AI / HPC GPU Accelerator NVIDIA
Tesla Referral P100 12GB PCIe Graphics Accelerator Card | HBM2 GPU for AI & HPC Original Pull, Tested Working | 12GB HBM2 | PCIe 3.0 x16 | Pascal Architecture Why Buy This ✓ Pascal Architecture — powerful GPU for AI inference, deep learning and HPC ✓ 12GB HBM2 — high-bandwidth memory for large models and datasets ✓ PCIe 3.0 x16 — fits standard PCIe slots in workstations and servers ✓ Tested Working — original pull, verified fully functional What You'll Get
An NVIDIA Tesla Referral P100 12GB PCIe graphics accelerator card as pictured. This is an original pull from working servers, fully tested and verified working.
Built on the Pascal architecture with 12GB HBM2 high-bandwidth memory and a PCIe 3.0 x16 interface, the Tesla Referral P100 is designed for AI inference, deep learning training and high-performance computing (HPC).
Real photos shown — exactly what you will receive. A cost-effective way to add GPU acceleration to your workstation or server for AI, machine learning and compute workloads.
Key Features
- NVIDIA Tesla Referral P100 graphics accelerator card
- 12GB HBM2 high-bandwidth memory
- Pascal architecture — 3,584 CUDA cores
- PCIe 3.0 x16 interface — fits standard PCIe slots
- Designed for AI inference, deep learning and HPC
- Original pull, tested working, as pictured
- Passive cooling — requires chassis/active airflow
Specifications Item Detail Brand NVIDIA Model
Tesla Referral P100 Memory 12GB HBM2 Architecture Pascal Interface PCIe 3.0 x16 Use Cases AI inference, deep learning, HPC, GPU compute Condition Used, Original Pull, Tested Working, as Pictured Condition Original pull, tested working. As pictured. Cosmetics may show normal signs of server use — fully tested and verified functional.
- Original pull from working servers
- Fully tested and verified working
- As pictured — exactly what you will receive
- Passive cooling (no fan) — requires chassis airflow
⚠ Important — Read Before Buying:
1. This is a passive server GPU — it has no built-in fan. It requires adequate chassis airflow / server-grade cooling. It is NOT designed for typical desktop gaming cases.
2. No video output ports — it is a compute GPU. Use an iGPU or a separate display card for your monitor.
3. Your motherboard must support Above 4G Decoding for proper operation.
4. Please confirm your power supply has sufficient wattage and the required PCIe power connectors.
5. As pictured — minor cosmetic signs of use are normal for server pulls. Package Includes
1 × NVIDIA Tesla Referral P100 12GB PCIe Graphics Accelerator Card
Note: Passive cooling (no fan). As pictured.
Warranty & Returns Warranty 30-day functional warranty from delivery date Returns 30-day returns accepted (buyer pays return shipping if not DOA) Testing GPU fully tested before dispatch Shipping
- Dispatched within 1–3 business days after payment
- Tracking number provided for every order
- Shipped in anti-static packaging with protective foam
- Standard & Expedited options at checkout
Why Buy From Us
- Tested working — every GPU verified before shipping
- Original pull — genuine NVIDIA Tesla Referral hardware
- Cost-effective AI — Pascal-class GPU at a great price
- Fast response — message us for compatibility or bulk inquiries
FAQ
Q: Can I use this for gaming?
A: It is a compute/server GPU with no video output. It is designed for AI, HPC and GPU compute — not gaming. You need a separate display card for your monitor.
Q: Does it come with a fan?
A: No. This is a passive server GPU. It requires adequate chassis airflow / server-grade cooling.
Q: What motherboard do I need?
A: A motherboard with a free PCIe x16 slot and Above 4G Decoding enabled in BIOS.
Q: What condition is it in?
A: Original pull, tested working, as pictured. Minor cosmetic signs of server use are normal.
Questions about compatibility, cooling or bulk orders? Message us before ordering — we reply fast.
