NVIDIA Blackwell B200 GPU (SXM5)

Brand: NVIDIA | Category: GPUs

SKU: NVIDIA-900-21001-0000-000 | Part #: 900-21001-0000-000 | MPN: 900-21001-0000-000

Contact for Pricing — Request a Quote

The NVIDIA Blackwell B200 GPU in SXM5 form factor delivers enterprise-grade performance with passive cooling and liquid-cooled backplate technology. This accelerator is engineered for mission-critical workloads requiring maximum computational density and memory bandwidth. The B200 provides 192 GB of HBM3e memory, enabling processing of exceptionally large datasets without frequent host transfers. With a memory bandwidth of 19.1 TBps, the architecture sustains high-throughput data flow essential for complex algorithms. Peak FP64 performance reaches 44.1 TFLOPs, while peak TF32 performance delivers 1408 TFLOPs, positioning this GPU for both traditional HPC and modern AI inference and training applications. The SXM5 interconnect includes 12x NVLink 4.0 channels, establishing low-latency, multi-GPU communication pathways within a single system. Operating at 1000 W thermal design power, the B200 is cooled via passive design with a liquid-cooled backplate, reducing active component failure modes in data center environments. NVIDIA backs this accelerator with a standard 3-year manufacturer warranty, with extended coverage available upon request.

Enterprise AI infrastructure teams, HPC systems integrators, and data center architects select the B200 for workloads spanning large language model inference, scientific simulation, and real-time analytics at scale. The SXM5 form factor integrates seamlessly into NVIDIA-certified server platforms, while the 12V-2x6 / 16-pin server-class power connector ensures compatibility with standard enterprise power delivery architectures. Operating temperature range of 0–85 °C accommodates diverse thermal management strategies, from air-cooled to fully immersive cooling deployments.

Omnixon Global stocks the NVIDIA Blackwell B200 GPU (part number 900-21001-0000-000) for immediate quotation and delivery across the UAE and broader MENA region. Our technical team can validate integration requirements, confirm platform compatibility, and expedite procurement for large-volume deployments. To request a quote or discuss your GPU infrastructure needs, please submit an RFQ through our Omnixon Global portal.