Nvidia Potential Roadmap Update for 2017: Volta Architecture Could Be Landing As Early As 2H 2017

Gregster · 24 Jul 2016 at 21:12

Something pretty interesting has been making the rounds lately. Word around the grapevine is, the Volta GPU is going to be landing one year early in May next year at the GTC event held annually by Nvidia. There are two particular sources in play on this report, both as critical to the authenticity of this information as the other. Before we go any further, I would like to point out that according to the official roadmaps, Nvidia has revealed to the public so far, Volta architecture was originally scheduled for release in 2018. That said however, I believe there has been a significant update.
NVIDIA Volta GPU
Nvidia launching the 16nm FinFET based Pascal successor on GTC on 8-11th May 2017 ?
Nvidia CEO Jen Hsun Huang has previously stated that every GPU Technology Conference needs to involve a new GPU but as far as Volta was concerned – this was going to be impossible. Volta was originally designed for a process node that was smaller than the 16nm FinFET (namely 10nm FinFET TSMC), a node that was not going to be anywhere near maturity for large scale GPUs by mid-2017. This was of course the primary reason why everyone assumed we will not be seeing Volta by GTC 2017 next year.
So why am I still writing bout this? Well, the reason I think this report might actually turn out to be true is because though the Volta GPU was originally slated for release in 2018, a report recently published by Fudzilla claimed that Nvidia is giving the 16nm FinFET treatment to the Volta architecture. This of course, is something that completely changes the dynamics of Nvidia’s roadmap for 2016 and 2017. If Volta will indeed be manufactured on the 16nm FinFET node than we can actually expect to see it by May 2017 – which is less than a year away.

Read more: http://wccftech.com/nvidia-roadmap-2017-volta-gpu/#ixzz4FMJfyouX

It wouldn't surprise me to see Volta that early in truth and NVidia seem to be flying GPUs out quicker than we can get them in our machines

kalniel · 24 Jul 2016 at 21:15

Have they given up on consumer pascal with HBM2 then?

tommybhoy · 24 Jul 2016 at 21:18

Hope this one can do proper hardware Async then we can get more DX12 titles.

Dirk Diggler · 24 Jul 2016 at 21:21

Between Nvidia ripping it with price premiums and the 10% Brexit tax (thanks, little Englanders :rolleyes:

) ... I'm seriously thinking of sticking with my 980ti and sitting Pascal out. The bang for buck is abysmal at the minute.

If Volta really is due next year then the relatively small wait would be worth it for a proper upgrade. Now, all I have to do is stay away from the MM when the 1080's are getting dumped for the new Titan.

Jono8 · 24 Jul 2016 at 21:23

Could it be just a low range card first though ( like the 750ti)

spikerules · 24 Jul 2016 at 22:04

String said:
Between Nvidia ripping it with price premiums and the 10% Brexit tax (thanks, little Englanders ) ... I'm seriously thinking of sticking with my 980ti and sitting Pascal out. The bang for buck is abysmal at the minute.

If Volta really is due next year then the relatively small wait would be worth it for a proper upgrade. Now, all I have to do is stay away from the MM when the 1080's are getting dumped for the new Titan.

I'm going to do just that I think!

Orangey · 24 Jul 2016 at 22:18

wccf are proven liars that make stuff up constantly.

Orangey · 24 Jul 2016 at 22:21

String said:
thanks, little Englanders

You know you're not on buzzfeed now right?

Not everyone shares your political opinions. So please don't force those of us who don't to read them on a GPU forum.

tommybhoy · 24 Jul 2016 at 22:39

Orangey said:
wccf are proven liars that make stuff up constantly.

Like these guys???

Greebo · 24 Jul 2016 at 22:42

kalniel said:
Have they given up on consumer pascal with HBM2 then?

Whoever said they were ever doing consumer pascal with hbm2? Nvidia never did.

Just geeks on forums thought they would. As I have said many times before, pascal doesn't need it.

Xplo · 24 Jul 2016 at 22:43

I can see this being true you know.

10nm is bloody ages away in the GPU world. Look how long 16nm took, 10nm will be 2018 and thats even if it doesnt run into any problems, which is likely as the node gets smaller.

