GQ
Groq develops LPU inference systems focused on deterministic low-latency serving for large language models.
Tag
Inference
AI inference chips, serving platforms, runtime optimization, and deployment resources.
Tagged Resources
H100
Hopper-generation data center GPU widely used for AI training, inference, and HPC workloads.
H200
Hopper refresh with larger HBM3e memory footprint for memory-bound inference and training tasks.
Submit a resource
Help keep AIChipNav accurate by suggesting AI chip, GPU cloud, benchmark, policy, or semiconductor resources.