Skip to content

Gryf Resource Sled: GPU / Accelerator

Bring the GPU to the Mission.
Not the Mission to the GPU.

Delivering dedicated acceleration at the operational edge: GPUs, FPGAs, AISICs — including those with enough memory to run the latest AI inference models.

Gryf Resource Sled: GPU / Accelerator

GPU / Accelerator Sled

Purpose-Built for Inference at the Edge

AI inference, deep learning, computer vision, and signals processing all demand GPU acceleration. The Gryf GPU / Accelerator Sled puts that power right at the edge where decisions are made.

144GB Max GPU Memory

400 W Accelerator Power

One Integrated Environment

What makes the GPU / Accelerator Sled exceptional isn’t just its GPU, it’s how GigaIO’s patented software technology, FabreX, integrates it into the broader Gryf system. FabreX uses PCIe to support peer-to-peer transfers between GPUs — as well as between GPUs and storage — directly across the fabric, eliminating the CPU as an intermediary and dramatically reducing latency on deep learning and inference workloads.

Gryf’s sleds hold almost any full-length, full-height, double-wide GPU, so you can easily have enough GPU memory to run the latest AI models at the edge. And multiple accelerator sleds across a Gryf stack can be treated as a unified accelerator pool — disaggregated, composable, and dynamically allocated to whichever workload needs them most.

If your workload needs different types of accelerators, perhaps a video processing card or an audio card, no problem. They can all go in the accelerator drawer. Pair with the Compute Sled to feed the fabric, or add a Storage Sled to keep training datasets close to the accelerator.

Specifications

Supported Cards
Full-Length, Full-Height, Double-Wide PCIe
Fabric Bandwidth
256 Gbps Board-to-Board
Accelerator Power
Up to 400 W
Accelerator Memory
Depends on GPU Selection, but Supports Models Up to 144GB
Interconnect
GigaIO’s PCIe-based FabreX
Sleds per Gryf
Up to 6 Configurable Slots
Gryf Resource Sled: GPU / Accelerator Top

Simple to Service

Should a component need service, simply swap the bad sled with a good one. No waiting for special technicians to be flown in. No opening boxes, no cables and connectors to be confused by. Swapping a sled is just like swapping a battery.

Every sled includes remote management through IPMI/BMC, so your team can restart, diagnose, or reconfigure it through the chassis RJ45 port, no matter what state the primary OS is in. No hands-on-hardware required, no waiting on someone to be standing next to it.

Gryf Sleds

Complete Your Configuration

Each Gryf can hold six hot-swappable sleds. Add more than one GPU/Accelerator Sled or choose from other options:

Ready for Datacenter-class AI at the Edge?

Talk to a GigaIO expert about your field environment and see if Gryf is the right fit for your workload.

"*" indicates required fields

This field is for validation purposes and should be left unchanged.

Envelope

Contact Us

  • This field is for validation purposes and should be left unchanged.