NVIDIA A100 PCIe 80GB HBM2e GPU — Dell PowerEdge R750xa Add-in Card (Legacy)
Brand: Dell | Category: GPUs
SKU: 900-21001-0000-200 | Part #: 900-21001-0000-200 | MPN: 900-21001-0000-200
Contact for Pricing — Request a Quote
About the NVIDIA A100 PCIe 80GB HBM2e GPU — Dell PowerEdge R750xa Add-in Card (Legacy)
AI training and inference workloads demand the NVIDIA Ampere architecture's exceptional tensor performance, and the NVIDIA A100 PCIe 80GB HBM2e GPU delivers precisely that capability as a Dell PowerEdge R750xa add-in card. With 6912 CUDA cores, 432 third-generation Tensor Cores, and a massive 80GB HBM2e memory pool connected via 2,039 GB/s memory bandwidth, this accelerator is engineered for organizations running large-scale machine learning, scientific computing, and data analytics at enterprise scale. The passive cooling design integrates seamlessly with Dell PowerEdge R750xa chassis airflow, eliminating the need for dedicated thermal management while maintaining a 300W thermal design power envelope suitable for data center deployment.
Infrastructure teams and AI procurement specialists will appreciate the card's multi-workload flexibility: Multi-Instance GPU (MIG) technology partitions the accelerator into up to 7 isolated instances, enabling efficient resource allocation across concurrent applications. Third-generation NVLink support enables card-to-card connectivity at 600 GB/s bidirectional throughput for distributed training scenarios. Full ECC memory protection safeguards mission-critical computations, while the PCIe Gen4 x16 interface ensures compatibility with Dell PowerEdge R750xa servers. As a legacy platform component (part number 900-21001-0000-200), this add-in card suits organizations with established Dell infrastructure seeking proven, high-performance GPU acceleration. Contact Omnixon Global to request a formal quotation and explore integration options for your enterprise environment.
Key Specifications
- Manufacturer: Dell
- Manufacturer Part Number: 900-21001-0000-200
- GPU Memory: 80GB HBM2e
- Memory Bandwidth: 2,039 GB/s
- CUDA Cores: 6912
- TF32 Tensor Core Performance: 312 TFLOPS (with sparsity: 624 TFLOPS)
- Interface: PCIe Gen4 x16
- Thermal Design Power: 300W
- Compatible Server Platform: Dell PowerEdge R750xa
- NVLink Support: Third-generation NVLink (up to 600 GB/s bidirectional)
Technical Specifications
| Brand | Dell |
| Category | GPUs |
| SKU | 900-21001-0000-200 |
| Part Number | 900-21001-0000-200 |
| Condition | New |
| Manufacturer Part Number | 900-21001-0000-200 |
| GPU Architecture | NVIDIA Ampere |
| GPU Memory | 80GB HBM2e |
| Memory Bandwidth | 2,039 GB/s |
| Interface | PCIe Gen4 x16 |
| CUDA Cores | 6912 |
| Tensor Cores | 432 (3rd Generation) |
| FP64 Performance | 9.7 TFLOPS |
| FP32 Performance | 19.5 TFLOPS |
| TF32 Tensor Core Performance | 312 TFLOPS (with sparsity: 624 TFLOPS) |
| FP16 Tensor Core Performance | 624 TFLOPS (with sparsity: 1,248 TFLOPS) |
| INT8 Tensor Core Performance | 1,248 TOPS (with sparsity: 2,496 TOPS) |
| Multi-Instance GPU (MIG) | Up to 7 MIG instances (1g.10gb to 7g.80gb profiles) |
| TDP (Thermal Design Power) | 300W |
| Form Factor | Full-height, full-length (FHFL) PCIe add-in card |
| Cooling | Passive (server-managed airflow via Dell PowerEdge R750xa chassis) |
| Compatible Server Platform | Dell PowerEdge R750xa |
| NVLink Support | Third-generation NVLink (card-to-card, up to 600 GB/s bidirectional) |
| ECC Memory | Yes |
| Display Outputs | None |
Frequently Asked Questions about NVIDIA A100 PCIe 80GB HBM2e GPU — Dell PowerEdge R750xa Add-in Card (Legacy)
What server platforms accept the NVIDIA A100 PCIe 80GB HBM2e GPU — Dell PowerEdge R750xa Add-in Card (legacy)?
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.
How long is the lead time on AI GPUs?
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.
Do you supply matched networking (Quantum InfiniBand / Spectrum-X)?
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.
Can you help with NVIDIA AI Enterprise licensing?
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.