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

Page 56 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Ottonomous

Senior member
May 15, 2014
559
292
136
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.
 

uzzi38

Platinum Member
Oct 16, 2019
2,698
6,393
146
It would be funny if it were fake but turned out to be really close anyway just because it's entirely plausible.

If truth can be stranger than fiction then why can't fiction be more real than true leaks?

The thing is it's mostly in line with the other leaks barring the 80Ti name (which is literally all names anyway - could always change at any time) and TSMC 7nm instead of SS 8nm. Otherwise those actual GPU specs excluding the node are basically almost exactly where I personally expect Ampere to land anyway.

But I mean, I could easily be wrong. I'm pretty much just speculating, something I've learnt the hard way I'm not all that great at hahaha.
 
Reactions: Mopetar and Konan

dr1337

Senior member
May 25, 2020
379
635
136

Yet another 3080ti rumor, this time it being a partner card thats supposedly leaked. If this is legit then we should be seeing consumer ampere fairly soon if AIB manufacturers already have hands on it.
 

FaaR

Golden Member
Dec 28, 2007
1,056
412
136
My bog, that thing looks like butt. The current-gen strix is perhaps no great beauty, but still several steps up from that abomination... Ugh.
 

GodisanAtheist

Diamond Member
Nov 16, 2006
7,035
7,456
136
Who knows if that's real or not, frankly its barely a thing given that all we're looking at is a fan shroud that might be a mock-up of a mock-up of a mock-up, but I've always had a preference for nice clean fan shroud designs.

Just give me a nice squared off box in a nice neutral color without a bunch of gamer bling and rgb and I'll be happy.

Love me some of the older XFX/Evga/PNY/Sapphire designs.
 

DooKey

Golden Member
Nov 9, 2005
1,811
458
136
Who knows if that's real or not, frankly its barely a thing given that all we're looking at is a fan shroud that might be a mock-up of a mock-up of a mock-up, but I've always had a preference for nice clean fan shroud designs.

Just give me a nice squared off box in a nice neutral color without a bunch of gamer bling and rgb and I'll be happy.

Love me some of the older XFX/Evga/PNY/Sapphire designs.

I've always been a fan of efficient cooler designs that don't sound like jet engines. I really don't care what they look like to be honest. Just work and work fast baby! That's my motto.
 
Reactions: Konan and Elfear

alcoholbob

Diamond Member
May 24, 2005
6,271
323
126

So buyers get to choose to spend $500 on a GPU or $500 on a console that has an equivalent GPU. Nvidia is brilliant at milking gamers for every dollar they have.
 

Glo.

Diamond Member
Apr 25, 2015
5,758
4,666
136
So buyers get to choose to spend $500 on a GPU or $500 on a console that has an equivalent GPU. Nvidia is brilliant at milking gamers for every dollar they have.
But consoles will be more efficient . /joke

500$ for RTX 3070 Ti - sure. I can live with it. If it will be the price of 3070 SKU - hell no. That would be absurdly stupid price tag, if there is also 3070 Ti SKU above that GPU.

Personally, I believe that 3070 Ti price tag will be 100$ lower - at 399$, and 3070 will be at 350. Like in the olden days...
 

alcoholbob

Diamond Member
May 24, 2005
6,271
323
126
But consoles will be more efficient . /joke

500$ for RTX 3070 Ti - sure. I can live with it. If it will be the price of 3070 SKU - hell no. That would be absurdly stupid price tag, if there is also 3070 Ti SKU above that GPU.

Personally, I believe that 3070 Ti price tag will be 100$ lower - at 399$, and 3070 will be at 350. Like in the olden days...

Well, they are. The Xbox One X can run games at 4K on a RX580 class card that you need at least a 1070 to run at 30fps in multiplatform games. So in reality, probably the $800 3080 will be roughly equivalent in practice to the next gen console hardware. Nvidia video cards will be a very bad value proposition across the board until Ampere refresh or Hopper launches IMO. It'll be like the price of Intel CPUs when Zen first launched. At least Turing was unequivocally king of the hill for the last 2 years. However if there is a similarity with Turing, I expect the 3080 Ti be the only card that's really "worth it" to get on day one.
 
Last edited:
Reactions: Elfear

BenSkywalker

Diamond Member
Oct 9, 1999
9,140
67
91
So JHH stating explicitly that the majority of Ampere chips are coming from TSMC was accepted, but then the Huwai ban happened removing what was their second largest customer and the COVID global slowdown happened which according to TSMC had serious negative impacts and saw their orders reduced with a large chunk of the slack picked up by nVidia(this according to TSMC themselves) but now it has to be Samsung because...... Twitter....?

The same source that said zero consumer Ampere parts? The same source that said no consumer parts until 2021?

Guessing core counts is pretty easy, stating that three governments, three multinational corporations and every semiconductor news outlet in the world is spreading lies to mask nVidia's launch because some guy on Twitter disagrees.....

