I had the privilege of sitting down with SiMa.ai board member Daniel Docter, who is also Managing Director at Dell Technologies Capital and Brett Rosenbaum, Partner at Adage Capital Management. Both investors believe in our ability to disrupt the embedded edge market with our industry-first MLSoC platform.
Daniel invests in data and data analytics, AI/ML, cloud management and orchestration software as Managing Director of Dell Technologies Capital. Daniel has worked in venture capital for the past 23 years. He has an impressive technical background with an ECS degree and Ph.D. in solid state physics.
Brett is a Partner at Adage Capital Management. He manages investments for public/private technology companies focused on semiconductors, hardware, and communication equipment. Specifically, Brett has successfully run a semiconductor portfolio for the past 14 years at Adage. In this discussion, both Daniel and Brett share their views of the embedded edge market and why they think SiMa.ai is the best positioned to solve the embedded edge market’s software problem.
Krishna: Thanks for joining us Daniel and Brett. Given your roles as venture investors, let’s begin with some market observations and context before we get into talking about SiMa.ai. How do you view the macroeconomics of the general market?
Daniel: I think it’s a tale of two cities. There are certain areas in the market that have become very hot. One obvious example is Generative AI (large language and foundation models), where the market has become overhyped, with an excess of funding and inflated valuations. But in many other areas of the market, companies are delaying their fundraising as long as possible to try to continue to grow and “catch up” to their last valuation from the pandemic era. Later stage companies can more easily manage this, but earlier stage companies with no real customer and revenue traction may be forced to sell or shut down.
Krishna: Daniel, what are your thoughts on investments made during COVID compared to investments happening today?
Daniel: The pace of venture capital investment since COVID has slowed dramatically. The statistics for Q1 2023 show VC investment is down to about ⅓ of the peak COVID investment pace. That statistic indicates the massive decline that we saw from the height of the pandemic. As I mentioned earlier, companies are trying to stretch themselves further before they go back to market. I would describe it as a “bid/ask” problem. There’s a bid, and there’s an ask, and there’s the gap and that gap is too large. But the gap will eventually close and deals will get done. It’s just a matter of time.
Krishna: Thank you, Daniel. Brett, what’s your outlook on the macroeconomics of the general market? You’ve worked with both public and private companies, so I’d love to hear your sentiment.
Brett: The SPAC parade that’s taken place over the past couple of years is a tremendous bubble that will pop. We haven’t seen these SPACS go under yet because they are living out the SPAC proceeds; we call it a market of zombies because it includes companies that should be dead but aren’t. As those companies go under, it’s going to tighten the funding belts, but we haven’t seen that yet.
The reality is, we must get through these hard economic times to realize that money should have never been invested in companies that weren’t ready to go public. From where I sit, the biggest trend I see are companies becoming platform companies. It’s very much widening the moat. Having software, having the platform, having a turnkey solution is paramount to survive in the public market.
Krishna: Now, a lot of companies are operating around a 40-50 percent haircut, do you see them staying at that percent for a while? Or do you think the market will rebound?
Brett: Unfortunately, I don’t see a nice path forward anytime soon, although there might be some hard winners.
Krishna: I’m hearing from my fellow startup founders and entrepreneurs that they expect a very difficult journey, especially this year. Do you both share the same sentiment?
Brett: I see it a little differently because there are still so many powerful trends like autonomous, EVs, and automation. If the United States really wants to globalize, we can’t do it with the same human workforce, there needs to be automation, there needs to be more technology. With these technology requirements comes a ton of opportunity for those who are nimble. I don’t think it’s so doom and gloom.
Daniel: Great thought, Brett. I’m a technology person so I always think it’s going to be technology that leads us out of these tough times. That being said, I do believe it will be harder this year and going forward in the near future. But I don’t see it being so doom and gloom either. It’s going to be a tough road and we need to focus on the improvements we see from quarter to quarter.
Krishna: Okay, thank you. A little closer to home for us. How do you read the ML hardware market today? And more specifically to SiMa.ai, how are we positioned to disrupt it?
Daniel: At Dell Technologies Capital, we looked at 32 ML hardware companies before we invested in SiMa.ai. Many of those companies weren’t sustainably differentiated or compelling in our view. They were founded by amazing and super smart people, who were funded by great investors, and are doing really interesting things, but nothing looked “lasting” to us.
I divide the machine learning chip market up into three parts: the data center market, the cellphone (or far edge / consumer) devices market, and the embedded edge market. The embedded edge market is diverse but huge in aggregate, and definitely less served by high quality chip and solution providers.
SiMa.ai stood out to us because of its team’s deep understanding of the embedded edge space, knowing the architecture needed to enable customers to switch over easily and cost effectively, and support all existing software. I go into more detail about our investment thesis in SiMa in this blog: Software Solves a Silicon Problem, How a SiMa.ai Chip Outperformed the Industry Leader.
Really, the ML Hardware market isn’t so different from other venture capital markets; there’s always 20-30 companies that are funded to try to solve a new problem, and there are 1-2 that go public, and another 1-2 that get acquired, and then the rest go away. That’s the game.
Krishna: Brett, how do you read the ML Hardware market?
Brett: First thing is you’re going to need to know and understand who the competitors are. If you’re competing with goliaths like NVIDIA, Intel etc. a lot of money and time is going to have to go into it. Know who your competitors are and what it takes to compete with them. Then finding what’s unique in the market. For the embedded edge market, a system on a chip that can solve the embedded edge software problem is unique and stands out to us.
We were most excited about SiMa.ai of its pedigree – I knew you and Gopal from Xilinx and the success you had there. And again, the uniqueness of SiMa.ai’s MLSoC Platform and its ability to solve the software problem. No other technology can do this yet; that is what excited me.
Krishna: Daniel, how about you?
Daniel: We view SiMa.ai as a company that has great people, a great vision and is already executing to solve a problem. There’s nothing better than seeing a silicon company deliver its very first chip with breakthrough performance, and that can ship for customer revenue. That matters a lot. The Sima story around push button, 10x, any model is really compelling and that resonates with customers.
Krishna: How do you think we are doing now after already investing in us?
Brett: We have seen near flawless execution and discipline from SiMa.ai. Most recently, you’ve made some incredible and strategic hires, like Harry Kroger to lead your automotive business. You’ve met your timelines and like Daniel said, have executed on everything. And it’s not easy, yet you’ve made it look easy. To have an A-zero chip that works is a huge accomplishment.
Krishna: Thank you both for this conversation and your continued trust in us. Looking forward to speaking again soon!