Article by Ryan Engel, Isabel Glass
Aug 6, 2024
Hardware is hard. That’s part of what makes it a great investment. Software may seem less capital-intensive and easier to scale, but “easy” has its disadvantages; easy results in low differentiation and by extension, high sales and marketing costs.
Hardware is hard. That’s part of what makes it a great investment.
Software may seem less capital-intensive and easier to scale, but “easy” has its disadvantages; easy results in low differentiation and by extension, high sales and marketing costs. In contrast, hardware’s barriers to scale create competitive moats and are difficult to replicate. When combined with intellectual capital and complex software integration, these moats become even more defensible. As a result, while capital requirements differ, hardware and pure software are just as likely to create fund returning outcomes (1).
Hardware is misunderstood and in turn, its value is underestimated; Hardware is the single greatest limitation on software innovation, presenting an opportunity for founders and investors alike.
A History of Hardware in Venture
As SaaS companies have boomed in recent decades, venture dollars have been increasingly biased against hardware. US VC deals in consumer electronics have been declining for nearly a decade (2). In the last 2 years of seed stage deals, hardware investing has been as much as 10x less competitive than software investing (1).
Despite the difference in deal volume, long-term data indicates that hardware and software returns are just as likely to produce 10x returns, and there is no evidence that hardware startups take longer to exit than their software counterparts (3). Instead, the hardware/software venture gap is likely attributed to a lack of experience and understanding of hardware among investors.
As a result of this funding gap, and to overcome lower survival rates, hardware companies have had to be more resourceful with funding. They increasingly rely on non-dilutive funding sources, particularly in key areas of government investment, like climate change (e.g., fusion reactors, carbon capture, mining robots, EV charging networks, solar developers, etc.).
The history of Silicon Valley, built on hardware ventures (literally referring to the “silicon” used to create commercial semiconductors), demonstrates that capital-intensive investments can pay off in the long term. In terms of venture capital raised, many hardware titans are actually much more capital lean than software giants; Uber alone raised $15B vs. NVIDIA and SpaceX's raising of $12B and $9.4B respectively (4).
A Hardware-driven Cycle of Innovation
Hardware and software are inextricably linked; innovation in software is not possible without innovation and investment in hardware.
Over the past five decades, there’s been significant improvement in the underlying computational power and storage hardware. As Moore’s Law predicted, every 2 years, we’ve approximately doubled computing capability (through CPUs) at continuously declining costs and energy usage over time (5). Software advancements have only been able to demonstrate strong economies of scale because of this foundation.
However, as Moore's Law begins to slow, CPU performance is now increasing linearly and unable to keep up with software growth. The cost of hardware has increased and development timelines have lengthened, impeding efficiency. As a result, new custom hardware models will be necessary in order to power the next wave of software advancements, particularly in AI.
CPUs, which have sufficed in the past, will need to be supplemented by DPUs (Data Processing Units) and GPUs (Graphics Processing Units, both optimized for power efficiency. DPUs can accelerate networking, data encryption, key management, and storage visualization, while GPUs are essential for machine learning, video editing, and gaming applications.
Without a hardware-driven cycle of innovation, software advancements may begin to slow and then, soon, peak. Kevin Deierling of Nvidia underscored this point: “The modern datacenter is software-defined and hardware-accelerated. They go hand in hand.”
Conclusion
Ultimately, hardware can’t be ignored; hardware represents 50% of the most valuable tech companies globally and nearly 75% across all sectors (6). Hardware is here to stay, and as technology improves at a rapid pace, there’s no better time to invest in the key to this innovation.
With less competitive deals, an equally attractive return profile, and an essential role in software innovation, connected hardware/software businesses represent a unique and undervalued investment opportunity.
Sources:
Photo by Alexandre Debiève on Unsplash