Qinggang Zhou

Senior Staff Tl Tlm, Waymo Ml @Waymo

Sunnyvale, CA, US
MOBILE NUMBERS
+14•••••••33

Signup · Get unlimited contacts

WORK HISTORY

Jan 2023 — Present

Senior Staff Tl Tlm, Waymo Ml @Waymo

View department →

TL/TLM of the Runtime & Optimization team. Our team is responsible for the reliability and performance cross the model cycle, from model efficiency, quantization, sparsity, custom kernel libraries, MLIR/graph optimization, XLA, Triton, runtime, export, serving.* Spearheaded a transition in ML optimization strategy from late-stage compiler tuning to a proactive, holistic system-level approach. Transformed the Runtime & Optimization organization into a Center of Excellence.* Catalyzed \'0-to-1\' initiatives for Foundation Model acceleration, targeting both autoregressive and diffusion decoding. Achieved d a 1.4x speedup through speculative decoding and 2x through FFN sparsity.

EDUCATION

N/A

University of Washington

MSEE, EE

SKILLS

Embedded SystemsSemiconductorsOpenglField-Programmable Gate Arrays (Fpga)Device DriversArmDigital Signal ProcessorsDebuggingAlgorithmsVerilogAsicDigital Signal ProcessingParallel ComputingAndroidArm ArchitectureMpeg-4GpuSocComputer ArchitectureProcessors

ABOUT QINGGANG ZHOU

ML Systems Leader with a proven track record of spearheading 0-to-1 AI initiatives and orchestrating complex, cross-company collaborations. Expert in the full-stack optimization of training and inference systems, spanning ML hardware (GPU/TPU), kernels, and compilers to ML frameworks. Deeply specialized in the system implications of ML workloads, including the latest models, foundation models, decoder efficiency, and diffusion models across both hyperscale data centers and edge environments.

This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.

Qinggang Zhou — Email, Phone Number & Contact Info | Unifers