Nic Brathwaite, founding CEO of Celesta Capital, is optimistic about the global semiconductor industry. “Great technical innovations are based on semiconductors,” explained Brathwaite in an interview. He modestly says there wasn’t enough VC money available for hardware startups, leaving Celesta “picking from the litter”. Brathwaite explained that Celesta has three investment themes:
- Technology platforms that enable mass diffusion of technology. There are many new applications falling under this theme, including AI (Artificial Intelligence), IoT (Internet of Things), and technologies that are transforming existing industries such as construction, education, and agriculture.
- Technologies that transform and update existing technology platforms.
- Technologies at the intersection of high tech and biotechnology that Brathwaite calls “bioconvergence.” An example of this type of convergence is medical diagnostics.
Not surprisingly, Brathwaite says he and his partners are big fans of the US CHIPS and Science Act. “Semiconductors are becoming increasingly important to the global economy,” he says, “and it makes sense for governments to pay attention.” Although semiconductors may not account for a large portion of a country’s GDP (gross domestic product), the portion of a country’s GDP generated by semiconductors is enormous and continues to grow.
While semiconductors have become ubiquitous in all sectors of the economy (industrial, commercial, agricultural, medical, consumer, military, aerospace and government), the supply chain has consolidated into fewer and fewer semiconductor manufacturers and OSATs (Outsourced Semiconductor Assembly and Test Providers). Brathwaite says this consolidation brings with it a huge global risk in terms of vulnerability. The failure or elimination of just one semiconductor manufacturer or OSAT puts the entire semiconductor ecosystem at risk.
Brathwaite’s bullish stance on hardware and semiconductors is not surprising. His first job in the 1980s was as a semiconductor process engineer at Intel. After Intel, he co-founded nChip, became its VP of Operations and Technology Development, and worked on the development of an advanced multi-chip module assembly process. A little over five years later, Flextronics acquired nChip and Brathwaite became the company’s CTO, where he helped transform the contract manufacturer into Flex, an EMS (Electronic Manufacturing Services) and product development company. Brathwaite is unusual among VCs (venture capitalists) and knows the semiconductor and electronics industry inside out.
After 12 years at Flex, Brathwaite founded Riverwood Capital in 2008 and became a VC. His first three investments at Riverwood were Aptina, Ambarella and GoPro. DRAM manufacturer Micron Technology released Aptina in 2008 to manufacture CMOS imagers based on the technology developed at Micron. Founded in 2004, Ambarella develops video compression and image processing SoCs (Systems on Chips). Founded in 2002, GoPro is the famous maker of compact digital action cameras that, unsurprisingly, feature Aptina image sensors and Ambarella image processors. Aptina was later acquired by On Semi, which continues to make CMOS image sensors. Ambarella continues to make SoCs for video and imaging, and GoPro continues to lead the action camera market it created. All three companies were and are very successful.
Brathwaite left Riverwood in 2013 as the company’s focus shifted to today’s most popular startup investment targets: software and internet startups. He co-founded a new VC firm, Celesta Capital, with Michael Marks (also of Riverwood and Flex), Sriram Viswanathan (also of Intel background) and Lip-Bu Tan (founder of Walden Capital and former CEO of Cadence Design Systems). . Celesta continues Brathwaite’s preferred focus on investing in early-stage deep tech startups. His preference is for hardware startups, which is no surprise given his previous VC background, and he’s clearly had success investing in hardware companies. By the end of 2013, Celesta had invested in thirteen semiconductor startups. Celesta’s semiconductor investments now include:
- Alif Semiconductor – secure, low-power and high-performance edge processors
- Altair – LTE chipsets acquired by Sony
- Annapurna Labs – Hypervisor and network stack acceleration hardware acquired by Amazon
- Aquantia – High-speed Ethernet physical layer chips acquired by Marvell
- Aurasemi – IoT radio chips,
- CNEXLabs – NVME and open channel solid state disk controller chips
- Credo – High-speed SerDes transceiver
- Dust Photonics – high-speed optical transceivers
- Eliyan – Chiplet Technology
- EtherWhere – GPS chips
- Fungible – data processing units for disaggregated storage in the data center
- Habana Labs – Artificial Intelligence (AI) processors acquired by Intel
- HiDeep – Touch Interface Chips, IPO
- Ineda – Low power automotive and IoT processors acquired by Intel
- Innovium – Ethernet switch chips for data centers acquired by Marvell
- Lion Semiconductor – Battery Management Chips
- Nebulon – hybrid, cloud-defined data storage
- Netspeed Systems – Network on Chip (NoC technology, adopted by Intel
- Nuvia – Arm-based CPU cores inherited from Qualcomm
- ProteanTecs – Semiconductor IP for failure prediction
- Quantenna – High performance WiFi chips acquired from ON Semi
- RF Pixels – Beamforming IP for 5G acquired from Skyworks
- Recogni – AI-based image processing systems for automotive applications
- SambaNova Systems – generative AI platforms
- Stathera – Microelectromechanical (MEMS) timing chips
- Talgore Technology – GaN (gallium nitride) semiconductors.
- VeriSilicon – System-on-Chip (SoC) design services and IP, IPO
Note that Celesta has exited several of these investments through IPOs and/or acquisitions.
During his interview, Brathwaite noted that the US is still at the forefront of semiconductor innovation, on the design side but not on the manufacturing side. “Fabless semiconductor companies don’t innovate on the process side,” he said. In Brathwaite’s view, the US is at risk of losing its market leadership in devices if it lacks sufficient semiconductor manufacturing capacity.
Although the U.S. CHIPS and Science Act emphasizes the importance of semiconductors to national security and reduces the risk of losing additional U.S. semiconductor manufacturing capacity by providing government subsidies, Brathwaite believes the U.S. government can still be more innovative when it comes to supporting domestic semiconductor production. For one, says Brathwaite, the US government could cut through the VC community’s queues and provide non-dilutive funding alongside VCs. Perhaps this opinion isn’t particularly surprising coming from a VC. But Brathwaite goes further, pointing out that several countries, including Canada and China, have sovereign wealth funds that invest directly in companies. Six of these funds (Singapore, China, United Arab Emirates, France, Norway and Saudi Arabia) have assets of $1 trillion or more. The US has no such fund, he says. Although eleven US states have sovereign wealth funds, they are all much smaller than the funds supported by the states.
Brathwaite sees US investment in domestic semiconductor manufacturing as a given as it would create many good paying jobs in the US and therefore represents a solid investment for the future. As a VC specializing in semiconductor startups, Celesta’s investments clearly align with Brathwaite’s optimistic views on the US CHIPS and Science Act.
Tirias Research wholeheartedly agrees with Brathwaite. If semiconductors are a strategic asset, which seems self-evident given their pervasiveness in every industry, then the US government must help the country retain that asset. Adam Smith’s invisible hand making this asset managed like a commercial commodity by default has predictably led to a mass relocation of semiconductor manufacturing to other parts of the world. The US CHIPS and Science Act represents a single, albeit large, lever for the industry. It is unlikely that this single attempt will be enough to maintain US leadership in the semiconductor space. Regular consultations with experts from the VC community like Brathwaite and CEOs of semiconductor companies like Intel’s Pat Gelsinger, AMD’s Lisa Su and Nvidia’s Jensen Huang represent a unique and valuable knowledge resource that offers the US government far more insight into the complicated ecosystem could become what the global semiconductor industry has become and possibly lead to better strategic investment decisions in the future.