Question 'Ampere'/Next-gen gaming uarch speculation thread

[Ottonomous]

How much is the Samsung 7nm EUV process expected to provide in terms of gains?
How will the RTX components be scaled/developed?
Any major architectural enhancements expected?
Will VRAM be bumped to 16/12/12 for the top three?
Will there be further fragmentation in the lineup? (Keeping turing at cheaper prices, while offering 'beefed up RTX' options at the top?)
Will the top card be capable of >4K60, at least 90?
Would Nvidia ever consider an HBM implementation in the gaming lineup?
Will Nvidia introduce new proprietary technologies again?

Sorry if imprudent/uncalled for, just interested in the forum member's thoughts.

 

[AtenRa]

Physics

The physics are tied to the engine. The GPU has zero impact on in game physics. It really comes down to the work the engine maker puts into the physics engine. Some games use a 3rd party physics engine like Havok. Some like Unreal Engine, use their own.

Correct, its another thing the Physics Engine inside the Game Engine and another thing the NVIDIA PhysX.
Also, you can use Havok in Unreal.

 

[Glo.]

Enjoy.

 

[Dribble]

Really more of an AMD video then an Nvidia one, particularly as the guy is a strong AMD fan. In short he says (for non RT) Nvidia vs AMD stays about the same as it is today. What he conspicuously avoids talking about is RT and DLSS, probably because here there will be no competition and Amphere will just crush everything.

 

[Stuka87]

Really more of an AMD video then an Nvidia one, particularly as the guy is a strong AMD fan. In short he says (for non RT) Nvidia vs AMD stays about the same as it is today. What he conspicuously avoids talking about is RT and DLSS, probably because here there will be no competition and Amphere will just crush everything.

Huh?

The majority of the video is spent on nVidia, and there was no perceivable bias anywhere, unless to you math is biased.

Both AMD and nVidia have RT in the coming cards. No comparison was done because nothing is known. A100 has no ray tracing at all, so we can't look at that. And all we know about AMD's solution is that it is vastly different from nVidia on the hardware side of things because every shader has RT functionality. While DLSS 2.0 is a big step up over the initial version, its only marginally better than existing up-scaling techniques (which were superior to DLSS 1.0).

 

[Glo.]

Huh?

The majority of the video is spent on nVidia, and there was no perceivable bias anywhere, unless to you math is biased.

Both AMD and nVidia have RT in the coming cards. No comparison was done because nothing is known. A100 has no ray tracing at all, so we can't look at that. And all we know about AMD's solution is that it is vastly different from nVidia on the hardware side of things because every shader has RT functionality. While DLSS 2.0 is a big step up over the initial version, its only marginally better than existing up-scaling techniques (which were superior to DLSS 1.0).

"The only biases we see are those that are in our own minds."

 

[DeathReborn]

Huh?

The majority of the video is spent on nVidia, and there was no perceivable bias anywhere, unless to you math is biased.

Both AMD and nVidia have RT in the coming cards. No comparison was done because nothing is known. A100 has no ray tracing at all, so we can't look at that. And all we know about AMD's solution is that it is vastly different from nVidia on the hardware side of things because every shader has RT functionality. While DLSS 2.0 is a big step up over the initial version, its only marginally better than existing up-scaling techniques (which were superior to DLSS 1.0).

I think it might be more focus on RDNA2 than on Ampere/Next-gen, which is what this thread is about. There is less info about Ampere (at least non HPC/AI/DL/ML Ampere) which is probably why not much in that video applies to this thread.

 

[nurturedhate]

I think it might be more focus on RDNA2 than on Ampere/Next-gen, which is what this thread is about. There is less info about Ampere (at least non HPC/AI/DL/ML Ampere) which is probably why not much in that video applies to this thread.

It's the screencap of the video, it's all AMD, if someone were to lackadaisically scroll past it one might think it was a video/presentation directly from AMD. Book/cover and all that.

 

[Glo.]

Adored talked about what Nvidia did with TSMC/Samsung Process for next gen cards, he talked about how next gen stacks compared to next gen AMD lineup.

Its all about context. For Nvidia, as Adored said its all about Context.

AMD being competitive is the key to understanding next generation GPUs, from both companies.

 

[ozzy702]

Adored talked about what Nvidia did with TSMC/Samsung Process for next gen cards, he talked about how next gen stacks compared to next gen AMD lineup.

Its all about context. For Nvidia, as Adored said its all about Context.

AMD being competitive is the key to understanding next generation GPUs, from both companies.

Are we back to believing anything he says? He has a really long road of being on point before I put stock into what he says.

 

[Glo.]

Are we back to believing anything he says? He has a really long road of being on point before I put stock into what he says.

His analysis usually is pretty accurate. Whether his info is accurate - remains to be seen, on the topic of next gen Nvidia GPUs.

Basically he repeated everything that leaked out about Next gen Nvidia GPUs and the reasons of what is happening with their GPUs.

 

[AtenRa]

Has this been posted before ??

 

[Stuka87]

Has this been posted before ??

It has not been, although its somewhat out of date as it was posted before nVidia showed A100 and talked a bit about gaming cards.

 

[DXDiag]

While DLSS 2.0 is a big step up over the initial version, its only marginally better than existing up-scaling techniques (which were superior to DLSS 1.0).

Nope, it basically blew every upscaling technique out of the water. No blur during motion, upscale 4X the resolution with minimal artifacts and double the performance, it's basically the jack of all trades.

 

[SirCanealot]

particularly as the guy is a strong AMD fan

He isn't an AMD fan, more so isn't a fan of the monopoly we have on GPUs at the moment. That and he hates NVidia's constant anti-consumer practices.

