Unleash the Computing Potential Of All Your GPU Accelerators at Once.


Fungible GPU-Connect™ solves one of the most serious data center challenges caused by the growing demand for AI/ML and Edge computing. For the first time ever, using a GPU enabled composable infrastructure, you can be free of traditional GPU architecture constraints.

FGC provides the capability to dynamically compose GPUs and connect them to CPUs over a standard Ethernet network. Centralizing GPUs into a single pool that can be attached to the host servers on demand eliminates the need for dedicated GPUs. This disaggregation solution removes the constraints of having GPUs and CPUs physically co-located within the same server. Deliver GPU processing power independent of application’s CPU requirements. Expensive GPU resources are no longer left stranded. FGC allows any host server to receive physical GPU resources from a centralized GPU pool anywhere within the same data center.

Product Brief

What You Get

Fungible GPU-Connect components are FC200 cards powered by the Fungible DPU™, FX-108 expansion shelves to hold your GPUs, and the Fungible Composer software.

  1. Any standard x86 server fitted with an FC200 card can be connected over an IP/Ethernet network to the remote GPUs in the FX-108 shelves.
  2. Each FX-108 shelf can pack up to 8 full height, full length GPUs and connect them to a network at 8x100GE. Supports both actively and passively cooled GPUs.
  3. Use Composer to assign users to specific GPUs within the pool manually or through an API.


Differentiate service value

Offer higher compute power and time at lower COGS

Increase ROI

Use existing GPU assets to provision new users without having to buy new servers

Decrease utility expenses

Reduce the size of the physical footprint requiring server level environmental controls

Speed up cash flow

Faster new client onboarding means you can start billing sooner

Achieve GPU efficiency at data center scale

Give your entire data center dynamic access to the central GPU pool

Increased profitability

Serve more clients with existing GPUs, dropping profit straight to the bottom line


Request a Demo