave. Not because they wouldn't want to be there, but because they were unsuccessful and it may have been better to find an alternative market than to continue with frivolity. Consider also that they most likely have pursued automotive with more fervor than they otherwise would have if they had been successful with mobile (look at Qualcomm's automotive efforts by comparison, for instance). Pursuing the automotive infotainment market has perhaps allowed them to establish relationships they would not have otherwise had with automakers for ADAS and self-driving cars, which present opportunities for revenue, not just for Tegra, but for the data center as well.
Sorry, that's bullshit. Tegra lost $239M on $559M sales last year. There is no way Automotive is going to make it profitable in the forseeable future. They would have to triple sales
They can triple sales. Multiple companies (10?) obtain over $1.5 billion in revenue from automobile infotainment. It's a $40+ Billion market that's growing. In addition, Tegra is integral to NVIDIA's plan to capture some of the deep learning inference compute market. Whether that's successful or not remains to be seen, but many of these trained networks will run on low power devices and if NVIDIA can prove their GPUs are the most advantageous way to run them it will open up many doors for Tegra products. If GPUs turn out to be good processors for inference, there's a good chance NVIDIA might have a software/middleware advantage that would carry them over companies like ARM and PowerVR just like it's carrying them over AMD in the data center and in HPC.
My, you are an optimist. It's always the next thing will make Tegra profitable, but Tegra has lost money every single year since it started in 2008. It's only that Nvidia makes so much money for GPU's that keeps this turkey afloat. If that GPU profit ever stops growing Tegra will be the first thing on the chopping block. That's what happaned to Intel's smartphone platform.
I'm not certain but I think Tegra may have turned a profit for NVIDIA in FY12 or FY13. There were definitely quarters in there where it was profitable. But, in any case, AMD has lost money every quarter for a while but they're still making stuff. Tesla has lost money for the last several years, ever since it came into existence.
And there's absolutely nothing clever about saying that if a company gets into financial trouble that it is going to have less resources to invest towards speculative areas. But in this case you are wrong. The GPU profit could stop growing and that would likely only cause NVIDIA to invest more heavily in automotive. Why? Because automotive is a growth segment and your supposition was that the GPU segment wasn't so attractive as before. They would still have plenty of money to make the investment too, since NVIDIA not only does NVIDIA have plenty of cash and low debt at the moment, your supposition was just that the GPU profits "stopped growing", so they are still bringing new cash in.
But as far as the struggles with the Tegra unit, a lot of it had to due with Qualcomm. Qualcomm was already fined for their practices by China and will probably be facing more punishment in the future by the EU and the US. NVIDIA was not the only company forced out by their practices. Texas Instruments was, too. Icera was trying to get regulatory attention against Qualcomm even before NVIDIA bought Icera. Now that China stepped in, MediaTek and a new Chinese SOC maker which I forget the name of have been able to compete. It takes time to shift focus from mobile to automotive. Yes, I am very optimistic that Tegra will be profitable in perhaps 2 years. it is also a strategically important part of the company even if its margins are lower. It gives them a jumping off point to future possible opportunities. It's a smart endeavor as long as they can make it profitable or at least break even, which I think they can.
I think it is a good call. I'm not sure they left so much as refused to compete with Intel's contra revenue scheme. Now that Intel has dropped out of the market that leaves an opening for Pascal and Tegra for VR enabled mobile devices, and anyone willing to license a GPU, or build around Pascal.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
36 Comments
Back to Article
poohbear - Thursday, May 12, 2016 - link
what's most interesting is that nearly half their sales aren't from GPUs! they have really diversified and protected themselves incase PC gaming goes south, but that isn't about to happen any time soon. good on them!Yojimbo - Thursday, May 12, 2016 - link
Half their sales aren't from GPUs? What do you mean by that?lilmoe - Thursday, May 12, 2016 - link
Are we reading the same charts?alistair.brogan - Thursday, May 12, 2016 - link
"GPUs accounted for most of the revenue"people don't usually read and comprehend before commenting :)
tipoo - Friday, May 13, 2016 - link
Benefit of the doubt, maybe he meant consumer GPUs. They really did make inroads into HPC, cloud compute, etc.Yojimbo - Friday, May 13, 2016 - link
Maybe by "GPUs" he meant the "Gaming" segment of their "Revenue by Market Platform" graph. That's what immediately occurred to me but I asked him because I don't really know, and "GPU" covers more than just gaming.ThomasS31 - Friday, May 13, 2016 - link
Or just irony? :))michael2k - Friday, May 13, 2016 - link
I think he meant gaming. If you look for a division that accounts for half their revenue then you get gaming platform revenue.Gaming platform revenue was up 17% to $687 million, consistent with the strong growth in PC gaming compared to the rest of the PC market.
webdoctors - Friday, May 13, 2016 - link
Its still mostly GPU money, but:Automotive revenue was $113 million of the $160 million for Tegra, up 47% year-over-year.
