Introducing an expanded portfolio of modules that provides builders a single, scalable structure to energy every thing from low-power gadgets to enterprise edge infrastructure
As AI strikes from cloud environments into factories, transportation methods, robotics, and clever infrastructure, builders face rising challenges associated to energy consumption, reminiscence bandwidth, deployment complexity, and infrastructure prices. MemryX Inc., a fabless semiconductor firm growing edge AI processing options powered by proprietary compute-at-memory know-how, as we speak expanded its Cascade deployment platform household with three new merchandise constructed on its MX3 accelerator structure that allow AI deployment throughout embedded methods, edge gadgets, and high-density edge server infrastructure. The merchandise are being launched this week at Automate 2026 at Sales space 15045 within the North Constructing of McCormick Place in Chicago.
MemryX Cascade 100P PCIe Accelerator
The MemryX Cascade 100P PCIe Accelerator is engineered for high-density edge server deployments and demanding multi-camera AI functions. Constructed on a 16-chip array of MX3 processors, the cardboard permits parallel inference on tons of of real-time video streams. That includes a single-slot, passively cooled type issue, the Cascade 100P simply integrates into normal server environments—permitting operators to maximise compute density with out including complicated thermal or power-budget constraints. This makes it an excellent answer for sensible manufacturing, clever transportation, video analytics, and multi-camera AI deployments.
MemryX Cascade 100U USB Accelerator
The MemryX Cascade 100U USB Accelerator is engineered to carry AI acceleration to current methods by means of a regular USB Sort-C connection. That includes two MX3 accelerators, the system permits builders to quickly consider, prototype, and deploy AI imaginative and prescient functions with out redesigning current {hardware} platforms.
Additionally Learn: AiThority Interview with Matej Bukovinski, Chief Know-how Officer at Nutrient
MemryX Cascade 100R Raspberry Pi HAT+
The MemryX Cascade 100R Raspberry Pi HAT+ is engineered to carry hardware-accelerated AI inference to the Raspberry Pi ecosystem for robotics, machine imaginative and prescient, training, and low-power industrial functions. Constructed round two MX3 processors linked by means of PCIe, the HAT could be cascaded to assist bigger AI deployments.
All three merchandise are supported by MemryX Developer Hub, offering a standard software program atmosphere for compiling and deploying AI fashions throughout the MX3 product household.
“AI doesn’t have a mannequin downside anymore, it has a deployment downside,” stated Ross Jatou, CEO of MemryX. “Whether or not builders and system builders are prototyping on Raspberry Pi, including AI acceleration to current methods by means of USB, or deploying tons of of video streams on edge servers, they want a standard structure that scales seamlessly and delivers efficiency with out the ability and reminiscence penalties of conventional approaches. Our expanded Cascade platform household offers that basis, enabling prospects to maneuver from improvement to manufacturing on a single, scalable, extremely environment friendly AI structure powered by MX3.”
Collectively, the brand new merchandise set up an entire MX3 deployment ecosystem, enabling organizations to scale AI functions from improvement and embedded deployment to high-density edge server environments utilizing a standard {hardware} and software program basis.
