Comment on Deep Learning: Our Miraculous Year 1990-1991parentComments−tlb6yNot for clock speed, but yes for parallelism. It might look like the Cerebras [0] wafer-scale monster becoming a commodity you could fire up 1000 of in the cloud.[0] https://www.cerebras.net/
Comments
Not for clock speed, but yes for parallelism. It might look like the Cerebras [0] wafer-scale monster becoming a commodity you could fire up 1000 of in the cloud.
[0] https://www.cerebras.net/