Nvidia leaving Tablet/Smartphones Tegra a failure

I love this part

"Huang: [Tegra 4i] wasn't that successful for us."

Translation "Tegra is an UBER failure and 4i was the nail in the coffin"

and also

"If you want to be an innovative company, you have to fail."

Well, I guess Tegra is "innovative".
 
Stoly said:
I love this part

"Huang: [Tegra 4i] wasn't that successful for us."

Translation "Tegra is an UBER failure and 4i was the nail in the coffin"

and also

"If you want to be an innovative company, you have to fail."

Well, I guess Tegra is "innovative".
Click to expand...

Tegra wound up being the GeForce Titan of the tablet/smartphone space - not only overkill, but pricey.

Look at the hardware out there with Tegra - none of it is, in any way, inexpensive. That is, if anything, Tegra's biggest flaw. (Huang himself said as much.) Android hardware sells based on low prices - period. Unless you're Apple, the only way you are going to compete with Android is based on cheap hardware for the same price AS Android hardware. (Look at the two generations of Microsoft SurfaceRT - the only non-Android tablets to use Tegra. They have been universally whacked for higher prices than Android hardware that did NOT use Tegra - nobody compared Surface - in terms of either features OR price - to a Tegra-based Android tablet.)

In fact, has anybody come up with any flaws with Tegra OTHER than price?
 
PGHammer said:
Tegra wound up being the GeForce Titan of the tablet/smartphone space - not only overkill, but pricey.

Look at the hardware out there with Tegra - none of it is, in any way, inexpensive. That is, if anything, Tegra's biggest flaw. (Huang himself said as much.) Android hardware sells based on low prices - period. Unless you're Apple, the only way you are going to compete with Android is based on cheap hardware for the same price AS Android hardware. (Look at the two generations of Microsoft SurfaceRT - the only non-Android tablets to use Tegra. They have been universally whacked for higher prices than Android hardware that did NOT use Tegra - nobody compared Surface - in terms of either features OR price - to a Tegra-based Android tablet.)

In fact, has anybody come up with any flaws with Tegra OTHER than price?
Click to expand...

features, performance,temp, battery life just on top of my head.
 
I think the major flaw with tegra is "too little too late" and almost ZERO design wins.

Since the original tegra, nvidia hasn't been able to bring it on time.

The original tegra looked good on paper. but by the time it came out there were already cheaper, faster and more feature rich alternatives.

History repeated itself with every generation except Tegra 3.

Now, K1 looks great, but again nobody wants it for tablets, much less phones.
 
Nvidia hyped up Tegra too much and couldn't deliver. I bought into the Tegra3 hype and it turned out to be inferior at the time to Exynos 4412 for performance, power consumption and stability. Worse of all the Tegra advantage that they touted with Tegra enabled games like Dead Trigger was just a paid configuration handicap (sort of like Intel compiler intentionally producing slower code for non-Intel CPU and Apple paying developers to delay Android app release behind iOS).

http://www.jayceooi.com/2012/07/05/how-to-enable-dead-trigger-tegra-3-extended-effects-on-non-tegra-3-device/
 
How about the Tegra K1? I heard of one smartphone or tablet coming out with it, but for the East Asian market. Supposedly it's coming out this year. Other than that, I haven't heard of actual hardware coming out that may use the K1 outside of hardware demos from Nvidia.

Nvidia is going to stop production on that entirely as well?
 
The thread title and Guru3D headline is not exactly accurate and arguably highly misleading.

The question was framed with "smartphones and tablets" however the actual reply to the question during the interview specifies smartphones.

No need for secondary sources - http://www.cnet.com/news/nvidia-ceo-sees-future-in-cars-and-gaming-q-a/

octoberasian said:
How about the Tegra K1? I heard of one smartphone or tablet coming out with it, but for the East Asian market. Supposedly it's coming out this year. Other than that, I haven't heard of actual hardware coming out that may use the K1 outside of hardware demos from Nvidia.

Nvidia is going to stop production on that entirely as well?
Click to expand...

It is used in the new Xiaomi tablet.

See above for the rest of your question in regards to Tegra.
 
limitedaccess said:
The thread title and Guru3D headline is not exactly accurate and arguably highly misleading.

The question was framed with "smartphones and tablets" however the actual reply to the question during the interview specifies smartphones.

No need for secondary sources - http://www.cnet.com/news/nvidia-ceo-sees-future-in-cars-and-gaming-q-a/



It is used in the new Xiaomi tablet.