Speculating used to revolve around using facts to make educated guesses, not hitching your cart to some random guy on Twitter and putting blinders on when reality says he's wrong.
 

uzzi38

Platinum Member
Oct 16, 2019
2,698
6,393
146
So JHH stating explicitly that the majority of Ampere chips are coming from TSMC was accepted, but then the Huwai ban happened removing what was their second largest customer and the COVID global slowdown happened which according to TSMC had serious negative impacts and saw their orders reduced with a large chunk of the slack picked up by nVidia(this according to TSMC themselves) but now it has to be Samsung because...... Twitter....?

The same source that said zero consumer Ampere parts? The same source that said no consumer parts until 2021?

Guessing core counts is pretty easy, stating that three governments, three multinational corporations and every semiconductor news outlet in the world is spreading lies to mask nVidia's launch because some guy on Twitter disagrees.....

Speculating used to revolve around using facts to make educated guesses, not hitching your cart to some random guy on Twitter and putting blinders on when reality says he's wrong.
Fine, I'll bite. Ugh why do I do this to myself.

1. Nvidia are far from the only company ordering wafers at TSMC or ramping orders.

2. All wafers allocated to Huawei that are currently in production will continue to be in products. It's future orders that have been killed off. That wafer supply isn't going to shift overnight. That effectively means that Huawei still has several months of wafers getting churned out from TSMC which will not be affected, especially as companies like this book wafer supply months in advance of when they can even begin HVM.

3. TSMC did not see "serious" negative impacts from Covid. It's my understanding that absolutely nothing has changed for them and that they're going as strong as ever - but I've not been following TSMC's financials etc all that well. Can you provide some proof to back up that claim?

4. If Nvidia have actually chosen 8LPP for consumer GPUs, then additional wafer supply is rather meaningless for everything short of a mid-generational refresh (which we probably won't see because Nvidia would probably be better off trying to accelerate their next uArch as much as possible instead). Reason being is you need tapeouts on the other node and to perform validation to ensure everything works, and only once that is done can you begin HVM.

5. What's this statement from TSMC about 'a large chunk of the slack' being picked up by Nvidia you talk about? Literally never heard that one before.
 

beginner99

Diamond Member
Jun 2, 2009
5,221
1,594
136
I doubt a 3070ti will be offered alongside a 3070. Maybe down the road when they need to spice up the mid-range segment

I think they will be offered at the same time. If that leak is true, the main take away is that the 3070 and 3070 Ti are based on the GA104 chip. Meaning a 3080 will have trouble fitting in there as well so will most likely be GA102.
This at least gives a change for steady or even lower prices (per xx70/xx80 model) and chance of some actual relevant performance/$ increases.

AMD is ultimately in the same boat. Will be dififuclt to charge $300-$400 for a CPU in a $500 console. Albeit I'm still expecting $600 launch prices for consoles.
 

BenSkywalker

Diamond Member
Oct 9, 1999
9,140
67
91
1- The sky is also blue.

2-

Of course it takes months, but it will be a done deal by September, that leaves at least all of Q4 with their second highest volume customer gone.

3- https://semiwiki.com/semiconductor-...smc-covid-19-and-double-digit-growth-in-2020/

That was news on every semi site I follow, it was literally everywhere. Compared to their earlier projections it was a large correction. There were at least dozens of articles, you got the first link.

4- This thread, back in the first few pages JHH stated Ampere would be mostly TSMC either in January of this year or the end of last year. This was acknowledged at the time, and then Twitter.

5- https://www.digitimes.com/news/a20200505PD209.html

*Edit* Meant to link this one https://www.digitimes.com/news/a20200423PD202.html

First one was about further out orders, this one is about nVidia and AMD picking up the slack left by others due to COVID */Edit*

That one might be behind a pay wall, again though, was covered widely.

Actual news, not tweets from anonymous sources. The amount of faith put into Twitter is utterly bizarre when we know what they are stating is wrong simply by following the actual industry news instead of the silly rumors from anonymous people.

They've thrown out so many different configurations at this point one of them will likely be correct and people will act like it's because of insider information or sources when the most basic elements of the claims we already know are wrong.
 
Last edited:

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
1- The sky is also blue.

2-

Of course it takes months, but it will be a done deal by September, that leaves at least all of Q4 with their second highest volume customer gone.

3- https://semiwiki.com/semiconductor-...smc-covid-19-and-double-digit-growth-in-2020/

That was news on every semi site I follow, it was literally everywhere. Compared to their earlier projections it was a large correction. There were at least dozens of articles, you got the first link.

4- This thread, back in the first few pages JHH stated Ampere would be mostly TSMC either in January of this year or the end of last year. This was acknowledged at the time, and then Twitter.

5- https://www.digitimes.com/news/a20200505PD209.html

*Edit* Meant to link this one https://www.digitimes.com/news/a20200423PD202.html

First one was about further out orders, this one is about nVidia and AMD picking up the slack left by others due to COVID */Edit*

That one might be behind a pay wall, again though, was covered widely.

Actual news, not tweets from anonymous sources. The amount of faith put into Twitter is utterly bizarre when we know what they are stating is wrong simply by following the actual industry news instead of the silly rumors from anonymous people.

