02 Dec Nuvia Datacenter Processor Company Comes Out Of Stealth And It Looks Provocative
I have been in and around processors for nearly 30 years as a systems OEM, search engine, chipmaker and now CEO of Moor Insights & Strategy. I tell anyone who will listen that semiconductors are sexy and it doesn’t take much convincing nowadays. You just need to look to the trillion-dollar (or near) club Apple, Amazon, Google, and Microsoft to see how these companies are leveraging custom silicon to enhance their competitive advantages. While I normally don’t cover silicon startups, I am going to make an exception with NUVIA as I believe the company is very unique.
NUVIA came out of stealth Friday with a $53M round A funding round led by Capricorn Investment Group, Dell Technologies Capital, Mayfield, Nepenthe LLC and WRVI Capital. What on earth would compel this grade of VCs to invest nearly $53M in what is currently 60 people?
After meeting with NUVIA founders John Bruno, Manu Gulati and Gerard Williams III over the past few months, I believe it comes down to principals who have a consistent track record of success on processors, SoCs and systems. Were you impressed by those 20% Apple iPhone and iPad performance improvements? Well… John Bruno likely modeled the performance gains, Gerard Williams was likely its CPU architect, and Manu Gulati was likely its SoC architect. I may have been critical of some of Apple’s dealings with its silicon suppliers, but I always had the utmost respect for its SoCs performance pace. Together, the three NUVIA principals have 100 patents granted spanning 20 chips developed at AMD, Apple, Arm, Broadcom and Google. An impressive pedigree.
Intel currently dominates the high-performance, general-purpose datacenter processor market, AMD is gaining share and IBM POWER is a smaller volume but very high-performance player in accelerated workloads. Arm-based designs are prevalent in networking, storage, and server rack offload like at AWS. If you look at the entire datacenter and edge market, I believe NUVIA is looking at a $35B market opportunity, conservatively.
NUVIA is targeting the datacenter market, which may sound interesting, given the founders’ success in mobility, but then again, these were Apple SOCs which were anything but “small cores.” Also consider the competitive field is thinner in the highest performing datacenter silicon market. When I look at the track record of the team, I am convinced there’s a good likelihood it will be able to power through walls other datacenter CPU startups hit, which had a hard time executing a “big core” strategy. Many of these startups didn’t have a lot of experience in big cores, but I believe this team knows how to architect and deliver a “big core” in an SOC.
One of NUVIA’s co-founders John Bruno reiterated one of the differences with company’s approach:
This company is targeting performance leadership, which isn’t a narrative we’ve heard much before from some of the past and present new entrants in the server space. We’re taking a clean-sheet approach and building a custom CPU design for the hyperscale world, without the legacy infrastructure or overhead.Nuvia co-founder John Bruno
While I was impressed at the overall level of VCs, I was most intrigued about Dell Technologies Capital. Dell Technologies knows a little about the datacenter (sarcasm added). It has the #1 share in servers, HCI systems, external storage, all-flash arrays, virtualized network ports, and server virtualization software. In the NUVIA release, Scott Darling, President of Dell Technologies Capital, said: “The best companies start when founders with outstanding track records of performance come together to identify a big problem and line up the best investment team to help them succeed. As part of that team, we’re focused on providing NUVIA with DTC’s unique value and market leverage.”
This is interesting, right? I don’t want to put words in Darlings mouth, but I believe that “big problem” he is referring to is the unfulfilled need of consistent large CPU performance gains at increasingly lower power draw. As a reaction to the long-term slowdown in raw server CPU performance, the industry has reacted with every accelerator under the sun, including GPU, FPGA, DSP and ASIC. I’m not saying those wouldn’t be needed if CPU vendors had delivered more performance sooner and consistently, but it would have likely hastened the timing and many of the requirements.
While it will likely be three years before we see final, production silicon given the company was founded early this year, I believe this is one of the most interesting, if not the most interesting chip startups out there right now. The company has a lot to prove and execute over those next three years, but this is a good start as the team has a strong pedigree in high-performance designs backed by a high-quality set of investors. Stay tuned as I will be watching and analyzing with interest. Here we go.