SiMa AI Unveils ‘MLSoC,’ a Purpose-Built Software-Centric Machine Learning System-on-Chip Platform for the Embedded Edge

'MLSoC Platform addresses any computer vision application while delivering 10x better performance per watt with a push-button experience'

Products for the “edge” market include drones, internet of things devices, and phones. And This market is served by the very broad field of artificial intelligence computer chips.

Numerous businesses have received funding in venture capital funding to provide chips for AI in smartphones and other embedded computing usages. The AI chip startup SiMa.ai presented its “MLSoC,” a system-on-chip for accelerating neural networks while consuming less power. The manufacturer claims that the new chip, which has already started to be shipped to clients, is the only component that is “purpose-built” to handle jobs that heavily rely on computer vision.

Description

The SiMa.ai chip includes many elements that are assembled into a single chip using Taiwan Semiconductor’s 16-nanometer production process. Among the components are a machine learning accelerator and “Mosaic,” a program devoted to matrix multiplications, the cornerstone of neural network processing.

 A standalone computer vision processor, a video encoder, and a decoder are among the functional units onboard, along with 4 megabytes of on-chip memory and a plethora of communications and memory access chips, including an interface to 32-bit LPDDR4 memory circuits. The ARM A65 processor core, which is frequently found in automobiles, is also present. 

The chip hardware includes SiMa.ai software to facilitate performance tuning and support a wide range of workloads.

Use Cases

Robots, drones, autonomous cars, industrial automation, and applications in the healthcare and government markets are just a few of the markets SiMa.ai’s solution aims towards.

How is it different

Numerous embedded and mobile rivals face SiMa.ai. The intellectual property juggernaut ARM, Qualcomm, Intel, and Nvidia are competitors in the edge market along with AMD, now the parent company of Xilinx. However, those businesses have historically concentrated on larger chips that consume much more power, on the scale of tens of watts, while this focuses on smartphones to embedded ones that consume only a few microwatts of power.

According to its developers, the SiMa.ai chip has one of the lowest power budgets of any chip now available for carrying out standard tasks like ResNet-50, the most popular neural net for processing the ImageNet tasks of classifying images.

References:

  • https://sima.ai/sima-ai-ships-first-industry-leading-purpose-built-machine-learning-soc-platform-to-customers-for-embedded-edge-applications/
  • https://www.zdnet.com/article/ai-startup-sima-debuts-purpose-built-ai-chip-for-edge-computing/
Please Don't Forget To Join Our ML Subreddit

Prathvik is ML/AI Research content intern at MarktechPost, he is a 3rd year undergraduate at IIT Kharagpur. He has a keen interest in Machine learning and data science.He is enthusiastic in learning about the applications of Machine learning in different fields of study.