Hardware Acceleration with vSphere Bitfusion
What is vSphere Bitfusion?
vSphere Bitfusion supports AI- and ML-based workloads by virtualizing hardware accelerators such as GPUs. Multi-vendor hardware accelerators and the ecosystem around them are key components for delivering modern applications. These accelerators can be used regardless of location in the environment.
How does vSphere Bitfusion work?
vSphere Bitfusion decouples physical resources from servers within an environment. The platform can share GPUs in a virtualized infrastructure, as a pool of network-accessible resources, rather than isolated resources per server. Bitfusion works across AI frameworks, clouds, networks, and in environments such as virtual machines, containers, and notebooks. And Bitfusion stands ready to carry virtualization forward as new hardware accelerators are introduced.
Features
Dynamic GPU Attach Anywhere
Bitfusion disaggregates your GPU compute and dynamically attaches GPUs anywhere in the datacenter, just like attaching storage.
Fractional GPUs for Efficiency
Bitfusion enables use of any arbitrary fractions of GPUs. Support more users in the test and development phase.
Standards Based Accelerator Access
Leverage GPUs across an infrastructure plus integrate evolving technologies as standards emerge.
Application Run Time Virtualization
Bitfusion attaches GPUs based on CUDA calls at run-time, maximizing utilization of GPU servers anywhere in the network.
Any Application
Bitfusion is a transparent layer and runs with any workload, framework, container or notebook.
Components
The most advanced GPU virtualization technology to accelerate AI, ML and HPC.
vSphere Bitfusion
vSphere Bitfusion connects any compute servers remotely, over the network to GPU servers pools while attaching and detaching GPUs. This virtualizes the GPUs, allowing multiple workloads to run in parallel. Bitfusion does for GPUs what vSphere did for CPUs.
vSphere Bitfusion Manager
vSphere Bitfusion Manager is the management tool which allows the administrator to Monitor health, utilization and efficiency and availability of all GPU servers in the network. The tool also provides viewing and monitoring of client consumption of GPUs, plus assigning quotas and time limits.
vSphere Bitfusion Pricing and Packaging
vSphere Bitfusion requires an Enterprise+ License on the GPU server plus the Bitfusion Add-On feature.
Resources
Learn and Evaluate
- Mike Adams discusses vSphere Bitfusion availability with Doug Black of Inside HPC Publication
- VMware Kubernetes Integration with Bitfusion
- VMware vSphere 7 with Bitfusion on Dell EMC PowerEdge Servers
- Scaling Distributed Machine Learning with Bitfusion on Kubernetes
- Mohan Potheri at Supercomputing 2019 Discusses Bitfusion with Inside HPC Publication
- Bitfusion Solution Brief
- AI Composability and Virtualization: Mellanox Network Attached GPUs
- Elastic AI Edge Computing
- Bitfusion Provides an Elastic AI Infrastructure Disaggregation Platform for Cloud and On-Premise
- Network-Attached Full or Partial GPUs to Any Kubernetes Cluster Using Bitfusion FlexDirect
Technical Documentation – Legacy Bitfusion
- Bitfusion Guide to TensorFlow
- Bitfusion Guide to CUDA Installation
- Bitfusion using 3rd Party Software
- Bitfusion Docs FAQ and Troubleshooting
- Bitfusion Docs FlexDirect On-Premise Evaluation Guide
- Bitfusion Docs FlexDirect
- Bitfusion Docs Installing Bitfusion FlexDirect
- Bitfusion Docs Integration and Advanced Usage
- Bitfusion Docs System Requirements
- Bitfusion Docs FlexDirect Health Checks
- Using Bitfusion Flex Direct