NVidia - how is it going as a company?

[ams23]

@AnandThenMan:

Kepler.M (in NVIDIA's Logan) vs. A6X (in ipad 4): www.youtube.com/watch?v=TPtgsrv5xqM

 

[Granseth]

@AnandThenMan:

Kepler.M (in NVIDIA's Logan) vs. A6X (in ipad 4): www.youtube.com/watch?v=TPtgsrv5xqM

I read ATM as saying that it doesn't matter until you have the product and see how it works. The chips can have as much promise as they want but it won't matter unless the products they are in utilizes the promises.

 

[ams23]

Of course perf. per watt matters. In fact, it matters more than anything else in a power constrained environment. The perf. per watt of Kepler.M will not go down between now and launch date. The number of CUDA "cores" [192] in Kepler.M will not go down. The massive headroom and ability of Kepler to reach very high clock operating frequencies (up to ~ 1GHz) on a 28nm fabrication process will not go away. Kepler.M is a pretty well known entity if you put it all together.

 

[Granseth]

You don't like reading what other people write, do you?
Tegra 5 could cure cancer, but it won't matter unless it is put into a product that utilizes that ability. And in the same way Tegra 5 won't matter unless it is put into products that makes good use of the APU, and most importantly it has to be put into a product people are willing to buy.

 

[ams23]

What you are saying doesn't make sense in the context of this discussion. Perf. per watt is a measure of efficiency. The better the GPU perf. per watt, the better the power efficiency of the entire system (all else equal). Obviously all ultra mobile products will make heavy use of the GPU.

Here is more of Logan in action: www.youtube.com/watch?v=Ta8WL0iwUDI&desktop_uri=/watch?v=Ta8WL0iwUDI

 

[blackened23]

You're not understanding him. All of your assertions with regard to efficiency are correct. Tegra 5, however, is not a complete or actual product yet so those theories completely go to waste until they're in a finalized product.

Tegra 4 demos existed eons before the actual product did. I don't believe those demos helped T4's cause.

 

[ams23]

Well if that is what he wants to know, then he will probably have to wait a few months until CES 2014 when new products are announced and demoed. Logan-powered tablets will not be available until Q1 2014 at the earliest.

 

[tviceman]

bay holy grail will achieve nothing.

The tegra 4 story will repeat again with the tegra 5.

Remember these words and come back in one year. i'm always right. Always.

now you can can go on with your delusions and hopeless expectatives. Seriously, go on.

Guys he is always right close the thread.

 

[Mondozei]

Guys he is always right close the thread.

The hilarious thing is, people who call out that obvious trolling behaviour get warned, but Imouto gets away with nothing.

But even pointing this out, as you did, is an offence subject to warning/banning. Especially publically.

Infraction issued for moderator callout.
-- stahlhart

 

[blackened23]

I just don't see the need to ever call someone a "fanboy". By doing so, you're admitting that you can't argue on a rational basis - besides this fact, there is absolutely nothing wrong with having a brand preference. That being said, someone arguing on the basis of "i'm always right" without real substantiation can get somewhat annoying. I get that.

Anyway, with regard to the topic, Tegra 4 didn't do as well as it should have but as I've mentioned before - nvidia knows exactly what mistakes were made with the T4. They won't do the same again with the Tegra 5; Tegra 5 is their make it or break it moment. Nvidia isn't stupid, they know what has to be done. They need integration on the level of qualcomm, and they need to release on time. I expect that to happen despite T4's failure to gain traction - if they do this, there is a good chance they can potentially regain all of the design wins that move large volumes, such as the Nexus 7.

 

[tviceman]

Anyway, with regard to the topic, Tegra 4 didn't do as well as it should have but as I've mentioned before - nvidia knows exactly what mistakes were made with the T4. They won't do the same again with the Tegra 5; Tegra 5 is their make it or break it moment. Nvidia isn't stupid, they know what has to be done. They need integration on the level of qualcomm, and they need to release on time. I expect that to happen despite T4's failure to gain traction - if they do this, there is a good chance they can potentially regain all of the design wins that move large volumes, such as the Nexus 7.

T4 was lackluster from the onset; it's presentation at CES by JHH was really, really dull and things did not get much better in the weeks following, even with demos and whatever else showing it's performance. Tegra 5, however, has much more excitement / hype / promise from the get-go. I was surprised and initially disappointed that T5 is going to be 28nm, but with every indication that 20nm won't even have products on the market until the 2nd half of 2014, 28nm may be a great decision. It would be nice to see Nvidia integrate a few custom tweaks to the A15 cores to bring down power consumption and/or increase performance without increasing power.

 

[Imouto]

I was mocking ams23 and blackened23 cheerleading on both Tegra and Bay Trail products.

Of course I'm not always right, I just find funny how you react to ppl behaving in the same way as you do.

Infraction issued for callout.
-- stahlhart

 

[blackened23]

I'm "cheerleading" Tegra products. That's hilarious, because I was precisely doing NOT that. It's obvious that Tegra 4 did not do well.

Anyway, If you're mocking other posters, you're admitting that you're trolling? Very nice. Guess we can't have normal adult like conversations about these things.

 

[Imouto]

Sorry I forgot a "respectively" at the end of the first sentence.

 

[Crap Daddy]

I'm "cheerleading" Tegra products. That's hilarious, because I was precisely doing NOT that. It's obvious that Tegra 4 did not do well.

Anyway, If you're mocking other posters, you're admitting that you're trolling? Very nice. Guess we can't have normal adult like conversations about these things.

It's just the beginning for Tegra 4. After the Xiaomi deal which should be big, the new Surface from Microsoft will be powered by Nvidia. They are doing just fine.

 

[2is]

I think nVidia is fine as a company. That said, I think Shield will be a flop.