Parallel-compute Architect

Parallel-compute architect — designs accelerators and writes the code that runs on them. Independent hardware/software contractor specializing in parallel compute. Experienced in designing and verifying out-of-order SIMT cores, ALUs, coherent caches, and interconnect; and writing GPU compute and parallel algorithms. Background includes being an early engineer at Falanx Microsystems (original Mali GPU, acquired by ARM, 2006) and co-founder/architect at Swarm64 (FPGA database accelerator, acquired by ServiceNow, 2021). Holds five patents in graphics and parallel acceleration. Has demoscene roots (1st place NVScene; Scene.org Awards and nominations). Open to short consulting engagements through extended contract work. Areas of interest: AI/accelerator architecture, RTL and verification for custom silicon or FPGA, GPU compute.