newsroom
AI Infrastructure · GPU power-performance curve saturates, creating optimization opportunity for memory-bound inference
now playing · AI Infrastructure
GPU power-performance curve saturates, creating optimization opportunity for memory-bound inference
The power-performance relationship in GPU clusters is non-linear and saturates because inference workloads are memory-bound — decode stages consume power reading weights into HBM while SMs…
Data centers can unlock 100 GW flexible power by becoming grid-responsive
Today's data centers are static power consumers; by making AI workloads curtailable during peak grid periods without violating SLOs, operators can access 100 GW of flexible power capacity i…