See above for the rest of your question in regards to Tegra.
Click to expand...
Thank you for the link.

It sounds like they're shifting their focus from smarphones to tablets and Android gaming devices like the Shield.
 
PGHammer said:
Look at the hardware out there with Tegra - none of it is, in any way, inexpensive. That is, if anything, Tegra's biggest flaw. (Huang himself said as much.) Android hardware sells based on low prices - period. Unless you're Apple, the only way you are going to compete with Android is based on cheap hardware for the same price AS Android hardware. (Look at the two generations of Microsoft SurfaceRT - the only non-Android tablets to use Tegra. They have been universally whacked for higher prices than Android hardware that did NOT use Tegra - nobody compared Surface - in terms of either features OR price - to a Tegra-based Android tablet.)
Click to expand...

The Nexus 7 was pretty cheap and one of the cheaper tablets at the time. I think the rumor mill was that a Tegra 3 cost about $25-35. I'm not sure if that was considered cheap or expensive in the ARM world at the time.

But Jen-Hsun is saying what everyone knows, and even Qualcomm knows: there isn't a huge market for top-end SoCs. Qualcomm mentioned this awhile ago, that their big chips are where the profit margins are, but volume is low and it's a hard sell. The lower-end chips are going to be going up against the cheapy commodity MediaTek stuff.

For phones there is almost no need for a big GPU. Even the off-the-shelf ARM designs are more than sufficient for 99.9% of everybody.

NVIDIA is smart to get out of the smartphone market. It's become commoditized and now it's a race to the bottom, just like Dell experienced in the PC market. They should focus Tegra on specialty devices and gaming systems. Their GPU tech is what sets them apart from other companies so they need products that leverage that strength. A phone is not one of those products. (That could change if we eventually get into these hybrid "phone docks as your desktop" type devices like Ubuntu is trying to make. And NVIDIA has done a lot of work to get Ubuntu running well on their Tegra chips as it's the OS of choice for their dev boards.)

In fact, has anybody come up with any flaws with Tegra OTHER than price?
Click to expand...

Radio support is the big one for smartphone wins, especially in America.
 
jwcalla said:
The Nexus 7 was pretty cheap and one of the cheaper tablets at the time. I think the rumor mill was that a Tegra 3 cost about $25-35. I'm not sure if that was considered cheap or expensive in the ARM world at the time.

But Jen-Hsun is saying what everyone knows, and even Qualcomm knows: there isn't a huge market for top-end SoCs. Qualcomm mentioned this awhile ago, that their big chips are where the profit margins are, but volume is low and it's a hard sell. The lower-end chips are going to be going up against the cheapy commodity MediaTek stuff.

For phones there is almost no need for a big GPU. Even the off-the-shelf ARM designs are more than sufficient for 99.9% of everybody.

NVIDIA is smart to get out of the smartphone market. It's become commoditized and now it's a race to the bottom, just like Dell experienced in the PC market. They should focus Tegra on specialty devices and gaming systems. Their GPU tech is what sets them apart from other companies so they need products that leverage that strength. A phone is not one of those products. (That could change if we eventually get into these hybrid "phone docks as your desktop" type devices like Ubuntu is trying to make. And NVIDIA has done a lot of work to get Ubuntu running well on their Tegra chips as it's the OS of choice for their dev boards.)



Radio support is the big one for smartphone wins, especially in America.
Click to expand...

Except for the radio issue (which isn't really an issue with Tegra), Huang was right about SoCs becoming commoditized, especially in terms of tablets and smartphones. That is why I commented specifically about the price issue - which is where Tegra is outmatched. Tegra is quite capable as a high-end SoC - which is exactly why Microsoft chose it for Surface, and it also explains the design wins it HAS gotten; all are where price was NOT a major concern. However - except for Apple - the tablet/smartphone market has indeed devolved down to a race to the bottom. Even in North America (including the US), radio support is not THAT big of an issue; price, on the other hand, is. (Seriously, what is the price difference between the multiple sorts of LTE in North America in terms of radio support?)
 
I wonder if this includes the Nvidia Shield. I don't know what they were thinking when they thought that was a good idea. Mobile tech moves a break neck speed and to try and break into the console/handheld market with the Shield wasn't a good move imo.
 
Shield is their last chance lloose.

