Search papers, labs, and topics across Lattice.
2
0
2
3
Remote HBM traffic can vary by 90x for the same GEMM dimensions, and a 2D block-swizzle traversal can slash this traffic by over 5 times.
Achieving up to 24.7x reduction in remote HBM traffic for LLM GEMM tasks could redefine efficiency standards in multi-chiplet GPU architectures.
