Product presentations, technical data, availability information and price indications are non-binding. Final specification, price, delivery time and availability are confirmed individually in the quotation.
CUDA, HPC, scientific computing, FP64-intensive workloads, legacy AI training and research clusters.
A cost-effective enterprise accelerator with HBM2 memory, strong double-precision performance and a mature CUDA software ecosystem.
Requires directed, high-static-pressure server airflow because the PCIe model is passively cooled. It has no display outputs and should be deployed only in a validated server chassis.
Get in touch with us now. Our infrastructure experts will review your inquiry and reach out to you personally within 24 hours.