Everyone should be an AMD fan really in light of NVidia's years of anti-consumer practices (yes I know AMD aren't perfect).

 

[Dribble]

Has this been posted before ??

He did a later video (11 May vs this one which is 3rd May) and that was in this thread many pages back.

He isn't an AMD fan, more so isn't a fan of the monopoly we have on GPUs at the moment. That and he hates NVidia's constant anti-consumer practices.

Everyone should be an AMD fan really in light of NVidia's years of anti-consumer practices (yes I know AMD aren't perfect).

So he's not an AMD fan but is an Nvidia hater, which means he should be an AMD fan? Either way he's not known for being an unbiased source of information.

 

[Glo.]

So he's not an AMD fan but is an Nvidia hater, which means he should be an AMD fan? Either way he's not known for being an unbiased source of information.

Shoot the message, not the messenger.

 

[Mopetar]

I’ve never had a problem with any of his stuff. He painstaking goes through the information and steps he uses to arrive at his conclusions which makes them interesting even if they’re not spot on.

Even if he doesn’t lIke certainly companies for their business practices I don’t think he lets it influence his opinions about a product. Not many people who supposedly hate a company will go out of their way to explain why they think that company will still have a better product.

 

[Glo.]

And since when has pointing out clear fraud and anti-consumer practices has become OBJECTIVELY hating on a company, and made something devalue someone's message?

Its like complaining that a journalist has made an article in which he REPORTED what currently elected president of any country in the world has done wrong during his election, and calling him president's hater.

Before we will ever call anyone of their biases, lets look into the mirror about our OWN biases we have, about companies.

 

[itsmydamnation]

Nope, it basically blew every upscaling technique out of the water. No blur during motion, upscale 4X the resolution with minimal artifacts and double the performance, it's basically the jack of all trades.

nope , your wrong.

DLSS is crap

DLSSx2 looked really good, never appeared and is yet to appear

DLSS2 is not DDLSx2

 

[Konan]

nope , your wrong.

DLSS is crap

DLSSx2 looked really good, never appeared and is yet to appear

DLSS2 is not DDLSx2

Not sure what you mean by this “x2” insert.
DLSS 1.0 release Feb 2019
DLSS 2.0 release Aug 2019
- DLSS 2.0 (second update APR 2020)

DLSS 2.0 games tested.

DLSS 2.0 tested: why you should finally turn on Nvidia's AI-powered upscaling

We let Nvidia's super sampling artificial intelligence take the wheel in Wolfenstein: Youngblood and Control.

www.pcgamer.com

As this technology improves I reckon it’s going to be one of the best value additions in software in a long time. It’s only going to be enhanced with Ampere. The competition is really going to need some thing to compete with here.

 

[DXDiag]

DLSS is crap

Not according to the people who saw, tested and tried it.

 

[coercitiv]

Not sure what you mean by this “x2” insert.

It was probably meant as DLSS 2X, which Nvidia called the Ultra High Quality mode, trained on 64x super sampled reference images.

 

[Tup3x]

It was probably meant as DLSS 2X, which Nvidia called the Ultra High Quality mode, trained on 64x super sampled reference images.

It would be interesting to see what kind of result DLSS 2.0 would give when using at native internal resolution. Should offer rather nice quality without humongous performance drop.

 

[IntelUser2000]

Not according to the people who saw, tested and tried it.

The first DLSS sucked, it was true. It used power-hogging datacenters to result in something worse than simple software upscaling taking up megabytes of storage space.

The second version is much better and a genuine option to get better image quality or performance.

 

[tajoh111]

Shoot the message, not the messenger.

The thing I hate most about adored is his clear bias towards AMD which is most abundantly clear with his failure to call out AMD when they do something anti consumer but then accuse Nvidia of raising prices.

No clearer example of this is calling out GK104 for raising consumer prices, without calling Tahiti and Pitcairns first. These cards much like Turing kept price to performance vs last gen flat and actually had a negative price to performance vs last gen.

Remember 28nm was not really an expensive node as this was prior to finfet and AMD charged $350 for a 213mm2 die and $550 for a 351mm2 die. This was without Nvidia setting prices first(so you can't use that excuse). It was AMD initial 28nm pricing which set off the price escalation charging big money for small dies.

More thing he fails to call out is when AMD does a silent spec change like using a cut down variant of a chip, while keeping the name the same with no memory difference or pricing difference which means only AMD reaps the benefit. Yet he will call out something like the 3gb variant of the 1060, which atleast provides some cost savings and an easier way parts across.

AMD Silently Cuts Down Core Count on Radeon RX 560 Graphics Card - Now Shipping With 896 Cores Instead of 1024 SPs

AMD has silently introduced a new variant of the Polaris 11 based Radeon RX 560 which cuts down the core count.

wccftech.com

I would say one of the worst way AMD hurts even knowledgeable consumers is using viral and guerrilla marketing channels to get people to wait months to years for their products with fake leaks(fake release dates, extreme performance leaks) like they did with Fiji and Vega to stall potential buyers. And when these products come to light with similar price to performance as existing nvidia cards that have been on the shelf for months or over a year, they drum out the guerrilla marketing team to add rumors like Vega create a loss and can't be priced lower, nvidia is evil.

The funny thing is Adored finally caught on to this last point and semi terminated his relationship with AMD as he realized he was being used as a Guerilla marketing machine for AMD. He was pissed off he wasn't getting paid and while other youtubers were. It was in the video below.

https://www.reddit.com/r/realAMD/comments/cgsyhd

This video has since been made private or deleted and I suspect AMD has contacted him and decided to give him what he wanted.