413 Views

Cutting-Edge RISC-V Processor and TSN Switch for AI Space Applications

LinkedIn Facebook X
July 09, 2024

Get a Price Quote

Microchip is making significant strides in the field of space technology by qualifying an eight-core fault-tolerant RISC-V processor designed specifically for artificial intelligence applications in space. The radiation-tolerant PIC64-HSPC octal core 1.2GHz switch, boasting 26K DMIPS, is constructed on a 12nm FINFET process at GlobalFoundries in the United States.

The PIC64-HSPC represents the second addition to Microchip's new 64-bit RISC-V processor family. It leverages the X280 RISC-V cores from SiFive, incorporating vector extensions tailored for machine learning and AI tasks. This development marks the initiation of the 64-bit PIC64 family with RISC-V architecture.

The eight-core processor features advanced capabilities such as 240Gbit/s time-sensitive networking (TSN), CXL2.0 memory interconnect, and RMAP-compatible SpaceWire ports with internal routers. With a three-level cache, the device can deliver up to 2TOPS of INT8 AI inference performance.

Microchip is fostering a new ecosystem for space design, collaborating with more than a dozen system and software partners to expedite the adoption of PIC64-HPSC on Single-Board Computers (SBCs), space-grade companion components, and a network of open-source and commercial software partners.

The radiation- and fault-tolerant PIC64-HPSC is being deployed to NASA and the wider defense and commercial aerospace sector for AI and machine learning applications. Notable fault-tolerance features include Dual-Core Lockstep (DCLS) operation, WorldGuard hardware architecture for partitioning and isolation, and an onboard system controller for fault monitoring and mitigation.

Recent Stories


Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.