Nvidia thinks "Pascal is just unbeatable" and already decided not to move Volta ahead

[Mira Yurizaki]

Just now, i_build_nanosuits said:

and i see it the other way around...stock clock is the data point and ''average overclock'' is the real indication of how good the hardware really is.

Which is flawed because of the huge amount of variation not only between the quality of manufacturing at the time affecting how good the GPU can get to, but how the board was designed itself. I can't overclock my GPU to any of the, say EVGA Classified cards because my board literally does not have the support circuitry to drive a GPU that hard. All an overclock test does is tell me how good that particular board is. Not how good the GPU is in its entirety across the board.

 

I want to go in knowing I'm guaranteed at least X amount of performance. Not that I have the potential to achieve Y amount of performance. Because the latter's case is basically the same as internet speed marketing.

[i_build_nanosuits]

1 minute ago, M.Yurizaki said:

Which is flawed because of the huge amount of variation not only between the quality of manufacturing at the time affecting how good the GPU can get to, but how the board was designed itself. I can't overclock my GPU to any of the, say EVGA Classified cards because my board literally does not have the support circuitry to drive a GPU that hard. All an overclock test does is tell me how good that particular board is. Not how good the GPU is in its entirety across the board.

 

I want to go in knowing I'm guaranteed at least X amount of performance. Not that I have the potential to achieve Y amount of performance. Because the latter's case is basically the same as internet speed marketing.

we'll just have to disagree then...personally i hate to see the nvidia cards tested at 1650mhz when 95%+ of them can hold 2000mhz stable with the pushing of a single slider in afterburner and 100% of them can do 1900mhz...i think it's miss-representation of actual performance.

[The Benjamins]

I said something like this would be bad for the community on PCper Vega article, and this is what someone responded with

 

zme-ul said

Quote

to the contrary, it's very good for the market because it will be the 1st time in a very long time that GPUs will have a larger life span, and that's very good for the consumer considering the price premiums we had to pay for Pascal

also, nVidia expressed no intention to launch consumer Volta this year - taken from the recent earnings call

 

[i_build_nanosuits]

5 minutes ago, The Benjamins said:

I said something like this would be bad for the community on PCper Vega article, and this is what someone responded with

 

zme-ul said

 

OMG that's the stupidest thing i've read today...''we paid a lot for these old ass cards, so it's good that they keep being relevant longer''

meanwhile...4K 120hz isn't happening, and i can still barely run project cars on medium settings in my oculus rift..

[The Benjamins]

Just now, i_build_nanosuits said:

OMG that's the stupided thing i've read today...''we paid a lot for these old ass cards, so it's good that they keep being relevant longer''

meanwhile...4K 120hz isn't happening, and i can still barely run project cars on medium settings in my oculus rift..

and now that I think about it the name seems familiar :^)

[i_build_nanosuits]

4 minutes ago, The Benjamins said:

and now that I think about it the name seems familiar :^)

...and now nvidia will start making those +5% performance leap per generations, just like intel did for the past 7 years...

Seriously, what we really need is for AMD to GTFO and be replaced by another major corporation that has more budget for R&D and that is not driven by a bunch of drunken monkeys.

[Zodiark1593]

1 minute ago, i_build_nanosuits said:

...and now nvidia will start making those +5% performance leap per generations, just like intel did for the past 7 years...

Seriously, what we really need is for AMD to GTFO and be replaced by another major corporation that has more budget for R&D and that is not driven by a bunch of drunken monkeys.

*Rise of the Matrox*

[dorin]

So, this year Nvidia will release... nothing? Or we will have a Pascal "refresh"?

[The Benjamins]

2 minutes ago, i_build_nanosuits said:

...and now nvidia will start making those +5% performance leap per generations, just like intel did for the past 7 years...

Seriously, what we really need is for AMD to GTFO and be replaced by another major corporation that has more budget for R&D and that is not driven by a bunch of drunken monkeys.

honestly Navi is make or break point, sense it is the GPU that is FULL designed by the new leadership.

If Nvidia does the 5% scheme then if will allow AMD to bounce back like they did in x86 but if Nvidia keeps doing health improvements then it could force AMD out of the GPU marked. (assuming Navi fails)

 

But it is definitely showing that Nvidia doesn't have to work that hard to keep them selves at the top.

[XenosTech]

29 minutes ago, M.Yurizaki said:

Which is flawed because of the huge amount of variation not only between the quality of manufacturing at the time affecting how good the GPU can get to, but how the board was designed itself. I can't overclock my GPU to any of the, say EVGA Classified cards because my board literally does not have the support circuitry to drive a GPU that hard. All an overclock test does is tell me how good that particular board is. Not how good the GPU is in its entirety across the board.

 

I want to go in knowing I'm guaranteed at least X amount of performance. Not that I have the potential to achieve Y amount of performance. Because the latter's case is basically the same as internet speed marketing.

Didn't paul's hardware and Hardware unboxed compare vega 56 to AIB cards and didn't vega 56 match and beat those as well ?

[Dietrichw]

2 minutes ago, XenosTech said:

Didn't paul's hardware and Hardware unboxed compare vega 56 to AIB cards and didn't vega 56 match and beat those as well ?

Paul used an overclocked Galax 1070

[XenosTech]

3 minutes ago, Dietrichw said:

Paul used an overclocked Galax 1070

I know and the 56 kept up and beat it in some games... think steve used a factory oced 1070 too and at the end of his i think  vega was like a couple % faster

[Bananasplit_00]

4 hours ago, Flavio hc 16 said:

even though  under heavy oc it could draw someting retarded like 750w alone).

