AMD Debuts Peano, An Open-Source LLVM-Compiler Designed For Ryzen AI NPUs & AI Accelerators

AMD has announced their new AI-focused project, the "Peano", which is an open-source LLVM designed to cater to Ryzen's cutting-edge AI engines.

With the AI hype now transitioning into the personal computing segment, the trend of dedicated AI engines or so-called "NPUs" in modern-day processors has become the new norm, and every other firm is involved in racing against each other in the race of "TOPS", which represents AI compute power. Amid all, AMD has made a pretty exciting addition to Linux, revealing a new project called the Peano, which is targeted directly towards AMD's Phoenix, Hawk Point, and the Strix APUs along with their respective XDNA AI engines.

Let's talk about what Peano is. Team Red's new LLVM compiler is explicitly designed for Ryzen AI engines. In simple terms, a dedicated LLVM compiler will allow for much more efficient use of resources and allow developers to leverage existing programming languages and tools to develop applications that can harness the processing power of these AI engines. Here is how AMD's Stephen Neuendorffer describes the new project:

On behalf of AMD, I’m pleased to announce the open sourcing of an LLVM backend for AMD/Xilinx AI Engine processors. These processors exist in a number of devices including RyzenAI SoCs. The repository currently focuses on supporting the AIE2 architecture implemented by the XDNA accelerators in “Phoenix” and “Hawk Point” devices.

Note that these accelerators include an array of processors, while the LLVM backend only supports a single processor. Support for devices as a whole is available in open source tools based on MLIR.

via LLVM.org

The Peano LLVM compiler is currently compatible with AMD's Phoenix and Hawk Point, while support for the new "Ryzen AI 300" APUs is in progress. Phoronix has highlighted the importance of Peano being an open-source project and how it will enable massive development relevant to Ryzen AI engines, ensuring rapid progression of the platform and a wider ecosystem adoption. However, it's important to note that AMD's competitor Intel has already shipped out its IVPU kernel driver with the debut of the Meteor Lake series, so in terms of enablement, Team Red is a bit behind.

Dedicated AI engines such as AMD's XDNA and Intel's VPUs will play a vital role in shaping the next generation of computing markets, and the effects are already visible with the AI PC narrative.

News Source: Phoronix