Nvidias hardware async is pretty much non-existant, their software implementation is getting spanked by AMD with lesser cards, and DX12 is starting to rear its head. For and by a launch of next May DX12 is the hot topic, and nvidia just won't be able to compete, especially when Vega gets here. They can't afford to wait until 2018 and beyond to put this right

So they pull forward the launch of Titan Pascal, a quiet launch and with no HBM. They redesign Volta onto 16nm and pull this forward to May 2017 with HBM.

Bobs ya uncle

Wucked · 24 Jul 2016 at 22:47

Interesting. The internal monologue for skipping gen grows stronger.

Rroff · 24 Jul 2016 at 22:47

Xplo said:
I can see this being true you know.

10nm is bloody ages away in the GPU world. Look how long 16nm took, 10nm will be 2018 and thats even if it doesnt run into any problems, which is likely as the node gets smaller.

Nvidias hardware async is pretty much non-existant, their software implementation is getting spanked by AMD with lesser cards, and DX12 is starting to rear its head. For and by a launch of next May DX12 is the hot topic, and nvidia just won't be able to compete, especially when Vega gets here. They can't afford to wait until 2018 and beyond to put this right

So they pull forward the launch of Titan Pascal, a quiet launch and with no HBM. They redesign Volta onto 16nm and pull this forward to May 2017 with HBM.

Bobs ya uncle

There were a lot of challenges (both in the development of the node and for the companies designing product using the process) in the transition below 28nm (especially in regard to finfets) that won't be the case for the move to 10nm. I believe the first load of tape outs of 10nm test products are about kicking off so we are potentially (though with TSMC it doesn't pay to hold your breath) going to see it sooner rather than later.

EDIT: Going by what we've seen so far AMD would have to make an immense jump with Vega to offset what nVidia could potentially do performance wise via brute force with Pascal if their back was to the wall even assuming for arguments sake they had no async at all.

shankly1985 · 24 Jul 2016 at 22:50

tommybhoy said:
Like these guys???

killswitch73 · 24 Jul 2016 at 23:02

Very interesting and making want to buy that 3rd cheap 980ti and put it under water for 4K goodness . Scaling is fine I think as I had 3x 980s and that was fine at scaling in some games , some games I had to turn off SLI namely Advanced Warfare !

Boomstick777 · 24 Jul 2016 at 23:12

Possible the next X80 card could match / surpass the new Titan X then. That's not long away at all.

Might just hold off until next year then. Still can't decide on a 4K monitor and this RX480 will last me until I do.

no_1_dave · 24 Jul 2016 at 23:12

Could explain why we have Titan coming out already.

1080ti by Christmas and Volta next year...

NirK · 24 Jul 2016 at 23:15

Xplo said:
their software implementation is getting spanked by AMD with lesser cards, and DX12 is starting to rear its head. For and by a launch of next May DX12 is the hot topic, and nvidia just won't be able to compete, especially when Vega gets here. They can't afford to wait until 2018 and beyond to put this right

Not sure what planet you live on but the 1080 and 1070 are faster than anything from AMD, including anything with Async compute.. pascal does get some gains from async compute, about 8% in 3Dmark timespy, but yes I agree that they will likely release volta early as they need to update architecture for full DX12 support... either way they are in no danger at all of "not being able to compete"

Griffildur · 24 Jul 2016 at 23:19

NirK said:
Not sure what planet you live on but the 1080 and 1070 are faster than anything from AMD, including anything with Async compute.. pascal does get some gains from async compute, about 8% in 3Dmark timespy, but yes I agree that they will likely release volta early as they need to update architecture for full DX12 support... either way they are in no danger at all of "not being able to compete"

can't say I am surprised, fully expected an architectural update to deal with all this Directx 12 and Vulkan stuff. NVidia won't fall behind especially now.

Greebo · 24 Jul 2016 at 23:29

NirK said:
Not sure what planet you live on but the 1080 and 1070 are faster than anything from AMD, including anything with Async compute.. pascal does get some gains from async compute, about 8% in 3Dmark timespy, but yes I agree that they will likely release volta early as they need to update architecture for full DX12 support... either way they are in no danger at all of "not being able to compete"

Potentially if big vega is time as fast as a 480 and with the lead amd have with a sync and dx12, it could easily beat the 1080 and nvidia won't be happy about not having the fastest card for a year.