Cisco UCSC-GPU-M10 | Tesla M10 GPU, 32GB GDDR5, 4x Maxwell GPUs, PCIe 3.0

Q: What is the primary use case for the UCSC-GPU-M10?
A: The UCSC-GPU-M10 is optimized for density-driven virtual desktop infrastructure (VDI) and application virtualization. It works with NVIDIA GRID software to deliver high user density, supporting up to 64 concurrent virtual desktop sessions per board, making it ideal for enterprise organizations deploying virtual workspaces at scale.

Q: How many virtual users can this GPU support simultaneously?
A: With a 1GB memory profile, the Tesla M10 supports up to 64 users per board (16 users per GPU). When installed in a dual-GPU server configuration, it can support up to 128 users per server, providing exceptional density for VDI deployments.

Q: Which Cisco servers are compatible with this GPU?
A: The UCSC-GPU-M10 is designed for Cisco UCS C-Series rack-mount servers, including models like the UCS C240. It requires a PCIe 3.0 x16 slot and dual-slot clearance with passive cooling support, as the card uses server airflow for thermal management.

Q: Does this GPU require active cooling or special airflow?
A: The Tesla M10 uses passive cooling and relies on the server's internal airflow for heat dissipation. Ensure your UCS server has adequate front-to-back airflow and sufficient PCIe slot clearance for the dual-slot, full-height form factor.

Q: What software is required to enable virtual GPU functionality?
A: The UCSC-GPU-M10 works with NVIDIA GRID virtualization software to enable vGPU profiles and desktop/application virtualization. Hypervisor support includes VMware vSphere and other enterprise virtualization platforms compatible with NVIDIA vGPU technology.