This is amazing. That means cars are getting a lot smarter, and $113M is a lot of money, ~10% for a company that is primarily known for GPUs. So its a great diversification step and one no one 5 years ago thought would be a market with significant money in it.
Its cool how companies leverage their strength in one area and apply it to others.
The_Assimilator - Friday, May 13, 2016 - link
It's cool how companies *that aren't run by incompetent morons* leverage their strength in one area and apply it to others. I'm sure if AMD's board weren't absolutely clueless and didn't change every year, their balance sheet would be looking a whole lot better.JoeyJoJo123 - Friday, May 13, 2016 - link
That's a huge baseless assumption made by someone without any marketing, financial, or managing experience, let alone, an opinion made by someone who isn't intimately aware of the decisions being made by AMD's board of presidents.AMD's not doing as well now due to Intel's much stronger leverage over the CPU market, and Nvidia's aggressive marketing. They're caught between a rock (Nvidia) and a hard place (Intel), with very little they can do to push out without sacrificing limited resources in either CPU or GPU advancements.
michael2k - Friday, May 13, 2016 - link
AMD sold their mobile GPU to Qualcomm. They failed to focus their remaining tech on the mobile sector. Their Kaveri part, which powers the PS4, benches around Atom class, which is also about ARM v8, while consuming Core levels of power.Their Radeon part isn't much better, meaning they can't ship a handheld PS4 or XB1 part because their CPU and GPU are incapable.
ARM, Tegra, Pascal, and Atom, and Core M all have the ability to hit those power levels however.
ddriver - Friday, May 13, 2016 - link
What is interesting is your poor ability to compare numbers.Ice-Tea - Friday, May 13, 2016 - link
This is now the second site that mixes up 2017 and 2016. More troubling perhaps: 6 comments and nobody notices...Dribble - Friday, May 13, 2016 - link
Explain - figures look fine to me.Nvidia made 13% more in Q1 17 then Q1 16, but 7% less then Q4.
Ice-Tea - Friday, May 13, 2016 - link
IT'S 2016!!!vladx - Friday, May 13, 2016 - link
Before bashing others I suggest educating yourslf. Nvidia financial reports are 1 year ahead in name.BurntMyBacon - Friday, May 13, 2016 - link
Talk to nVidia about that. According to them, it is Fiscal Year 2017.XZerg - Friday, May 13, 2016 - link
companies are free to pick their calendar year start and finish times. the fact that it is q1 2017 is because nvidia chose their reporting calendar as such. their q1 xxxx starts in [feb xxxx - 1] year, q2 starts in [may xxxx - 1] year, ...MrSpadge - Friday, May 13, 2016 - link
They're explicitly reporting on their "fiscal year 2017". A weird accounting thing, but quite common.BedfordTim - Friday, May 13, 2016 - link
US accounting dates are weird so while it is 2016 the accounting year is called 2017.HighTech4US - Friday, May 13, 2016 - link
Any yet again every time Nvidia posts an earning report someone who doesn't understand FISCAL years (FY) posts their uninformed rant when it is themselves who show they lack even basic knowledge.What is a 'Fiscal Year - FY'
A fiscal year (FY) is a period that a company or government uses for accounting purposes and preparing financial statements. The fiscal year may or may not be the same as a calendar year.
http://www.investopedia.com/terms/f/fiscalyear.asp
Nvidia's current FY ends January 2017, so yes this is their FY2017 Q1.
