Kandou buzzing with $15 million investment
30.07.2018
High-speed chip pioneer Kandou Bus has closed a series B round with financing from venture capitalists (and previous investors in the company) Bessemer Venture Partners and Walden Investment. The funds will be used to accelerate production of the company’s Chord™ signaling products.
 Amin Shokrollahi photographed by Martin Heimann
|
 |
Kandou’s breakthrough is a method of transmitting multiple bits over multiple correlated wires without reduced margin – delivering two to four times the bandwidth, with half the power consumption and excellent signal integrity. “Customer demand for Kandou’s technology continues to grow,” said founder Amin Shokrollahi. “This funding helps Kandou respond to increasing demand for Chord signaling architectures and solutions.”
Kandou’s success story started with Venture Kick back in 2011. Shokrollahi took the time to answer our questions about his entrepreneurial journey.
How did Kandou Bus originate (and what’s the meaning of the name)?
The idea behind Kandou’s technology originated in a casual conversation in my lab at EPF Lausanne about how to improve a ubiquitous method of data communication, in which one bit is transmitted across two correlated wires. This was back in April 2010. The name Kandou means “beehive” in my native tongue, Farsi. Just as a beehive is much more than the sum of its parts, in our method of data communication the total data throughput across a large set of correlated wires is much bigger than what one could achieve if data were transmitted on single wires. The collaboration of communication wires is what makes the technology work, in the same way that the collaboration of bees is what defines a beehive.
What was the impact of winning Venture Kick for your early growth?
It was huge and I cannot overstate it. To begin with, the regular cadence of the awards (starting small, ending big) forced us to introduce a corresponding cadence in the way we developed the company. There had to be a meaningful improvement in the state of the company at each stage. In addition, we were obliged to provide monthly updates, another cadence that helped us define realistic milestones and keep on track to achieve them. The actual monetary award was obviously also quite important, and equated to a good percentage of our annual burn rate. I dare say that the Venture Kick award was what made it possible for me to create enough value during the first year of the company to raise my seed round at a reasonable valuation. It prepared the ground for the future growth of the company.
Kandou’s mission is “to increase the energy efficiency of electronic devices and systems”. Your chips also increase processing speed – but do you see energy as a more pressing need?
Speed and energy are two sides of the same coin. You can spend less energy if you run slower, and will have to spend more if you run faster. You have to transmit more data per wire from one generation to the next, since the number of data wires is limited, but the amount of data to be processed grows roughly by a factor of two every 18 months or so. So, the question becomes: how do you send more data and not spend quite as much power? This is the problem we are solving. This said, power is king in devices such as mobile phones or tablets, and it is extremely important for data centers.
What innovation do you expect to drive the electronics industry forward in the next five to 10 years?
This is very hard to say. I don’t think it will be one innovation that will drive the electronics industry for the next decade, but many. The most pressing issue is that Moore’s Law has come to an end. This law describes something that can commonly be seen as “free lunch”: it states that the number of transistors per square millimeter of silicon will double every two years (roughly), while the production cost stays fixed (or even decreases). This law has been the driving force behind the many innovations in the electronics industry that have given us cellular phones, tablets, smart TVs, data centers, and practically everything that powers our digital lives. At a more technical level, this law has been responsible for massive integration of components onto a single chip (called a “system on chip”.)
The economy of cramming more transistors into a given area does not work anymore, though. Large systems on chip are so expensive that only a handful of companies are able to produce them. Innovation suffers as a result. One of our products, Glasswing, is designed to counter this downward spiral. Together with our partners, we are advocating disintegrating components – putting each component on a separate chip, connecting those chips with our Glasswing interface, and packaging the entire solution. The resulting development will be of much lower cost, and will lead to faster time-to-market for new products. In addition, it allows smaller companies to produce innovative solutions, and hence fuels innovation. If this type of solution takes off, it will drive the electronics industry for several decades.
You already have offices in Germany, the UK and Silicon Valley as well as Lausanne; you supply industry leader Marvell; you have won a raft of awards and been one of Switzerland’s Top 100 for six years in a row. In the whole history of Kandou, what achievement are you most proud of?
There are many things I am proud of, and it is difficult to put a linear order on them. First and foremost, though, I am proud of the team that is Kandou. Not only is the team incredibly talented, it is also extremely dedicated, and everyone works very hard to achieve our common goals. I am also very proud of our patent portfolio (~250 filed patents in ~110 different families). Our patent team has created the right strategy to harvest patents, together with our engineering team. In the last three years we have averaged around 40 patents per year, and this number is set to grow.
