JetStream is d-Matrix’s I/O accelerator connecting Corsair systems for scalable data center AI inference.
JetStream is an I/O accelerator that connects d-Matrix Corsair hardware across servers used for AI inference. Its transparent NIC architecture enables accelerator-to-accelerator communication through PCIe Gen5 and standard Ethernet interfaces, supporting model execution across multiple nodes.
Operators use JetStream in rack-scale Corsair deployments that need more compute or memory than one server provides. Its PCIe card connects to existing Ethernet switches, allowing data centers to scale inference clusters while retaining standard networking infrastructure.
| Attribute | JetStream | CN5000 Omni-Path | InfiniBand, Spectrum-X | |
|---|---|---|---|---|
| Provider |  |  |  | |
| Founded | 2019 | 2020 | 1993 | |
| Sells To | Enterprise | Enterprise, Government | Consumers, Enterprise | |
| Pricing Model | — | Recurring | Licensing, Product Sales, Subscription | |
| Ownership | Private | Venture-backed | Public |
Provider
Founded 2019
Sells To Enterprise
Pricing Model —
Ownership Private
Provider
Founded 2020
Sells To Enterprise, Government
Pricing Model Recurring
Ownership Venture-backed
Provider
Founded 1993
Sells To Consumers, Enterprise
Pricing Model Licensing, Product Sales, Subscription
Ownership Public
Cerebras CS-4 is a rack-scale AI system built for fast, large-scale inference.
High-bandwidth, low-latency cloud networking for distributed AI and GPU clusters.
AuraLinks is an open-standards network fabric for scale-up and scale-out interconnect in AI data centers.
Scale-out Ethernet systems are open fabrics built on NVIDIA Spectrum-X switch silicon and a SONiC-based NOS for AI clusters.
Arista AI Networking is a portfolio for high-performance AI cluster connectivity.
SkyHammer is a rack-scale, low-latency scale-up fabric designed from the ASIC up for synchronized AI workloads.
HPE Slingshot is a high-performance cluster network for AI and supercomputing workloads.
Cerebras CS-4 is a rack-scale AI system built for fast, large-scale inference.
High-bandwidth, low-latency cloud networking for distributed AI and GPU clusters.
AuraLinks is an open-standards network fabric for scale-up and scale-out interconnect in AI data centers.
Scale-out Ethernet systems are open fabrics built on NVIDIA Spectrum-X switch silicon and a SONiC-based NOS for AI clusters.
Cisco Silicon One is a unified networking silicon architecture.
Arista AI Networking is a portfolio for high-performance AI cluster connectivity.
SkyHammer is a rack-scale, low-latency scale-up fabric designed from the ASIC up for synchronized AI workloads.
HPE Slingshot is a high-performance cluster network for AI and supercomputing workloads.