Flex Logix Technologies, a supplier of embedded FPGA (eFPGA) IP, architecture and software, has been selected to be part of a team of microelectronic companies, led by Microsoft, to build a chip development platform for security as demonstrated by the DoD RAMP Project.

Flex Logix was chosen for its embedded (eFPGA) technology that enables chips to be reconfigurable after tape-out, allowing companies to adapt to new requirements, changing standards and protocols as needed.

“Security can mean a lot of different things. For eFPGA, it means providing a means for keeping circuitry secret by programming the eFPGA in a secure environment,” said Geoff Tate, CEO and Co-founder of Flex Logix.

The RAMP project intends to address and replace the obsolete practices of the US Government in support of state-of-the-art custom IC and system-on-chip design especially those associated with physical or “back-end” design. 

Given the long system lifetime of DoD systems, utilising eFPGA technology will provide the capability to make silicon changes to update algorithms or modify functionality and will also help to save money by avoiding costly tape-outs and installation of replacement parts.

"Our collaboration with Flex Logix will help to advance flexibility to chip design," said Mujtaba Hamid, General Manager, Silicon, Modeling and Simulation, Microsoft. "The company’s eFPGA technology has demonstrated success in supporting mission-critical applications and we look forward to bringing this capability to support national security priorities."

Using Flex Logix’s EFLX, chip developers are able to implement eFPGAs from a few thousand LUTs to hundreds of thousands of LUTs with performance and density per square millimetre similar to leading FPGA companies in the same process generation. EFLX eFPGA is modular so arrays can be spread throughout the chip, can have all-logic or be heavy-DSP, and can integrate RAM in an array of many types.

Leveraging Flex Logix’s patented interconnect used in EFLX, Flex Logix has developed InferX, an AI inferencing tensor accelerator that is said to deliver the highest throughput per dollar compared to other edge inferencing solutions.

The combination of EFLX, InferX and a host processor will create a cohesive programmable and reconfigurable cognitive system.