What comes next for Kandou?
Well, believe it or not, I am again in the fundraising mode! We have expanded our business model and will be selling standalone chips. Such an endeavor is quite expensive, and I will need to raise substantial capital to support it. On the product side, we will be delivering quite a few exciting new products to the market.
Kandou’s success story started with Venture Kick back in 2011. Shokrollahi took the time to answer our questions about his entrepreneurial journey.
How did Kandou Bus originate (and what’s the meaning of the name)?
The idea behind Kandou’s technology originated in a casual conversation in my lab at EPF Lausanne about how to improve a ubiquitous method of data communication, in which one bit is transmitted across two correlated wires. This was back in April 2010. The name Kandou means “beehive” in my native tongue, Farsi. Just as a beehive is much more than the sum of its parts, in our method of data communication the total data throughput across a large set of correlated wires is much bigger than what one could achieve if data were transmitted on single wires. The collaboration of communication wires is what makes the technology work, in the same way that the collaboration of bees is what defines a beehive.
What was the impact of winning Venture Kick for your early growth?
It was huge and I cannot overstate it. To begin with, the regular cadence of the awards (starting small, ending big) forced us to introduce a corresponding cadence in the way we developed the company. There had to be a meaningful improvement in the state of the company at each stage. In addition, we were obliged to provide monthly updates, another cadence that helped us define realistic milestones and keep on track to achieve them. The actual monetary award was obviously also quite important, and equated to a good percentage of our annual burn rate. I dare say that the Venture Kick award was what made it possible for me to create enough value during the first year of the company to raise my seed round at a reasonable valuation. It prepared the ground for the future growth of the company.
Kandou’s mission is “to increase the energy efficiency of electronic devices and systems”. Your chips also increase processing speed – but do you see energy as a more pressing need?
Speed and energy are two sides of the same coin. You can spend less energy if you run slower, and will have to spend more if you run faster. You have to transmit more data per wire from one generation to the next, since the number of data wires is limited, but the amount of data to be processed grows roughly by a factor of two every 18 months or so. So, the question becomes: how do you send more data and not spend quite as much power? This is the problem we are solving. This said, power is king in devices such as mobile phones or tablets, and it is extremely important for data centers.
What innovation do you expect to drive the electronics industry forward in the next five to 10 years?
This is very hard to say. I don’t think it will be one innovation that will drive the electronics industry for the next decade, but many. The most pressing issue is that Moore’s Law has come to an end. This law describes something that can commonly be seen as “free lunch”: it states that the number of transistors per square millimeter of silicon will double every two years (roughly), while the production cost stays fixed (or even decreases). This law has been the driving force behind the many innovations in the electronics industry that have given us cellular phones, tablets, smart TVs, data centers, and practically everything that powers our digital lives. At a more technical level, this law has been responsible for massive integration of components onto a single chip (called a “system on chip”.)
The economy of cramming more transistors into a given area does not work anymore, though. Large systems on chip are so expensive that only a handful of companies are able to produce them. Innovation suffers as a result. One of our products, Glasswing, is designed to counter this downward spiral. Together with our partners, we are advocating disintegrating components – putting each component on a separate chip, connecting those chips with our Glasswing interface, and packaging the entire solution. The resulting development will be of much lower cost, and will lead to faster time-to-market for new products. In addition, it allows smaller companies to produce innovative solutions, and hence fuels innovation. If this type of solution takes off, it will drive the electronics industry for several decades.
You already have offices in Germany, the UK and Silicon Valley as well as Lausanne; you supply industry leader Marvell; you have won a raft of awards and been one of Switzerland’s Top 100 for six years in a row. In the whole history of Kandou, what achievement are you most proud of?
There are many things I am proud of, and it is difficult to put a linear order on them. First and foremost, though, I am proud of the team that is Kandou. Not only is the team incredibly talented, it is also extremely dedicated, and everyone works very hard to achieve our common goals. I am also very proud of our patent portfolio (~250 filed patents in ~110 different families). Our patent team has created the right strategy to harvest patents, together with our engineering team. In the last three years we have averaged around 40 patents per year, and this number is set to grow.
What comes next for Kandou?
Well, believe it or not, I am again in the fundraising mode! We have expanded our business model and will be selling standalone chips. Such an endeavor is quite expensive, and I will need to raise substantial capital to support it. On the product side, we will be delivering quite a few exciting new products to the market.