Imagination launches line-up of RISC-V processors
Catapult, from Imagination Technologies, is a RISC-V processor family designed for next-generation computing and comes in different versions to suit a variety of performance and efficiency requirements.
Supported by the open-standard RISC-V ecosystem, Catapult CPUs are available as dynamic microcontrollers, real-time embedded CPUs, high-performance application CPUs and functionally safe automotive CPUs. The microcontrollers are already shipping in high-performance automotive GPUs in SoCs. Real-time embedded CPUs are available now, while high-performance application CPUs and automotive CPUs will follow in 2022.
Imagination says the new processors are designed for markets including 5G modems, storage, ADAS/autonomous vehicles, data centres, and high-performance computers. Imagination says the chips feature multi-threaded RISC-V cores and offer a host of configuration options including 32bit or 64bit variants, the ability to be scaled up to eight asymmetric coherent cores per cluster and offer the option to add custom accelerators.
As part of Imagination’s heterogenous computing solution, Catapult CPUs offer full hardware, software and debug support for SoCs using Imagination IP with the ability to work with its GPU, AI and Ethernet packet processor cores.
Based on industry-proven security concepts, meeting safety and certification requirements for target markets, Catapult automotive processors are developed to ISO 26262 automotive standards and provide a range of CPU solutions for each automotive safety integrity level.
They are delivered with a fully featured SDK, which includes enhanced versions of industry-standard build and debug tools such as GCC, LLVM and GDB, optimised C libraries, and Imagination’s IDE, Catapult Studio. Catapult Studio is based on visual studio code, with extra features focusing on embedded, RISC-V development and close integration to the wider SDK.
Catapult SDK is available for Windows, Ubuntu, CentOS and MacOS and includes FreeRTOS and full support for Linux including reference bootloaders, kernels and Yocto-based filesystems.
“As the demand for compute grows in the cloud, at the edge and in devices, there is an ever-greater challenge to process immense amounts of data in tightly constrained area and power budgets. Heterogeneous architectures are key to providing performance, flexibility and resilience when accelerating an increasingly diverse set of workloads,” said Tim Mamtora, chief of innovation at Imagination.
