Home > Workload Solutions > Container Platforms > Red Hat OpenShift Container Platform > Archive > GPU-Accelerated Machine Learning with OpenShift Container Platform > Dell EMC PowerEdge C4140
The PowerEdge C4140 server is the densest 1U GPU accelerator-optimized server to be released from an enterprise server vendor. The PowerEdge C4140 server in configuration M contains up to four NVIDIA V100 GPUs in the SXM2 form factor and an NVIDIA NVLink interconnect, as shown in the following figure, allowing GPU-to-GPU communication in a full mesh configuration.
Figure 3. PowerEdge C4140 server with NVLink (top view)
Other configurations of the C4140 support general-purpose computing on graphics processing units (GPGPU) form factors. The PowerEdge C4140 is ideal for scale-out training workloads where multiple C4140 servers can be connected to achieve faster time-to-completion of the model training by using a distributed framework and communications library.