The combination of the SiFive P870-D CPU and Arteris Ncore cache coherent interconnect IP helps to accelerate the work of SoC design engineers looking to address the demanding AI workload requirements of the datacentre and embedded systems markets.
SiFive’s P870-D CPU enables high compute density and supports the open AMBA CHI protocol so customers have greater flexibility when it comes to scaling the number of clusters to boost performance while minimising power consumption.
Arteris Ncore cache coherent interconnect IP is a configurable and scalable network-on-chip (NoC) interconnect for heterogeneous cache coherent systems-on-chip (SoCs).
The pre-verified solution enables faster, low-risk development cycles for datacentre SoCs with high performance AI workload and power efficiency requirements.
"The latest focus of the multi-year collaboration between Arteris and SiFive is centred on reducing the risk, development cost and timeline for customers that wish to create the highest performance, most scalable RISC-V based SoCs for datacentre use cases,” explained Ian Ferguson, Senior Director of Marketing at SiFive. “SiFive is excited to partner with Arteris on its leading work on a RISC-V based emulation platform with our P870-D processor being one of the first CPU cores to be supported as part of this program."
“Cache coherent network-on-chip interconnect IP is foundational in the creation of complex, high-performing SoCs and we support the processor IP choices of our customers,” said Michal Siwinski, CMO of Arteris. “We are pleased to continue the collaboration with SiFive and enable the RISC-V ecosystem to accelerate time to market while reducing project risk.”
Ncore cache coherent interconnect IP is able to enhance system performance by reducing latency and optimising data flow between processing units. Its proven design also ensures reliability and minimises design iterations and supports complex multicore SoCs with efficient cache management and scalability, making it suitable for the burgeoning datacentre market.
