The 10 Hottest Semiconductor Startups Of 2023 (So Far)

These semiconductor startups could benefit from the hype around generative AI that has some companies looking at alternatives to GPU powerhouse Nvidia and better ways to architect systems. Some of these young companies are working on foundational or adjacent technologies that could improve the way AI work is computed in the future.

While the semiconductor industry banks on a late 2023 comeback after several months of waning demand, the hype around generative AI has rejuvenated interest in specialized processors and other semiconductor technologies that can run such workloads faster and more efficiently.

Nvidia has been the largest beneficiary of demand for the explosion in demand for AI chips and adjacent technologies. The company in May said it expects revenue in the second quarter to grow 64 percent from the same period last year, largely due to interest driven by generative AI technologies like ChatGPT.

[Related: How 16 Vendors Are Helping Partners Build AI Businesses]

But while many companies are turning to Nvidia to fuel AI workloads, the chip designer has apparently been experiencing a shortage of data center GPUs. This, along with a desire to test out new chip architectures to optimize for performance and efficiency, has renewed interest in competing technologies, including those designed by semiconductor startups like Lightmatter and Tenstorrent.

It’s not just chip design startups that stand to benefit from the new wave of interest in AI computing. There are also other semiconductor startups working on adjacent or foundational technologies that could improve the way AI work is computed in the future.

These upstarts, such as Ayar Labs and Eliyan, are working on new ways to build chips that could benefit AI and many other workloads. They also include companies like Astera Labs and Pliops that are addressing other aspects of high-performance computing systems such as storage and connectivity.

But it’s not just AI where semiconductor startups are finding opportunity. There’s also EdgeQ, which is taking on incumbents in the telecom chip space with its “Base Station on a Chip.”

These and other young companies make up CRN’s 10 hottest semiconductor startups of 2023 so far. What follows is an overview of each company and its recent accomplishments.

Astera Labs

Founder, CEO: Jitendra Mohan

Astera Labs seeks to capitalize on the new wave of servers entering the market with chip-based products that take advantage of Compute Express Link, a new connectivity standard that enables faster and more efficient servers with greater flexibility and lower costs.

Last November, the Santa Clara, Calif.-based startup announced that it had raised a $150 million Series D funding round led by Fidelity Management and Research, with participation by Intel Capital and other investors, at a nearly $3.2 billion valuation. The funding arrived as the company saw “strong” traction with customers.

Astera’s connectivity portfolio includes the Leo Memory Connectivity Platform, which it said is the “industry’s first” memory controller to support memory expansion, memory pooling and memory sharing features that can help increase performance and lower costs for servers.

Axelera AI

Co-Founder, CEO: Fabrizio Del Maffeo

Axelera AI wants to trounce Nvidia with the “world’s most powerful and advanced” chip-based offerings for running AI applications on edge devices.

The Netherlands-based startup announced in May that it had raised an “oversubscribed” Series A funding round that brought its total funding to $50 million.

The investment announcement came after Axelera struck a strategic partnership with edge device vendor Advantech, unveiled its next-generation Metis AI hardware and software platform, and taped out two AI chips in the previous 18 months.

Axelera said its expansion cards and modules, powered by the Metis AIPU, can provide “faster, easy-to-use AI acceleration while minimizing power and cost.”

Ayar Labs

CEO: Charlie Wuischpard

Ayar Labs believes it can usher in a new era of high-performance chips with silicon photonics technology that moves data through chips using light instead of electricity.

The Santa Clara, Calif.-based startup announced in May that it had raised an additional $25 million funding for its Series C round, bringing the round’s total to $155 million. The startup’s investors include Nvidia, Intel Capital, Hewlett Packard Enterprise’s venture arm and GlobalFoundries.

The extra investment arrived after Ayar achieved multiple milestones. These included a demonstration of the industry’s first optical offering for delivering a connectivity speed of 4 Tbps, a partnership with the U.S. Department of Defense to prototype military applications using its optical I/O chiplets and lasers, and a collaboration with Nvidia to develop AI chips with optical I/O.

EdgeQ

Co-Founder, CEO: Vinay Ravuri

EdgeQ wants to usurp semiconductor giants like Intel and the chip-making capabilities of telecom giants like Ericsson with its “Base Station-on-a-Chip” offering for a range of network applications.

The Santa Clara, Calif.-based startup in May revealed that it had raised a $75 million Series B funding round to invest in the development of its next-generation chip technology, which can be used as the foundation for public and private 5G and 4G cellular networks.

This year, network and telecom vendors began unveiling new products that will use EdgeQ’s chip. These include Taiwan’s Wistron, which said it has developed the “first software-defined, multi-mode 4G/5G small cell” using EdgeQ’s chip. The startup has also unveiled a 4G/5G small cell reference design with MaxLinear as well as a 4G/5G small cell architecture platform with Actiontec.

