SiMa.ai, the software-centric, embedded edge machine learning system-on-chip company, today announced it has raised an additional $70M of funding led by Maverick Capital, with participation from Point72 and Jericho, as well as existing investors Amplify Partners, Dell Technologies Capital, Lip-Bu Tan and others. SiMa.ai will utilize the $270M raised to date to continue meeting customer demand for edge AI/ML with its first-generation Machine Learning System-on-Chip (MLSoC) while accelerating the delivery of its second-generation MLSoC for release in Q1 of 2025.
With the release of the second-generation MLSoC, SiMa.ai will unveil one software-centric platform for all edge AI. SiMa.ai’s ‘one platform for all edge AI’ scales with customers as their AI/ML journey evolves, from computer vision, to transformers to multimodal generative AI.
The rapid proliferation of generative AI is simplifying the way we interact with AI/ML, namely through multimodal inputs such as text-to-speech, text-to-image, speech-to-text, speech-to-image, audio-to-image, image-to-image and image-to-video. Edge devices such as robots, drones, diagnostic machines, autonomous vehicles and more must be equipped to handle this new, more collaborative interface. The first-generation SiMa.ai MLSoC delivers industry leading performance and efficiency for vision-centric edge inference. Future generations will power and process any modality customers prefer with even greater performance and power efficiency. Uniquely designed to help customers capitalize on the opportunity that the shift of generative AI to the edge represents, SiMa.ai’s MLSoC’s unique fusion of hardware and software processes any model size or type on its silicon architecture.
“AI — particularly the rapid rise of generative AI — is fundamentally reshaping the way that humans and machines work together. Our customers are poised to benefit from giving sight, sound and speech to their edge devices, which is exactly what our next generation MLSoC is designed to do,” said Krishna Rangasayee, Founder and CEO at SiMa.ai. “We have established undeniable technology leadership with our first generation MLSoC and with that momentum, also recognized the imminent need to equip our customers with one software-centric platform that supports all modalities from computer vision to GenAI. To that end, we’re accelerating our ability to execute and have strengthened our investor base even further with the addition of Maverick Capital, Point72 and others, as we remain laser focused on powering the edge the world relies on.”
“The computational intensity of generative AI has precipitated a paradigm shift in data center architecture. The next phase in this evolution will be widespread adoption of AI at the edge. Just as the data center has been revolutionized, the edge computing landscape is poised for a complete transformation,” said Andrew Homan, Senior Managing Director at Maverick Capital. “SiMa.ai possesses the essential trifecta of a best-in-class team, cutting-edge technology and forward momentum, positioning it as a key partner for customers traversing this tectonic shift. We are excited to join forces with SiMa.ai to seize this once-in-a-generation opportunity.”
SiMa.ai MLSoC: One Software-Centric Platform for All Edge AI
The current generation of SiMa.ai’s MLSoC consistently demonstrates embedded edge market leadership for performance and power efficiency, validated most recently by the MLCommons® MLPerf Inference benchmark (4.0) results, where SiMa.ai competed in the Inference 4.0 closed, edge, power division category. SiMa.ai’s best in class results in FPS/W in these benchmarks are a result of continuous innovation to MLSoC and Palette software to optimize end-to-end model execution and performance. For more detailed information about recent updates to Palette 1.2, which include expanded support for C++ and support for developing on Windows, read here.
Made possible by SiMa.ai’s proprietary combination of silicon and software, the MLSoC adapts to any framework, any network, any model or sensor, and with the addition of Gen 2, any modality: audio, speech, text, image and more. SiMa.ai is poised to address customers’ computer vision, transformer and multimodal generative AI needs all in one, software-centric platform.
Utilizing one platform for all edge AI, SiMa.ai customers will benefit from a range of high performance and power optimized AI/ML made available through multiple generations of the SiMa.ai MLSoC. For more information on SiMa.ai’s one software-centric platform for all edge AI, please read The Next Chapter at SiMa.ai: One Platform for All Edge AI on the SiMa.ai blog.
The second-generation SiMa.ai MLSoC includes innovations from Arm, Synopsys (Nasdaq: SNPS) and TSMC, specifically:
- An Arm® Cortex®-A CPU, which is a key component of SiMa.ai’s heterogeneous compute platform for overall application development. Arm’s large, diverse software partner ecosystem further extends SiMa.ai’s reach to customers with specialized use cases.
- Synopsys EV74 Embedded Vision Processors, which enable critical pre-and-post processing in computer vision applications on a single chip. Synopsys full-stack AI-driven EDA suite and ZeBu emulation system further accelerated the design, verification and software bring-up of MLSoC.
- TSMC’s N6 technology, which offers further performance and power optimizations for MLSoC customers.
“As edge AI scales, the ecosystem is navigating growing software complexity and ever-increasing demand for AI performance, and is converging on Arm as the computing platform of choice from cloud to edge,” said Paul Williamson, SVP and GM IoT Line of Business, Arm. “It’s exciting to see continued innovation from partners such as SiMa.ai in this space, and we look forward to continued collaboration as we build the future of edge AI on Arm.”
“Synopsys celebrates the success of SiMa.ai as the company continues to push the limits of AI/ML computing at the embedded edge,” said Ravi Subramanian, general manager of the Systems Design Group at Synopsys. “Our close collaboration with SiMa.ai across Synopsys’ full-stack AI-driven EDA suite, broad IP portfolio including the Synopsys ARC EV Vision Processor, and hardware-assisted verification solution has enabled SiMa.ai to achieve first-pass silicon success and accelerate development of their high-performance, power-efficient AI designs.”
“The first-time silicon success of SiMa.ai’s first-generation MLSoC marks a significant milestone in our partnership, delivering an embedded edge ML platform with industry-leading performance and efficiency,” said Sajiv Dalal, Executive Vice President of Business Development at TSMC North America. “We are committed to supporting innovators like SiMa.ai, harnessing our most advanced technologies to enable differentiated solutions that meet the rapidly evolving demands in AI.”