Brand: NVIDIA | Category: GPUs
SKU: 900-21001-0000-030 | Part #: 900-21001-0000-030 | MPN: 900-21001-0000-030
Contact for Pricing — Request a Quote
The NVIDIA A100 PCIe 40GB HBM2 is a data center GPU built on the NVIDIA Ampere architecture, delivering a substantial generational leap in compute performance for AI training, inference, and high-performance computing workloads. The Ampere architecture introduces third-generation Tensor Cores with TF32 precision, enabling up to 20x the performance of the previous Volta generation for AI workloads without requiring any code changes. The PCIe form factor allows broad compatibility with standard server platforms, making it accessible for enterprises that do not require NVLink-based multi-GPU scaling.
With 40GB of HBM2 high-bandwidth memory delivering 1,555 GB/s of memory bandwidth, the A100 PCIe 40GB handles large model training and complex simulation workloads that demand both capacity and throughput. The GPU supports multi-instance GPU (MIG) technology, which allows a single A100 to be partitioned into up to seven independent GPU instances, each with dedicated compute, memory, and bandwidth resources. This capability dramatically improves utilization efficiency in shared infrastructure environments such as cloud platforms and enterprise AI clusters.
The NVIDIA A100 PCIe 40GB HBM2 (manufacturer part number 900-21001-0000-030) is purpose-built for enterprise data centers and research institutions requiring reliable, sustained performance across diverse AI and scientific computing workloads. It supports NVIDIA NVLink for peer-to-peer GPU communication when paired in compatible systems, and is fully supported under NVIDIA's enterprise software ecosystem including CUDA, cuDNN, TensorRT, and the NVIDIA AI Enterprise software suite, enabling organizations across the UAE, GCC, EMEA, and APAC regions to accelerate their digital transformation initiatives.
| Manufacturer | NVIDIA |
| Manufacturer Part Number | 900-21001-0000-030 |
| GPU Architecture | NVIDIA Ampere |
| CUDA Cores | 6912 |
| Tensor Cores | 432 (3rd Generation) |
| Memory Capacity | 40 GB HBM2 |
| Memory Bandwidth | 1,555 GB/s |
| Memory Interface | 5120-bit |
| Peak FP64 Performance | 9.7 TFLOPS |
| Peak FP64 Tensor Core Performance | 19.5 TFLOPS |
| Peak FP32 Performance | 19.5 TFLOPS |
| Peak TF32 Tensor Core Performance | 156 TFLOPS (312 TFLOPS with sparsity) |
| Peak BFLOAT16 Tensor Core Performance | 312 TFLOPS (624 TFLOPS with sparsity) |
| Peak INT8 Tensor Core Performance | 624 TOPS (1248 TOPS with sparsity) |
| GPU Memory Error Correction | ECC (HBM2) |
| Multi-Instance GPU (MIG) | Up to 7 GPU instances |
| Interconnect | PCIe Gen4 x16 |
| Form Factor | PCIe Full-Height / Full-Length (FHFL) dual-slot |
| Thermal Design Power (TDP) | 250W |
| System Interface | PCIe 4.0 x16 |
| NVLink Support | NVLink Bridge (600 GB/s bidirectional, 2 links) |
Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.
| Brand | NVIDIA |
| Category | GPUs |
| SKU | 900-21001-0000-030 |
| Part Number | 900-21001-0000-030 |
| Condition | New |
| Manufacturer Part Number | 900-21001-0000-030 |
| GPU Architecture | NVIDIA Ampere |
| CUDA Cores | 6912 |
| Tensor Cores | 432 (3rd Generation) |
| Memory Capacity | 40 GB HBM2 |
| Memory Bandwidth | 1,555 GB/s |
| Memory Interface | 5120-bit |
| Peak FP64 Performance | 9.7 TFLOPS |
| Peak FP64 Tensor Core Performance | 19.5 TFLOPS |
| Peak FP32 Performance | 19.5 TFLOPS |
| Peak TF32 Tensor Core Performance | 156 TFLOPS (312 TFLOPS with sparsity) |
| Peak BFLOAT16 Tensor Core Performance | 312 TFLOPS (624 TFLOPS with sparsity) |
| Peak INT8 Tensor Core Performance | 624 TOPS (1248 TOPS with sparsity) |
| GPU Memory Error Correction | ECC (HBM2) |
| Multi-Instance GPU (MIG) | Up to 7 GPU instances |
| Interconnect | PCIe Gen4 x16 |
| Form Factor | PCIe Full-Height / Full-Length (FHFL) dual-slot |
| Thermal Design Power (TDP) | 250W |
| System Interface | PCIe 4.0 x16 |
| NVLink Support | NVLink Bridge (600 GB/s bidirectional, 2 links) |
Reference servers include Dell PowerEdge XE9680 / XE9712, HPE Cray XD670, Lenovo ThinkSystem SR685a / SR675 V3, Supermicro AS-A21GE / SYS-821GE, Gigabyte G593 / G894, ASUS ESC. Share your target platform in the RFQ and we will confirm chassis-to-GPU compatibility and recommended NIC pairing.
Highly model-dependent. L40S / RTX-class: typically 3-6 weeks. H100/H200/B200 in SXM form factor: 12-16 weeks for whole-platform allocations. We quote genuine-channel ETAs only — no grey-market promises.
Yes — Omnixon stocks the full NVIDIA networking lineup (Quantum-2 / Quantum-X InfiniBand, Spectrum-X Ethernet, ConnectX NICs, BlueField DPUs) so we can quote a complete training-cluster BOM, not just the GPUs.
Yes. We hold genuine-channels for NVIDIA AI Enterprise software subscriptions. Add it to your RFQ and we quote node-aligned licensing along with the hardware.