Other vendors that have backed EdgeQ include Vodafone, which plans to use its chip for next-generation, software-programmable 5G ORAN platforms. Network software vendor Mavenir also plans to use EdgeQ’s chip to develop a 4G/5G small cell offering.

Eliyan

Founder, CEO: Ramin Farjadrad

Eliyan thinks its chiplet interconnect offering can best advanced packaging technologies from chip giants like Intel and TSMC to enable the next wave of high-performance chip architectures.

The Santa Clara, Calif.-based startup last November disclosed that it had raised a $40 million Series A funding round from investors, including Intel Capital and Micron Capital, to develop and ramp up production of its NuLink die-to-die interconnect technologies.

In May, the company unveiled the availability of its NuLink PHY technology, which is manufactured using TSMC’s standard 5nm process to transport data at high speeds between chiplets without the drawbacks of advanced packaging technologies.

Lightmatter

Co-Founder, CEO: Nicholas Harris

Lightmatter said it plans to “bring a new level of performance and energy savings to the most advanced AI and HPC workloads” with its photonic chip-based technologies.

The Boston-based startup in May announced that it had raised a $154 million Series C round from several investors, including the venture arms of Hewlett Packard Enterprise and Google. The company plans to use the funding to equip cloud service providers, semiconductor companies and enterprises with its photonic technologies.

Lightmatter’s portfolio includes Envise 4S, a 4U server that includes 16 of the company’s Envise chips and only uses 3 kilowatts of power. The company said a rack-scale Envise inference system can run the BERT-Base SQuAD large language model at three times greater inferences per second and eight times greater inferences per second per watt than Nvidia’s DGX A100 system.

Pliops

Founder, CEO: Uri Beitler

Pliops said its Extreme Data Processor “overcomes storage inefficiencies to massively accelerate performance and dramatically reduce overall infrastructure costs for data-hungry applications.”

The San Jose, Calif.-based startup announced last August that it had raised a $100 million Series D funding round. The company’s investors include SK Hynix, Intel Capital, Nvidia, AMD and Western Digital.

In January, Pliops unveiled the launch of its XDP Data Services platform, which it calls a “seamless-to-deploy, transformational approach to optimize data infrastructure and accelerate modern workloads, while in tandem reducing [total cost of ownership] by 50 percent.”

Pliops’ Extreme Data Processor, the foundation of the new platform, brings together “key data and storage technologies into one device” that multiplies their benefits, according to the startup. These benefits include 10 times greater performance, five times greater reliability, six times greater capacity and up to 80 percent lower total cost of ownership compared with traditional RAID products.

SiMa.ai

Founder, CEO: Krishna Rangasayee

SiMa.ai said its edge AI technology combines “jaw-dropping performance” and “incredible power efficiency” with a “beautifully simple, push-button software experience.”

In June, the San Jose, Calif.-based startup said that its Machine Learning System on a Chip silicon for the embedded edge market entered mass production with the support of two PCIe-based production boards and Palette software for low-code machine learning development. The company also raised an additional $13 million in funding, bringing the total investment to $200 million.

SiMa.ai said its MLSoC beat Nvidia’s Jetson AGX Orin system-on-chip on power and latency with “stronger” frames per second per watt performance for inference, based on results submitted to MLPerf, a suite of machine learning benchmark tests for training and inference.

Tenstorrent

CEO: Jim Keller

Tenstorrent said it’s building the “future of AI hardware” with its machine learning accelerator chips, RISC-V CPU technology and AICloud compute service.

The Toronto-based startup said in May that it has partnered with LG to build a “new generation of AI and RISC-V chiplets to “potentially power” the Korean electronics giant’s premium TV and automotive products in the future. The two will also collaborate to integrate LG’s video codec technology into Tenstorrent’s future data center products.

The announcement came after Tenstorrent made appointments for key roles in the company’s leadership team. These included former Google and SiFive executive Keith Witek as COO, former Meta infrastructure hardware lead Olof Johansson as vice president of operating systems and infrastructure software, and former Intel GPU head Raja Koduri as a board member.

Ventana Micro Systems

Founder, Chairman, President, CEO: Balaji Baktha

Ventana Micro Systems wants to move the data center and edge markets away from x86 and Arm processors to the open standard RISC-V chip architecture with high-performance CPU designs.

The Cupertino, Calif.-based startup in December revealed the Veyron V1 chip, which it called the “highest-performing” RISC-V chip with a frequency of 3.6GHz. The company said the chip is also the “first RISC-V processor to provide single thread performance that is competitive with the latest incumbent processors for data center, automotive, 5G, AI and client applications.” The Veyron V1 has the “highest single-socket performance among competing architectures,” too, Ventana said.

The company revealed the processor—which it plans to offer in a chiplet form that other companies can integrate into chip packages—after raising $55 million in funding from investors earlier in 2022.