SiMa.ai Secures $70 Million to Launch Next-Gen Multimodal Generative AI Chipset
Santa Clara, CA – April 4, 2024 – SiMa.ai, a pioneering Silicon Valley startup specializing in edge AI system-on-chip (SoC) platforms, announced today a $70 million extension to its Series B funding round. This fresh capital injection fuels the company’s development and upcoming launch of its second-generation chipset, specifically designed for multimodal generative AI processing at the edge.
The global market for AI-enabled chips is projected to double by 2027, reaching a staggering $119.4 billion according to Gartner [1]. However, only a handful of companies are currently developing dedicated semiconductors specifically tailored for edge AI applications. Traditionally, AI processing has resided within centralized cloud environments. However, there’s a growing trend towards “AI on the edge,” where hardware performs AI computations closer to the source of data collection, bypassing the cloud entirely.
SiMa.ai, derived from the Hindi word “Seema” meaning “boundary,” capitalizes on this shift by offering its cutting-edge AI SoCs to organizations across diverse industries like industrial manufacturing, retail, aerospace, defense, agriculture, and healthcare.
The San Jose-based company, targeting the 5W-25W power usage segment, introduced its first ML SoC, seamlessly integrating AI and machine learning through a unique software-hardware combination. This includes SiMa.ai’s proprietary chipset and their no-code software suite called Palette. Krishna Rangasayee, founder and CEO of SiMa.ai, revealed that over 50 companies worldwide have already adopted this powerful combination.
SiMa.ai’s current ML SoC boasts the highest FPS/W results on the MLPerf benchmark across various categories, including MLPerf Inference 4.0 closed, edge, and power divisions. However, this first-generation chipset primarily focused on traditional computer vision tasks.
To address the burgeoning demand for generative AI (GenAI), SiMa.ai is poised to unveil its second-generation ML SoC in Q1 2025. This innovative chipset prioritizes equipping customers with multimodal GenAI capabilities.
“The future is uncertain, but we can choose a direction and persistently refine our approach around it,” said Rangasayee. “This philosophy guided our architectural decisions. We haven’t fundamentally overhauled our architecture; the core concepts remain. This is a testament to the flexibility and agility inherent in our software-centric approach.”
SiMa.ai leverages industry leaders like Taiwan’s TSMC for manufacturing both its current and next-generation AI chipsets. Arm Holdings supplies the compute subsystem. The upcoming second-generation chipset will be built on TSMC’s 6nm process technology and incorporate Synopsys’ EV74 embedded vision processors for pre- and post-processing in computer vision applications.
SiMa.ai acknowledges established players like NXP, Texas Instruments, STMicro, Renesas, Microchip Technology, and Nvidia as competitors, along with fellow AI chip startups like Hailo. However, they identify Nvidia as their primary competitor, echoing a sentiment shared by many AI chip startups.
According to Rangasayee, while Nvidia excels in cloud computing, they lack a dedicated edge AI platform. He contends that Nvidia falls short in power efficiency and software specifically designed for edge AI applications. He further asserts that other AI chip startups often neglect to address broader system-level challenges, solely focusing on ML acceleration.
“Hailo stands out amongst our peers,” said Rangasayee. “We’re not claiming to be definitively better, but our value proposition is demonstrably different.”
Rangasayee emphasizes that SiMa.ai delivers superior performance and power efficiency compared to Hailo. Additionally, he highlights SiMa.ai’s system software as a key differentiator, particularly well-suited for GenAI applications.
“As long as we effectively solve customer problems and deliver unmatched value, we’re well-positioned for success,” he concluded.
SiMa.ai’s latest all-equity funding round, led by Maverick Capital with participation from Point72 and Jericho, extends their $30 million Series B round announced in May 2022. Existing investors including Amplify Partners, Dell Technologies Capital, Fidelity Management, and Lip-Bu Tan also participated in the additional investment. This fundraising brings the five-year-old startup’s total capital raised to $270 million.
Currently, SiMa.ai employs a workforce of 160, with 65 engineers stationed at their R&D center in Bengaluru, India. The company plans to significantly expand its headcount by adding new roles and bolstering its R&D capabilities. Additionally, they aim to establish a dedicated go-to-market team specifically focused on Indian customers.
SiMa.ai : Global Expansion and A Look Ahead
SiMa.ai’s growth trajectory extends beyond funding and product development. They are strategically scaling their global presence to capitalize on the burgeoning demand for edge AI solutions. Here’s a glimpse into their expansion plans:
- Targeting Korea, Japan, Europe, and the US: SiMa.ai recognizes the vast potential in these markets and plans to establish robust customer-facing teams to cater to regional needs. This ensures they can effectively support a wider clientele and accelerate market penetration.
- Building a Strong R&D Presence in India: The company acknowledges the talent pool and strategic location benefits of India. By expanding their R&D center in Bengaluru, they can harness local expertise and fuel further innovation.
SiMa.ai’s future hinges on its ability to capitalize on several key trends:
- The Rise of Edge AI: As the limitations of cloud-based AI become more apparent, edge AI is poised for explosive growth. SiMa.ai’s focus on this space positions them to become a leader in this transformative technological shift.
- The Power of Generative AI: Generative AI holds immense potential across various industries. By offering a next-generation SoC specifically designed for multimodal GenAI, SiMa.ai equips businesses to leverage this powerful technology.
- Software-Centric Approach: SiMa.ai’s unique software-centric architecture offers greater flexibility and nimbleness compared to traditional hardware-focused solutions. This approach empowers them to adapt to evolving customer needs and technological advancements.
SiMa.ai’s latest funding round, coupled with their innovative technology and strategic expansion plans, positions them as a major player in the rapidly evolving edge AI landscape. Their focus on multimodal generative AI and software-centric approach equips them to address the challenges and opportunities that lie ahead. As the demand for edge AI solutions continues to surge, SiMa.ai is well-positioned to shape the future of intelligent computing at the edge.