Tuesday, December 8th 2020
GPU Shortage Hits Data Centers: NVIDIA A100 GPU Supply Insufficient
GPU supply has been one of the most interesting things this year. With a huge demand for the new GPU generations like NVIDIA's Ampere and AMD's RDNA 2 "Big Navi" graphics cards, everyone is trying to grab a card for themselves. Besides the huge demand, there is also a big problem. The supply of these GPUs is just too low to satisfy the demand, driving up the prices, and increasing the scarcity of them. Companies like NVIDIA have their priorities set: all of the major production will go for the data center expansion and data center customers. However, even that plan is proving not to be good enough.
The scarcity of GPUs has now hit data centers, with NVIDIA unable to satisfy the demand for its A100 GPUs designed for high-performance computing. "It is going to take several months to catch up some of the demand," said Ian Buck, vice president of Accelerated Computing Business Unit at Nvidia. That is an indicator of just how huge the demand for these accelerators is. With the recent company announcement of A100 GPU with 80 GB memory, partners expect to have the first cards in their systems in the first half of 2021. That means that this situation and inadequate supply will hopefully resolve sometime around that timeframe.
Source:
Tom's Hardware
The scarcity of GPUs has now hit data centers, with NVIDIA unable to satisfy the demand for its A100 GPUs designed for high-performance computing. "It is going to take several months to catch up some of the demand," said Ian Buck, vice president of Accelerated Computing Business Unit at Nvidia. That is an indicator of just how huge the demand for these accelerators is. With the recent company announcement of A100 GPU with 80 GB memory, partners expect to have the first cards in their systems in the first half of 2021. That means that this situation and inadequate supply will hopefully resolve sometime around that timeframe.
15 Comments on GPU Shortage Hits Data Centers: NVIDIA A100 GPU Supply Insufficient
Golden business right now.
China (!) has tried and latest news are about them kind off failing.
So... golden :DYou are looking at this wrong. By installed wafer capacity (and at current rate, all that is used) Samsung is still largest foundry by a bit over TSMC, followed by Micron, SK Hynix and probably a couple others in front of Intel.
TSMC should be biggest by foundry revenue - a metric which AFAIK usually has own chips excluded so a bad picture. No idea what the 7/5nm share on Samsung is but I would expect TSMC to slightly lead on revenue/profit due to better and pricier 7/5nm manufacturing.
This is where China is going now. By the time you're done copying all the minutia, you're old news. What TSMC is doing with EUV right now is unique in the world, and its not just ASML machinery but lots of very smart heads that get together and a supply chain that works in perfect harmony as well. Every tiny piece needs to check out.
EUV is at the border of make-or-break of 7nm class manufacturing processes but it is by far not the only hurdle to deal with.
www.anandtech.com/show/15803/tsmc-build-5nm-fab-in-arizona-for-2024
And this facility is supposed to push out some 20k wafers a month, a rather low number considering that TSMC has another facility in Taiwan that can push some 100k wafers a month.I think the US does have some people that are more than capable of doing those jobs, but they're probably already taken by Intel or others. There is also the matter of relocating them and screening them (to protect TSMC's IP and the IP of any potential work for the US military), but that's a lesser concern.
Its not like they disappeared. IBM / GloFo just couldn't compete with the prices and fell behind. They're still making 14nm chips and 12nm chips.