As a long-time chip designer I've been pondering over the causes of this new wave of hardware innovation, coming seemingly from any corner in the world by almost always relatively young entrepreneurs. The more it unfolds in front of me, the more I feel a sense of déjà-vu of the first wave of successful lean startups, which were all driven by very young entrepreneurs and grew ever fast into dominance.
I've postulated here and there on why this is happening. Today this new TechCrunch article by Zavain Dar of Lux Capital caught my eye. It's a good complementary to my previous article on the potential of silicon power waiting to be unlocked by the agile new generation of hardware entrepreneurs.
To be specific, this graph that Mr. Dar drew was what caught my eye:
Among the six graphs, if you only wish to keep two that would be the middle one of the upper row and the last one of the lower row.
Pace of Innovation vs. Infrastructure
VCs and tech people who are old enough to live through the dot-com craze would remember how fast infrastructures were laid down during the frenzy. During the height it wasn't just the dog-food-peddling websites whose share prices had skyrocketed. Hardcore tech firms such as Cisco and Broadcom all saw their market caps balloon. These are companies that contributed to the building of internet infrastructures around the world, including backbones such as undersea inter-continent fiber links and last-mile technologies such as DSLs.
The burst of the bubble exposed the overcapacity in infrastructures. Along with sinking dot-com firms came crashing the unfortunate Juniper Networks, one of the largest fibre networking companies (and once Cisco's biggest rival) that saw its share price tumble from the high of $236 in late 2000 to a low of $4.66 within two years.
The infrastructures these companies laid down, though, never went away. They laid quietly under the sea for the time to come. And it didn't take long. As ADSLs and Cable Modems became the first true broadband last-mile, giving hundreds of millions of households fast and reliable access to the internet, the application finally exploded.
Youtube. Flickr. On-line gaming. Social networking.
These data-heavy applications rode the latent power of the dormant infrastructures and grew as fast as they could through various viral effects. Their growths were only determined by the startups' strategies and executions, rarely by the infrastructure limitation either technologically or economically. This is because the infrastructures had already been laid out during the bubble. Lean startup debates dwell mostly on execution, rarely on technology.
This is a demonstration of the 2nd graph in the upper row of Mr. Dar's article. A linear growth of infrastructure would lead to an exponential growth in innovation, as network effect comes into play and triggers positive feedbacks.
To some degree we seem to be seeing the same thing happening in hardware innovation. A decade of non-stop SoC development as well as relentless cost-down of key components such as processors, memories and connectivity modules has laid down the infrastructure for hardware innovation. Hardware entrepreneurs now focus on ideas (design, product-market fit) and execution (crowdfunding, operation and logistics).
Life expectancy of status quo vs. Time
As a natural result, the last graph by Mr. Dar would become a foregone conclusion. And judging by how things are moving in all areas, it seems to be true that disruptions are everywhere and becoming more and more often. A couple of examples:
- A year ago all articles were talking about how Google was going to conquer the whole world someday. Today most of Google-related articles are either predicting its doom or comparing it to Microsoft.
- Two years ago Samsung and its army of Android smartphones served both as a b-school case study of successful marketing strategy as well as a validation of Google's power. Today journalists skip Samsung entirely and directly tell fortune of a war between Apple and Xiaomi.
- Going mobile was an imperative two years ago. Everyone was going mobile. Today people hardly mentioned it — as it has become a status quo — and AI, machine learning and robots are all the rave.
- Bitcoin seems to have lost its relevancy before it even really became relevant. The talk has shifted to Blockchain and how this underlying technology might change how each identity — both mortal and virtual — interact on this planet
I remember when I first got out school and started my career, one of the best-selling business books was « Who Says Elephants Can't Dance?: Inside IBM's Historic Turnaround » by the legendary former CEO of IBM, Lou Gerstner. The time it took Mr. Gerstner to make the elephant dance, however, looks like eternality in today's world of disruption.
Today the elephants might not have the time to turn around and will just die, even if they have Lou Gerstner. What's scarier is how disruptors quickly morph into the targets to be disrupted.
No one is safe.
Implication on valuation
And if the life expectancy of status quo is only gonna get shorter, the implication on valuation is not small at all. After all, what would be the point to send a hot startup into IPO with huge accounting losses and argue that the time for harvest will come big in a couple of years, if the disruptions are already waiting down the pipeline?
Peter Thiel once said that when PayPal went IPO, he looked at their Discounted Cash Flow valuation model and noticed that a huge part of the current valuation comes from 10 years down the line. If the pace of disruption does increase relentlessly and the expected lifetime of status quo inevitably shrinks, would that long-term-value-building argument still work?
Or maybe that's why more and more hot tech firms are going through acquisitions instead of structural innovations from within?
These are all gonna be interesting questions to answer in the coming years, I feel.