RISC-V AI Revolution: DeepComputing and Fractile Partner with Andes Technology for Next-Gen AI Hardware

RISC-V AI PC: A Collaborative Breakthrough

DeepComputing and Andes Technology Corporation have joined forces to develop the world's first RISC-V AI PC, powered by Andes' 7nm QiLai SoC 1

1

. This innovative low-power PC, featuring Ubuntu Desktop, aims to redefine AI computing by combining industry-leading hardware and software designed specifically for RISC-V architecture.

The QiLai SoC integrates two Andes RISC-V processors: a high-performance quad-core AX45MP cluster and an NX27V vector processor. Running at up to 2.GHz (AX45MP) and 1.GHz (NX27V), the SoC delivers high performance while maintaining low power consumption of approximately 5W at full speed 1

1

.

Frankwell Lin, Chairman and CEO of Andes, expressed excitement about the collaboration, stating, "This AI PC project will demonstrate the power of the RISC-V architecture for general application processing and AI acceleration, and provide a powerful RISC-V platform for application development and processor IP evaluation" 1

1

.

Fractile's AI Inference Accelerator

In a parallel development, Fractile, an AI hardware company, has licensed Andes Technology's RISC-V vector processor for its radical new chip designed to accelerate AI inference 2

2

. Fractile aims to run frontier AI models, including large language, vision, and audio models, two orders of magnitude faster than existing hardware at a tenfold reduction in cost.

Fractile's innovative approach uses novel circuits to execute 99.99% of the operations needed for model inference in on-chip memory. This architecture promises both higher energy efficiency and dramatically improved latency on inference tasks 2

2

.

Advancing RISC-V in AI Computing

Both collaborations underscore the growing importance of RISC-V architecture in AI computing. The RISC-V AI PC will come equipped with a suite of tools and frameworks optimized for AI workloads, including AndeSight™ toolchains, AndeSoft™ software stacks, and AndesAIRE™ NN SDK 1

1

.

Dr. Charlie Su, President and CTO of Andes Technology, highlighted the versatility of their RISC-V vector processors, stating, "Andes RISC-V vector processors have enabled many AI SoCs to break free from architecture limitation and achieve new levels of performance and efficiency" 2

2

.

Industry Impact and Future Prospects

These developments are set to have a significant impact on the AI hardware landscape. The RISC-V AI PC is scheduled to be unveiled at the RISC-V Summit NA 2024 and will be officially available in early 2025 1

1

. Meanwhile, Fractile's AI inference accelerator promises to address the growing demand for more efficient AI model serving, particularly for large language models and other foundation models 2

2

.

Dr. Walter Goodwin, CEO and founder of Fractile, emphasized the importance of their collaboration with Andes, stating, "Software-programmable vector processors like Andes' are a key part of staying robust to these changes. We're delighted to announce this collaboration as Fractile furthers its mission to supercharge inference" 2

2

.

As these innovations come to market, they are expected to accelerate the development of RISC-V-based AI solutions and expand the reach of RISC-V in the broader computing landscape, potentially reshaping the future of AI computing.