Go look at a 980TI or R9 Fury under LN2, they go beyond that with ease. 980TIs can blow the 24 pin under LN2 because of the PCIe draw, maybe not alone but definitely if you have 2+ of them lol. Furys are the same. The 1080TI is sadly not as fun... 

 

4 hours ago, Misanthrope said:

It is now and it has been for a while, but no. Some of their past processors were actually quite better than contemporary intel offerings for example. Goes for some generations under ATI as well.

I often wonder what would have happened with the GPU market if AMD never bought ATI... Fun things probably. 

[Bananasplit_00]

3 hours ago, NumLock21 said:

Volta will probably release along side navi.

I'm going to say Volta will win by a few months. I have always said Volta announced at CES 2018, probably will take some time from that to hit the market but I doubt AMD will crank out Navi before then. They might but I don't think they will. 

 

1 hour ago, i_build_nanosuits said:

and i see it the other way around...stock clock is the data point and ''average overclock'' is the real indication of how good the hardware really is.

If that was how reviews were done my 970 would have been returned within a week of purchase. Probably bottom silicon ever put out from the fab, any overclock, even 1mhz, will crash the card as it hits ~60C which it will, even light load will get it there. 

[marldorthegreat]

6 hours ago, Tiwaz said:

then why does it have such a shitty efficiency? 

It doesnt, but when they increase clock speeds they increase power consumption. If you lower the clock speeds, it actually approaches the efficiency of the 1080, but not the performance. 

[Brooksie359]

5 hours ago, Taf the Ghost said:

@snortingfrogs

 

There was one other little thing about Vega that LTT came across that I'm quite curious to see tested more.

 

There's quite a lot of room between the 1080 and the Vega 64 in CS:GO. This matters a lot. I also want to see Dota 2 testing, especially against the 1080 Ti. If Vega 64 = 1080 Ti in two of the most popular games available, suddenly Vega's value completely changes.

 

This needs further testing.

Well I always thought csgo had a 300 fps limit so I don't see this being  that big of a deal. Obviously having higher fps in a game like csgo is beneficial but I had thought the engine had a 300 fps limit.

[Bananasplit_00]

4 minutes ago, Brooksie359 said:

Well I always thought csgo had a 300 fps limit so I don't see this being  that big of a deal. Obviously having higher fps in a game like csgo is beneficial but I had thought the engine had a 300 fps limit.

No you can set the limit to anything you want, set it to 1800fps or something

 and its basically unlimited. 

[Brooksie359]

1 hour ago, i_build_nanosuits said:

we'll just have to disagree then...personally i hate to see the nvidia cards tested at 1650mhz when 95%+ of them can hold 2000mhz stable with the pushing of a single slider in afterburner and 100% of them can do 1900mhz...i think it's miss-representation of actual performance.

Did they disable gpu boost? Because  most will boost higher than that out of the box. Also it is kinda of unfair to compare a reference model of the Vega 56 with a AIB partner card gtx 1070. I mean Vega 56 will have partner cards with better overclocking and cooling  performance  which should compete better against the gtx 1070 partner cards.

[Brooksie359]

1 minute ago, Bananasplit_00 said:

No you can set the limit to anything you want, set it to 1800fps or something

 and its basically unlimited. 

Oh is it just set to 300 by default then?

[tgk300]

Well information has just recently been released stating that the prices of Vega 64 aren't the final prices. It has been revealed by Overclockers UK that the price of Vega 64 (499) is set to increase in the long term to around £599. This does interfere with AMD's price to performance ratio slightly as their Vega 64 is now more expensive for +/- 5% performance making Nvidia's GTX 1080 seem sike the better deal.

 

Read the Article here

http://wccftech.com/amds-rx-vega-64s-499-price-tag-was-a-launch-only-introductory-offer/

[VagabondWraith]

This article paints a different picture:

 

GeForce Volta is coming - Nvidia and TSMC to increase production of 12nm

 

Quote

This week, TSMC, Nvidia’s silicon partner, announced that it will be kicking off production of its 12nm FinFET chips later this year, making Volta-based GeForce graphics cards possible as soon as early 2018.

Quote

So we won’t be seeing any GeForce-series Volta graphics cards in time for the holiday season, but with production of 12nm FinFET ramping up, we could see something in the first half of next year, potentially even before the end of March 2018 but that is speculation at this time.

https://www.kitguru.net/components/graphic-cards/damien-cox/nvidia-and-tsmc-to-increase-production-of-12nm-volta-gpus-later-this-year/

[Misanthrope]

41 minutes ago, VagabondWraith said:

This article paints a different picture:

 

GeForce Volta is coming - Nvidia and TSMC to increase production of 12nm

 

https://www.kitguru.net/components/graphic-cards/damien-cox/nvidia-and-tsmc-to-increase-production-of-12nm-volta-gpus-later-this-year/

Not really a different picture: Article I quoted is Jan saying no consumer volta for holiday 2017. The ramp up could be related to quadro volta sales or just early 2018 release which is past the holiday season

[Dylanc1500]

4 minutes ago, Misanthrope said:

Not really a different picture: Article I quoted is Jan saying no consumer volta for holiday 2017. The ramp up could be related to quadro volta sales or just early 2018 release which is past the holiday season

Could also be the GV100 is ramping from increased demand since they are already implementing them in systems currently.

 

Its to bad they couldn't give specifics as to which silicon is being ramped up.

[Bananasplit_00]

3 minutes ago, Misanthrope said:

Not really a different picture: Article I quoted is Jan saying no consumer volta for holiday 2017. The ramp up could be related to quadro volta sales or just early 2018 release which is past the holiday season

Lines up pretty good with my theory about a CES announcement for Volta though, both articles do actually 

[Bananasplit_00]

1 hour ago, Brooksie359 said:

Oh is it just set to 300 by default then?

Yes. You can just change it in the console