Thinking On Paper GPU meets quantum computer: Nvidia and Infleqtion's four-microsecond bridge to hybrid computing
Mar 8, 2026
Sam Stanwyck, NVIDIA product lead uniting GPUs with quantum processors. Pranav Gokhale, Infleqtion CTO building neutral-atom QPUs and logical qubit roadmaps. They unpack NVQ-Link, real-time GPU–QPU co-processing, software that routes workloads, a space-elevator battery workflow, and the 2028 push toward 100 logical qubits.
AI Snips
Chapters
Transcript
Episode notes
NVIDIA Sees Quantum As Supercomputer Augmentation
- NVIDIA views quantum as a complementary accelerator, not a replacement, extending supercomputers to new problem classes.
- Sam Stanwyck explains NVQ-Link as a low-latency, high-bandwidth interconnect to tightly couple QPUs with GPU supercomputers.
NVQ-Link Enables Real Time GPU–QPU CoProcessing
- NVQ-Link achieves ~4 microsecond round-trip latency enabling real-time co-processing between GPU and QPU.
- Pranav Gokhale highlights that this latency lets GPUs orchestrate and decode quantum results within the same clock cycles as the QPU.
Software Orchestrates Quantum Versus Classical Tasks
- Software decides workload routing in real time using low-latency logic, not humans.
- Pranav describes a two-level stack: pre-run GPU simulation (CUDAQ) plus fast runtime switching for decoding and next-step control.