Dylan Patel on the GPU Shortage, Nvidia and the Deep Learning Supply Chain
The Inside View
00:00
The Acceleration of GPUs in Clusters
Since November, people have been watering GPUs and that's like accelerated as money has flooded in. It takes three plus months to just get the chip back from TSMC, let alone put it into a data center. The equipment supply chain for not just making the chips but actually packaging them together is one of the biggest bottlenks. HBM memory comes from Korean companies SKynix and Samsung; they dominate the market.
Transcript
Play full episode