They've thrown out so many different configurations at this point one of them will likely be correct and people will act like it's because of insider information or sources when the most basic elements of the claims we already know are wrong.
Man. If you say the sky is blue, then please also say that chips that go into production in Q4, don't come to market before the start of Q2 2021.
 

Glo.

Diamond Member
Apr 25, 2015
5,758
4,666
136
4- This thread, back in the first few pages JHH stated Ampere would be mostly TSMC either in January of this year or the end of last year. This was acknowledged at the time, and then Twitter.
Jenhsen was specifically asked about 7 nm product orders, not about ALL of their next gen products.

And he said SPECIFICALLY that 7 nm products will be handled in majority by TSMC, and in small portion - Samsung.

Nobody has asked Jenhsen about 8 nm from SS.

So in this sense: GA100 - 7 nm TSMC - correct, which we already have seen. Anything below GA100 - 8 nm Samsung.
 

uzzi38

Platinum Member
Oct 16, 2019
2,698
6,393
146
1- The sky is also blue.

2-

Of course it takes months, but it will be a done deal by September, that leaves at least all of Q4 with their second highest volume customer gone.

3- https://semiwiki.com/semiconductor-...smc-covid-19-and-double-digit-growth-in-2020/

That was news on every semi site I follow, it was literally everywhere. Compared to their earlier projections it was a large correction. There were at least dozens of articles, you got the first link.

4- This thread, back in the first few pages JHH stated Ampere would be mostly TSMC either in January of this year or the end of last year. This was acknowledged at the time, and then Twitter.

5- https://www.digitimes.com/news/a20200505PD209.html

*Edit* Meant to link this one https://www.digitimes.com/news/a20200423PD202.html

First one was about further out orders, this one is about nVidia and AMD picking up the slack left by others due to COVID */Edit*

That one might be behind a pay wall, again though, was covered widely.

Actual news, not tweets from anonymous sources. The amount of faith put into Twitter is utterly bizarre when we know what they are stating is wrong simply by following the actual industry news instead of the silly rumors from anonymous people.

They've thrown out so many different configurations at this point one of them will likely be correct and people will act like it's because of insider information or sources when the most basic elements of the claims we already know are wrong.

Fair enough on the Huawei point, I didn't read into the situation too much and didn't realise all shipments past Sept 15th where off the table.

However, COVID has had little to no impact on TSMC and the article you linked even states that:

"Looking ahead to the second half of this year. Due to the market uncertainty, we adopt a more conservative view as we expect COVID-19 to continue to bring some level of disruption to the end market demand. For the whole year of 2020, we now forecast the overall semiconductor market, excluding memory growth, to be flattish to slightly decline, while foundry industry growth is expected to be high single-digit to low-teens percentage.”

Flattish to slightly decline does not imply a "serious" hit. It implies a very minor one. COVID has had little to no effect on TSMC.

Digitimes articles are not reliable sources at the moment given several of their most recent ones with Zen 3 and 5nm. Those can be taken with a huge grain of salt.

But regardless, you originally stated a large chunk of the orders were made up by Nvidia who supposedly where always planning to use N7 but also didn't have enough wafers? Your point really actually doesn't make much sense if I'm honest, because you're simultaneuously trying to argue that Nvidia were both always planning to use N7, but at the same time are picking up a huge number of additional (read: more than the thousands they originally first booked) N7 wafers therefore they must be using Ampere for consumer dies. I don't get how that makes sense.

The two can't be related to one another - you do realise that, right? Nvidia picking up very significant amount of additional N7 wafers would be an effect of not having enough in the first place and planning on dual-sourcing more than normal or alternatively responding to a rapidly growing market that's shown immense growth as of late with additional volume. Like for example, the datacentre market.

Actually, Jensen's comment
 

Glo.

Diamond Member
Apr 25, 2015
5,758
4,666
136
The reason why people put faith into Twitter news, specifically kopite7Kimi's, words is because HE WAS CORRECT PREVIOUSLY when it goes to Nvidia upcoming products.
 

DeathReborn

Platinum Member
Oct 11, 2005
2,755
751
136
Jenhsen was specifically asked about 7 nm product orders, not about ALL of their next gen products.

And he said SPECIFICALLY that 7 nm products will be handled in majority by TSMC, and in small portion - Samsung.

Nobody has asked Jenhsen about 8 nm from SS.

So in this sense: GA100 - 7 nm TSMC - correct, which we already have seen. Anything below GA100 - 8 nm Samsung.

Just a quick one, if the majority of 7nm is TSMC but everything else is 8nm, where is the minority of 7nm coming from or is it 100% of 7nm is TSMC which is not what he said. This relates to GA100 - TSMC & everything else 8nm in your post btw.
 

Glo.

Diamond Member
Apr 25, 2015
5,758
4,666
136
Just a quick one, if the majority of 7nm is TSMC but everything else is 8nm, where is the minority of 7nm coming from or is it 100% of 7nm is TSMC which is not what he said. This relates to GA100 - TSMC & everything else 8nm in your post btw.

He said specifically that MAJORITY of 7 nm products are coming from TSMC. MAJORITY. Which means that there might be also 7 nm products from Samsung. What it is - only he knows.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |