We’ve haven’t heard that much about NVIDIA’s Pascal GPU since it was taped out back in June. According to many sources the Pascal GPU will be manufactured on TSMC’s 16nm FinFET node. This means that Samsung has failed to win the contract for the latest GPU. The NVIDIA Pascal GPU will be the latest graphics architecture from the company and will feature HBM2, NVLINK, and Mixed Precision Support.
According to a recent report NVIDIA has taped out the silicon for its upcoming Pascal GPU architecture, codenamed GP100. This chip will be the successor to the GM200 silicon that is currently being used in high-end graphics cards like the Titan X. A tape-out is when a company successfully produces a tiny quality of working prototypes for internal testing and further development.
A foundry partner is always important for companies like NVIDIA. TSMC has always given priority to Apple for orders, but it looks like NVIDIA will still be retaining the company as their “primary” foundry partner. NVIDIA also has Samsung as a foundry partner, but it remains unclear which products will be fabricated by TSMC and which ones by Samsung.
We first heard about AMD’s Greenland GPU when we details AMD’s upcoming 16-core APU. Greenland will be the successor to Fiji (which will be powering the upcoming R9 390X). It has been reported that this new core will not be a radically new core. AMD wants the main goal of Greenland to bring more performance per watt to AMD GPUs in 2016.
Late last year we reported that AMD had cut ties with TSMC and would be manufacturing their GPU and APU products at Global Foundries. NVIDIA still uses TSMC to manufacture their chips, but recently they have moved to Samsung for 14 nm production. TSMC has just announced that they will not make any 20 nm GPUs at all, if anything 20 nm comes out of TSMC it will be low power ASICs.
According to a new report from WCCFTech, both AMD’s and NVIDIA’s upcoming next-generation GPUs could be delayed, at least the 16 nm or 20 nm versions. For the past 3 years the GPU industry has been stuck on the 28 nm process and while 2014 did bring in TSMC’s 20 nm process much of the early 20 nm process capacity has been given to major clients like Qualcomm and Apple.
In our previous post we talked about NVIDIA working with TSMC on the 16 nm FinFET process. We wondered why AMD is was not the released list of companies who are working with TSMC. Well it looks like AMD will manufacture their GPUs at Global Foundries.
It looks like TSMC is ready to move to volume production on their 16 nm FinFET process. Based on a recent report we know that NVIDIA is joining them. Does that mean that NVIDIA is skipping the 20 nm fab process completely?
There have been many rumors circulating around about upcoming Apple devices, but not that much has been said about the processor that will be powering these devices. Apple’s A8 Processor, which will power these devices will is said to have a clock speed of 2 GHz or higher.