M10 Enterprise GPU Maxwell 28nm 28*H.264 Streams Virtual Workstation Support
Tesla M10
The Tesla M10 is a dual‑slot, PCIe 3.0 x16 data center GPU released in May 2016, based on the 28nm Maxwell (GM107) architecture.
| Architecture |
28nm Maxwell (GM107) |
| GPU Configuration |
4 * GPUs per card |
| Processing Cores |
2560 CUDA cores total (640 per GPU) |
| Memory |
32GB GDDR5 total (8GB per GPU) with ECC protection |
| Memory Bandwidth |
332GB/s total |
| Clock Speeds |
1033MHz base / 1306MHz boost |
| FP32 Performance |
4 * 1.67 TFLOPS |
| Power Consumption |
225W TDP with passive cooling |
Key Features & Applications
- Optimized for GRID vGPU technology
- Supports up to 32 virtual users per card
- Capable of 28 H.264 1080p30 streams
- Ideal for high-density VDI deployments
- Virtual workstation support
- Video transcoding applications
- Enterprise reliability with ECC memory protection