White Papers

Direct from

Development

PowerEdge Product Group

Figure 3: Performance comparison of Network Attached Remote GPUs with FlexDirect on Dell EMC PowerEdge Servers vs. Native GPU execution

Performance is represented on y-axis (images/sec, higher is better), for training different CNN models using Tensorflow.

Figure 4: Fractional GPUs with FlexDirect on Dell EMC PowerEdge Servers vs. using a full GPU. The

aggregate performance of fractional GPUs (which can be assigned to different users) is compared against a full

GPU when training different model types using TensorFlow.

Conclusions

Bitfusion FlexDirect with Dell EMC PowerEdge Servers can bring flexibility and composability to the IT infrastructure along

with reducing overall Capex and Opex with shared GPU pools across multiple organizations, business needs and use-

cases. In addition, Bitfusion FlexDirect on Dell EMC PowerEdge Servers can increase user and organization productivity

by meeting GPUs consumption from a consolidated resource pool based on real-time workload needs, to any machine in

the environment.

Learn More

Visit https://www.bitfusion.io/product/flexdirect to get more information on Elastic Network Attached GPUs.

Visit http://www.dellemc.com/AI to get more information on Dell EMC PowerEdge Solutions for AI.