NVIDIA A16-PCIE-64GB Quad-GPU for Dell PowerEdge
Brand: Dell | Category: GPUs
SKU: 900-2G160-0020-000 | Part #: 900-2G160-0020-000 | MPN: 900-2G160-0020-000
Contact for Pricing — Request a Quote
About the NVIDIA A16-PCIE-64GB Quad-GPU for Dell PowerEdge
The NVIDIA A16-PCIE-64GB Quad-GPU for Dell PowerEdge delivers four discrete NVIDIA Ampere architecture GPUs on a single PCIe card, providing a combined 64 GB of GDDR6 memory (16 GB per GPU) designed specifically for dense virtual desktop infrastructure and remote visualization workloads at enterprise scale. Each of the four GA107 GPUs operates independently, enabling the card to support up to 64 simultaneous virtual GPU (vGPU) instances per card through NVIDIA's vGPU software stack, making it one of the highest-density virtualization accelerators available for rack-mounted PowerEdge server platforms.
Built to the PCIe 4.0 x16 form factor with passive cooling requirements satisfied by the PowerEdge chassis airflow, the A16 integrates tightly with NVIDIA Virtual PC (vPC), NVIDIA Virtual Applications (vApps), and NVIDIA RTX Virtual Workstation (vWS) profiles. The card exposes hardware-based H.264 and HEVC encode engines on each GPU, reducing CPU offload for graphically intensive virtual sessions and enabling smooth delivery of professional applications, 3D content, and HD video streams to remote endpoints across the enterprise.
Deployed within Dell PowerEdge rack servers, the A16-PCIE-64GB addresses the growing demand for scalable, centrally managed GPU resources in hybrid cloud and on-premises datacenter environments. Its multi-GPU topology allows IT administrators to allocate fine-grained fractions of GPU resources per user or workload, supporting everything from knowledge workers requiring basic 2D acceleration to engineers running moderate 3D CAD and simulation tasks—all from a single, power-efficient half-height, full-length passive card drawing a maximum of 250 W.
Ideal for
- High-density virtual desktop infrastructure (VDI) deployments serving large numbers of concurrent remote users from a single PowerEdge server node
- Remote professional visualization for CAD, BIM, and design workflows delivered via NVIDIA RTX Virtual Workstation profiles to distributed enterprise teams
- Centralized virtual application hosting where graphically accelerated apps are streamed to thin clients and BYOD endpoints across branch offices
- Healthcare and life sciences organizations requiring GPU-accelerated clinical imaging workstations delivered securely from an on-premises datacenter
- Education and government environments deploying persistent or non-persistent virtual desktops at scale with strict data residency requirements
- Media and entertainment post-production studios providing GPU-enabled remote editing and review sessions without distributing sensitive content to endpoints
Technical specifications
| Manufacturer | Dell |
| Manufacturer Part Number | 900-2G160-0020-000 |
| GPU Architecture | NVIDIA Ampere (GA107) |
| Number of GPUs per Card | 4 |
| Total GPU Memory | 64 GB GDDR6 |
| Memory per GPU | 16 GB GDDR6 |
| Memory Interface per GPU | 128-bit |
| Total Memory Bandwidth | Up to 512 GB/s (combined) |
| Form Factor | PCIe, Half-Height Full-Length (HHFL) passive |
| PCIe Interface | PCIe 4.0 x16 |
| Max Thermal Design Power (TDP) | 250 W |
| Power Connector | Dual 8-pin PCIe auxiliary |
| Display Outputs | None (headless compute/virtualization card) |
| Video Encode Engines | 4 (one per GPU), H.264 and HEVC hardware encode |
| Video Decode Engines | 4 (one per GPU) |
| vGPU Software Support | NVIDIA Virtual PC (vPC), Virtual Applications (vApps), RTX Virtual Workstation (vWS) |
| Max vGPU Instances per Card | Up to 64 (with 1 GB vGPU profiles) |
| Cooling | Passive (requires system-level airflow from chassis) |
| Compatible Platform | Dell PowerEdge rack servers with PCIe 4.0 x16 slot and passive GPU support |
| Operating System Support | VMware vSphere, Citrix Hypervisor, Red Hat Virtualization, Microsoft Hyper-V (via NVIDIA vGPU software) |
Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.
Technical Specifications
| Brand | Dell |
| Category | GPUs |
| SKU | 900-2G160-0020-000 |
| Part Number | 900-2G160-0020-000 |
| Condition | New |
| Manufacturer Part Number | 900-2G160-0020-000 |
| GPU Architecture | NVIDIA Ampere (GA107) |
| Number of GPUs per Card | 4 |
| Total GPU Memory | 64 GB GDDR6 |
| Memory per GPU | 16 GB GDDR6 |
| Memory Interface per GPU | 128-bit |
| Total Memory Bandwidth | Up to 512 GB/s (combined) |
| Form Factor | PCIe, Half-Height Full-Length (HHFL) passive |
| PCIe Interface | PCIe 4.0 x16 |
| Max Thermal Design Power (TDP) | 250 W |
| Power Connector | Dual 8-pin PCIe auxiliary |
| Display Outputs | None (headless compute/virtualization card) |
| Video Encode Engines | 4 (one per GPU), H.264 and HEVC hardware encode |
| Video Decode Engines | 4 (one per GPU) |
| vGPU Software Support | NVIDIA Virtual PC (vPC), Virtual Applications (vApps), RTX Virtual Workstation (vWS) |
| Max vGPU Instances per Card | Up to 64 (with 1 GB vGPU profiles) |
| Cooling | Passive (requires system-level airflow from chassis) |
| Compatible Platform | Dell PowerEdge rack servers with PCIe 4.0 x16 slot and passive GPU support |
| Operating System Support | VMware vSphere, Citrix Hypervisor, Red Hat Virtualization, Microsoft Hyper-V (via NVIDIA vGPU software) |
Frequently Asked Questions about NVIDIA A16-PCIE-64GB Quad-GPU for Dell PowerEdge
What server platforms accept the NVIDIA A16-PCIE-64GB Quad-GPU for Dell PowerEdge?
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.