jwcalla - Friday, May 13, 2016 - link
The company was incorporated in February 1998. This means their fiscal year runs from February to February. Their first quarter involves February, March and April.lefty2 - Friday, May 13, 2016 - link
The Tegra division makes a loss every quarter. The automotive sector does not generate enough revenue to make the group profitable. They needed smartphone design wins to make that happen and they failed it that respect. So, the comment about the "good call" in leaving the mobile market is bullshit.jwcalla - Friday, May 13, 2016 - link
How much of a loss does the Tegra division make every quarter?lefty2 - Friday, May 13, 2016 - link
They try to hide that information. You have to look at their 10-Q form here: http://investor.nvidia.com/financials.cfmloss was $239 million for 2016
Yojimbo - Friday, May 13, 2016 - link
The automotive sector is growing and should soon generate enough revenue to make the group profitable. The mobile market is under pricing pressure, I believe, but I am sure NVIDIA would rather be involved in it than not. That doesn't mean it wasn't a good call to leave. Not because they wouldn't want to be there, but because they were unsuccessful and it may have been better to find an alternative market than to continue with frivolity. Consider also that they most likely have pursued automotive with more fervor than they otherwise would have if they had been successful with mobile (look at Qualcomm's automotive efforts by comparison, for instance). Pursuing the automotive infotainment market has perhaps allowed them to establish relationships they would not have otherwise had with automakers for ADAS and self-driving cars, which present opportunities for revenue, not just for Tegra, but for the data center as well.lefty2 - Friday, May 13, 2016 - link
Sorry, that's bullshit. Tegra lost $239M on $559M sales last year. There is no way Automotive is going to make it profitable in the forseeable future. They would have to triple salesHighTech4US - Friday, May 13, 2016 - link
The only BS is what you are spewing.jasonelmore - Friday, May 13, 2016 - link
you realize we are in the cusp of a EV and Self Driving Car revolution right?Yojimbo - Friday, May 13, 2016 - link
They can triple sales. Multiple companies (10?) obtain over $1.5 billion in revenue from automobile infotainment. It's a $40+ Billion market that's growing. In addition, Tegra is integral to NVIDIA's plan to capture some of the deep learning inference compute market. Whether that's successful or not remains to be seen, but many of these trained networks will run on low power devices and if NVIDIA can prove their GPUs are the most advantageous way to run them it will open up many doors for Tegra products. If GPUs turn out to be good processors for inference, there's a good chance NVIDIA might have a software/middleware advantage that would carry them over companies like ARM and PowerVR just like it's carrying them over AMD in the data center and in HPC.lefty2 - Friday, May 13, 2016 - link
My, you are an optimist. It's always the next thing will make Tegra profitable, but Tegra has lost money every single year since it started in 2008. It's only that Nvidia makes so much money for GPU's that keeps this turkey afloat. If that GPU profit ever stops growing Tegra will be the first thing on the chopping block. That's what happaned to Intel's smartphone platform.Yojimbo - Friday, May 13, 2016 - link
I'm not certain but I think Tegra may have turned a profit for NVIDIA in FY12 or FY13. There were definitely quarters in there where it was profitable. But, in any case, AMD has lost money every quarter for a while but they're still making stuff. Tesla has lost money for the last several years, ever since it came into existence.And there's absolutely nothing clever about saying that if a company gets into financial trouble that it is going to have less resources to invest towards speculative areas. But in this case you are wrong. The GPU profit could stop growing and that would likely only cause NVIDIA to invest more heavily in automotive. Why? Because automotive is a growth segment and your supposition was that the GPU segment wasn't so attractive as before. They would still have plenty of money to make the investment too, since NVIDIA not only does NVIDIA have plenty of cash and low debt at the moment, your supposition was just that the GPU profits "stopped growing", so they are still bringing new cash in.
But as far as the struggles with the Tegra unit, a lot of it had to due with Qualcomm. Qualcomm was already fined for their practices by China and will probably be facing more punishment in the future by the EU and the US. NVIDIA was not the only company forced out by their practices. Texas Instruments was, too. Icera was trying to get regulatory attention against Qualcomm even before NVIDIA bought Icera. Now that China stepped in, MediaTek and a new Chinese SOC maker which I forget the name of have been able to compete. It takes time to shift focus from mobile to automotive. Yes, I am very optimistic that Tegra will be profitable in perhaps 2 years. it is also a strategically important part of the company even if its margins are lower. It gives them a jumping off point to future possible opportunities. It's a smart endeavor as long as they can make it profitable or at least break even, which I think they can.
lefty2 - Saturday, May 14, 2016 - link
Tegra lost $157M in 2013 and $60M 2012, since then it's gone downhill.Yojimbo - Saturday, May 14, 2016 - link
Until now, now it's going uphill.michael2k - Friday, May 13, 2016 - link
I think it is a good call. I'm not sure they left so much as refused to compete with Intel's contra revenue scheme. Now that Intel has dropped out of the market that leaves an opening for Pascal and Tegra for VR enabled mobile devices, and anyone willing to license a GPU, or build around Pascal.