3 ms·
Read the article. It's about ML workloads which scale well across many cores. It's also being compared to GPUs. The whole point of what they're doing to is t
by Mikeb85 5y ago
Read the article. It's about ML workloads which scale well across many cores. It's also being compared to GPUs. The whole point of what they're doing to is to be able to pack more cores versus a CPU but with a larger instruction set than a GPU core.