SiFive’s Intelligence Gen 2 RISC-V IP portfolio combines scalar, vector, and matrix compute to speed up AI workloads. The Gen 2 lineup consists of the brand new X160 and X180, alongside the upgraded X280, X390, and XM sequence. All merchandise function enhanced scalar and vector processing, whereas the XM sequence provides a extremely scalable matrix engine.
With as much as 4 cores, the 32-bit X160 and 64-bit X180 goal embedded IoT on the far edge. They ship excessive effectivity in a compact footprint, extending AI to automotive, robotics, and industrial automation. Their vector engine boosts AI mannequin efficiency with minimal energy and space overhead.
Intelligence Gen 2 merchandise span a variety of efficiency, space, and energy choices inside a single scalable Instruction Set Structure (ISA). Options embody a dual-issue, in-order eight-stage superscalar pipeline, narrow-to-wide vector engines, and the XM sequence’ scalable matrix engine for numerous AI workloads. The CPUs additionally assist the SiFive Scalar Coprocessor Interface (SSCI) and Vector Coprocessor Interface eXtensions (VCIX) to hyperlink exterior AI accelerators and coprocessors.
All 5 Intelligence Gen 2 merchandise at the moment are out there for licensing, with first silicon anticipated in Q2 2026.
Intelligence Gen 2 product web page
The publish RISC-V IP expands AI capabilities on the edge appeared first on EDN.
