NVIDIA GOES 12nm

Maximum PC - - GAME CHANGER -

Much of Tur­ing’s ad­vance­ments are thanks to a die shrink, part of the never-end­ing quest to be smaller and faster. Tur­ing GPUs are man­u­fac­tured us­ing TSMC’s 12nm FinNET process. In re­al­ity, it is more of a re­fine­ment of the pre­vi­ous 16nm process than a mean­ing­ful re­duc­tion of fea­ture sizes. Wit­ness the fact that the TU104 is only slightly smaller than the GP102 used in the GTX 1080 Ti, and the TU106 has 10.8 bil­lion tran­sis­tors, against a GP102’s 12 bil­lion. De­spite mov­ing to 12nm, they re­ally aren’t much smaller at all. How­ever, it is a ma­ture process, with ex­cel­lent yields, and it should read­ily lend it­self to faster clock speeds.

The change that will help boost things is a move to 7nm. TSMC is close to full pro­duc­tion with this, and AMD’s next GPU is sup­posed to be out and about be­fore the end of the year. If this goes well, we hope to see a more ef­fec­tive shrink of Tur­ing to 7nm. This will get die sizes down to more man­age­able lev­els, and en­able more cores to be added, and maybe more RT units to re­ally push the ray trac­ing.

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.