Optimized kernels such as FlashAttention and FlashDecoding are crucial for accelerating today's large models. Most of them are handwritten by experts because existing ML compilers cannot match their efficiency. Producing such kernels requires fusing computations with multiple reductions, which requires both algebraic t...
David Pissarra, Jinkun Lin, Hai-Tian Jiang et al.· 0 citations
An extension to the compilation process with a post-verification optimization phase to address the eBPF verifier limitation and measure the performance overhead introduced by these restrictions.
Farbod Shahinfar, Aurojit Panda, Gianni Antichi· Proceedings of the 4th Works...· 0 citations
Beeswax overlaps the execution of program stages, and thus hides cache miss overheads, and is applied to Katran, a load balancer deployed in production data centers, and BMC, a key-value store accelerator.
Farbod Shahinfar, Marco Molè, Aurojit Panda et al.· Conference on Applications,...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.