SiMa.ai, a Silicon Valley-based startup producing an embedded machine learning system-on-chip (SoC) platform, today announced it has raised $70 million in extended funding as it plans to launch a second-generation chipset built specifically for intermodal transportation generation artificial intelligence processing and bring it to the market.

According to Gartner, the global AI-enabled chip market is expected to more than double from 2023 to $119.4 billion by 2027. However, only a few manufacturers have begun producing specialized semiconductors for AI applications. Most notable contenders are initially focused on supporting AI in the cloud. Still, various reports predict significant growth in the edge AI market, meaning the hardware handling AI computations is closer to the source of data collection than in a centralized cloud. SiMa.ai, named after Seema (Hindi for “border”), is committed to capitalizing on this shift by delivering its edge AI SoCs to organizations in industrial manufacturing, retail, aerospace, defense, agriculture and healthcare.

Targeting the market segment between 5W and 25W power consumption, the San Jose-based startup launched its first ML SoC to introduce AI and ML through an integrated software and hardware combination. This includes its proprietary chipset and no-code software called Palette. Krishna Rangasayee, founder and CEO of SiMa.ai, told TechCrunch that the combination is used by more than 50 companies around the world.

The startup claims that its current-generation ML SoC delivers the highest FPS/W results in the MLPerf Inference 4.0 MLPerf benchmark across closed, edge and power partition categories. However, the first generation of chipsets focused on classic computer vision.

As demand for GenAI continues to grow, SiMa.ai plans to launch its second-generation ML SoC in the first quarter of 2025, focusing on providing customers with multi-modal GenAI capabilities. Rangasayee said the new SoC will be a “revolutionary change” from its predecessor and feature “some architectural tweaks” on the existing ML chipset. The basic concept will remain the same, he added.

The new GenAI SoC will adapt to any framework, network, model and sensor – similar to the company’s existing machine learning platform – and will also be compatible with any modality, including audio, speech, text and image. The startup said it will serve as a single edge platform for all artificial intelligence across computer vision, Transformer and multi-modal GenAI.

“You can’t predict the future, but you can pick a vector and say, hey, this is the vector I want to bet on. I want to continue to evolve around my vector. That’s the approach we’re taking architecturally,” Rangasayee said. “But fundamentally, we really haven’t given up or had to completely change our architecture. That’s also the benefit of us adopting a software-centric architecture, which allows for greater flexibility and agility.”

SiMa.ai’s first- and second-generation AI chipsets are powered by Taiwan’s Taiwan Semiconductor Manufacturing Co., Ltd. (TSMC) as its manufacturing partner and Arm Holdings as its computing subsystem supplier. The second-generation chipset will be based on TSMC’s 6nm process technology and include the Synopsys EV74 embedded vision processor for pre- and post-processing in computer vision applications.

The startup counts existing players such as NXP, Texas Instruments, STMicroelectronics, Renaissance, Microchip Technology, Nvidia, and AI chip startups such as Hailo as competitors. However, it considers Nvidia to be a major competitor – just like other AI chip startups.

Rangasayee told TechCrunch that while Nvidia is “excellent in the cloud,” it hasn’t built an edge platform yet. He believes Nvidia lacks sufficient power efficiency and edge AI software. Likewise, he asserted that other startups building AI chipsets don’t solve system problems but simply provide machine learning acceleration.

“Of all our peers, Hailo is doing very well. It’s not that we are better than them. But from our perspective, our value proposition is completely different,” he said.

The founders went on to say that SiMa.ai offers higher performance and better energy efficiency than Hailo. He also said that SiMa.ai’s system software is very different from GenAI and is very effective.

“As long as we can solve customer problems and we’re better at it than anyone else, we’re in a good position,” he said.

SiMa.ai’s latest all-equity funding round, led by Maverick Capital with participation from Point72 and Jericho, expands on the startup’s $30 million Series B originally announced in May 2022. Existing investors including Amplify Partners, Dell Technologies Capital, Fidelity Management and Lip-Bu Tan also participated in additional investments. With this financing, the five-year-old startup has raised a total of $270 million.

The company currently has 160 employees, 65 of whom work at its R&D center in Bangalore, India. SiMa.ai plans to increase its headcount by adding new positions and expanding its research and development capabilities. It also hopes to build a go-to-market team for Indian customers.Additionally, the startup plans to expand its customer-facing team globally, starting with South Korea and Japan, as well as Europe and the United States

“The computational intensity of generative AI has triggered a paradigm shift in data center architecture. The next stage in this evolution will be the widespread adoption of AI at the edge. Just as the data center has been revolutionized, the edge computing landscape is about to be revolutionized as well. Transformation. SiMa.ai’s combination of a best-in-class team, cutting-edge technology and forward momentum make it a key player as customers experience this tectonic shift. We are excited to join forces with SiMa.ai to seize this once-in-a-lifetime opportunity opportunity,” Andrew Homan, senior managing director at Maverick Capital, said in a statement.

#SiMa.ai #raises #US70 #million #financing #launch #multimodal #GenAI #chip

Leave a Reply

Your email address will not be published. Required fields are marked *