Android touchscreen gaming is a saturated market, but for devices with real controllers it's still an open field. Most devices don't have built in joysticks, and up until now none of those that do have been built with any quality. Nvidia brings that to the table, along with enough industry heft to get some good ports.

Google left the door wide open here by not establishing standard controller specs for Android. Nvidia is attempting to bridge that gap, provide a superior gaming solution, and still leverage Android compatibility to ease porting. Given their situation, it's the best move they could have made, although they need to make a more portable version of shield before most will take it seriously.

There is no other growth market remaining in mobile or tablets.
 
PGHammer said:
Tegra wound up being the GeForce Titan of the tablet/smartphone space - not only overkill, but pricey.

Look at the hardware out there with Tegra - none of it is, in any way, inexpensive. That is, if anything, Tegra's biggest flaw. (Huang himself said as much.) Android hardware sells based on low prices - period. Unless you're Apple, the only way you are going to compete with Android is based on cheap hardware for the same price AS Android hardware. (Look at the two generations of Microsoft SurfaceRT - the only non-Android tablets to use Tegra. They have been universally whacked for higher prices than Android hardware that did NOT use Tegra - nobody compared Surface - in terms of either features OR price - to a Tegra-based Android tablet.)

In fact, has anybody come up with any flaws with Tegra OTHER than price?
Click to expand...

Stoly said:
features, performance,temp, battery life just on top of my head.
Click to expand...

Everything that Stoly said. nVidia hyped and didn't deliver on time. Time was critical since the other players weren't late. And to think, people here still used to believe that nVidia's next mobile GPU was going to kick everyone's ass. Like believing the rants of a crazy person... :D
 
PGHammer said:
Except for the radio issue (which isn't really an issue with Tegra), Huang was right about SoCs becoming commoditized, especially in terms of tablets and smartphones. That is why I commented specifically about the price issue - which is where Tegra is outmatched. Tegra is quite capable as a high-end SoC - which is exactly why Microsoft chose it for Surface, and it also explains the design wins it HAS gotten; all are where price was NOT a major concern. However - except for Apple - the tablet/smartphone market has indeed devolved down to a race to the bottom. Even in North America (including the US), radio support is not THAT big of an issue; price, on the other hand, is. (Seriously, what is the price difference between the multiple sorts of LTE in North America in terms of radio support?)
Click to expand...

When I think of radio support I'm thinking of the power efficiency of the solutions, where it seems that Qualcomm has the advantage. I think Qualcomm practically owns the NA market, even to the point that Samsung prefers Qualcomm here over their own in-house designs.
 
The problem with Tegra is the price. Nvidia need to accept lower margins in order to be a serious ARM competitor. They are already dominating the GPU market, so they can maintain high margins there, but the mobile/ARM space is a fiercely competitive market. Nvidia needs to leave it's apple-like "pinky-out" high margin behind.
 
KazeoHin said:
The problem with Tegra is the price. Nvidia need to accept lower margins in order to be a serious ARM competitor. They are already dominating the GPU market, so they can maintain high margins there, but the mobile/ARM space is a fiercely competitive market. Nvidia needs to leave it's apple-like "pinky-out" high margin behind.
Click to expand...

But then there's no sense being in the sector.

Why try to sell these things at one millionth of a penny profit, especially when it would be nearly impossible to compete with the cheap Chinese stuff on price?
 
KazeoHin said:
The problem with Tegra is the price. Nvidia need to accept lower margins in order to be a serious ARM competitor. They are already dominating the GPU market, so they can maintain high margins there, but the mobile/ARM space is a fiercely competitive market. Nvidia needs to leave it's apple-like "pinky-out" high margin behind.
Click to expand...

I don't think that pricing was really an issue so far with Tegra. Probably a littlebit more expensive than Mediatekt but not more expensive than Qualcomm and Samsung. I also don't think that margins were high. I think it was more vice versa, that margins were low. The problem so far was more missing LTE integration and time to market of certain Tegra versions like Tegra 4 and Tegra 4i for instance. If Nvidia wan'ts to grow that business the first thing they need to improve on is time to market. K1 in both versions i think will be a great SoC and offers great features and performence. The question is how fast are they able to get it into the market. That will be the main factor making it either a success or failure. Pricing i don't think is an issue on this one. Nvidia does not intend so sell it into mainstream and low end smartphones. In tablets, portable gaming devices, cars, TVs and high end smart phones the SoC is not the main cost driver.
 
You must log in or register to reply here.
Back
Top