Better Hardware Could Turn Zeros into AI Heroes
Posted by IEEESpectrum@reddit | hardware | View on Reddit | 3 comments
Researchers from Stanford use sparsity to create an AI chip that, on average, consumed one-seventieth the energy of a CPU, and performed the computation on average eight times as fast.
JuanElMinero@reddit
CGRAs seem like a fascinating development, unfortunately that title doesn't do the topic any justice.
anifail@reddit
they are hardly new and have been a favorite research topic at stanford for decades. What's "novel" here is the additional control architecture for augmenting between dense/sparse compute domains.
chapstickbomber@reddit
It is kind of ironic that meat based analog processors (humans) built silicon based digital processors (microchips) to run analog weighted processing more like their own meat (AI models). Removing the digital environment is going to improve weight based processing by orders of magnitude.