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


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 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.

In fact one additional argument in the favor of Volta in 2017 is the fact that the Pascal big chip has been designed for double precision as well as single precision unlike Maxwell, which was designed from the ground up for single precision. While this is invaluable for Nvidia's other ventures, the double precision units are a waste of space on the P100 die as far as gaming is concerned.  The Geforce TITAN X only has 3584 cores and based on a smaller die, which naturally means we haven't seen the biggest possible chip for consumer purposes on this node yet.

One thing is for sure, the 16nm FinFET process can do much better than a single precision core count of 3584 and until Nvidia has fully utilized this node, I very much doubt it will move on. So it is almost a given that there is at least one more generation of 16 FF GPUs coming from Nvidia, and it is a fair bet that this will be the Volta architecture since this only leaves behind the question of nomenclature.

Nvidia can choose to either brand the new 16nm architecture, Pascal 2nd Generation, or Volta

Considering the fact that the architecture will probably be shipping with HBM2 memory, it makes sense to differentiate it from Pascal. In many ways Pascal architecture was similar to Maxwell (albeit on the 16nm FF node) which raises the question whether Volta can be to Pascal as Maxwell was to Kepler. Allusion to Intel’s former Tick-Tock schema is probably justified here. Volta architecture will perfect the 16nm node and considering Nvidia has already accelerated its road map with the GP102, which we were expecting to be released in Gamescom this year, this doesn't really come as surprising. We know almost nothing about Volta at this moment, except for the fact that it will probably be rocking HBM2 memory and will build upon the Async Compute lessons learned by Nvidia during this round.

The Geforce TITAN X can claim to be the first fully 4K 60 capable card on the market yet it is clearly targeted to the prosumer market. With Volta, Nvidia could finally put an enthusiast level card that will match or beat the TITAN-X, conquering a plateu that has as yet not been scaled by either chip maker. The Volta based successors to the Pascal GPUs will be called the GV 100, GV104 and GV 106/107. Assuming of course, that Nvidia does not decide to brand them Pascal second generation.