With software evolving quickly, it has become increasingly common for hardware vendors to create specialist AI processors that run their software more efficiently than would be possible with out-of-the-box hardware. While these custom processors can offer the advantage of performance, they come with challenges for developers. Primary among these is the challenge of enabling the latest software on the latest generation of processors. This can involve major software porting efforts to proprietary and non-standard programming models, creating additional work for customers, who will need to work on time-consuming optimizations and porting for their applications.

The oneAPI Construction Kit was created to solve these challenges by bringing all the benefits of simplified heterogeneous programming to new and custom hardware. It brings the oneAPI ecosystem to these architectures and makes it easy to access a wealth of supported SYCL-based libraries. All of this means less time spent on porting efforts and maintaining separate codebases for different architectures, and more time for innovation.

With the SYCLOPS Project, Codeplay will bring together the RISC-V and SYCL standards into a portable software stack via the oneAPI Construction Kit. The hardware will come from another member of the project, Codasip. Their custom RISC-V based accelerator, together with the oneAPI Construction Kit, will demonstrate competitive performance with open standards and portable code. Three use cases – high-energy physics, oncology and autonomous systems – will highlight the versatility of the SYCLOPS software stack and show real world benefits.

On a smaller scale, Codeplay has previously demonstrated a full software programming environment using oneAPI and SYCL for the next generation of RISC-V vector processors using the oneAPI Construction Kit, and in this short clip below, our CEO and co-founder Andrew Richards demonstrates this by running code using an Intel CPU and a custom FPGA simulating a RISC-V accelerator.

Codeplay and Axelera AI jointly released a case study that highlighted the success Axelera AI had when using the oneAPI Construction Kit to bring open standards to their Metis AIPU. This was a real testament to what can be achieved with the oneAPI Construction Kit, and you can read more about this here!