Interview With Solana Blockchain Founder Anatoly Yakovenko

Solana has successfully launched a testnet, so here is an interview with Solana founder, highlighting the path that he went through to build the new technology. Anatoly Yakovenko, founder of the Solana blockchain, someone who works to solve technological challenges and came with a solution that he believes can make a big difference in the blockchain area.

Could you introduce yourself and say how and when you joined the Solana team?

So here is the long story. I spent most of my career at Qualcomm, working on mobile operating systems, usually low level kernel firmware and trying to make things faster. Around 2017 I was at Dropbox working on compression and this crazy blockchain boom happened. One day I had too much coffee and was up till 4 in the morning thinking about single threaded mining and had this realization that it’s possible to encode passage of time as data, and a lightbulb went off in my head.

Once you have a source of time, all the optimization techniques that are used in wireless protocols become available. That was the start of Proof of History.

I couldn’t find anything on the Internet at the time that mentioned this technique, later I discovered that this is called a Verifiable Delay Function, and they have been around in some form since the 90s, but that was enough for me to quit my job and start working on Solana. I was super lucky to have been close friends with Stephen our GPU expert, and Greg our CTO and compiler expert, and I was able to convince them that this crazy idea might work and join me to start Solana.

Can you recap the main points of the journey in building Solana?

We have been heads down focused on building the network. It was mid March 2019 when we had our first demonstration of a BFT system that could survive a single fault. By April we had ambitions plans of onboarding validators and having a live network by the end of summer. Of course things went a little slower.

When we first deployed on a global open internet we saw that our network stack was running with the assumption that 64kb UDP packets would arrive most of the time across the world. Of course that was wrong, and we had to rebuild the stack to deal with 1200 byte packets.

While it’s easy to make it work, it’s quite difficult to make it work fast, since the stack now needs to handle 50x more events per second. We spend the summer rapidly developing the network and finally getting a stable global release in the Fall. Making it fast took another 3 months :). Finally at the end of the year we ran a global fault tolerant replicated state machine that could peak over 50k with 400ms blocks.

What makes Solana stand out from the rest of the popular blockchains?

So we are the only scalability chain that is rejecting sharding. When I started working on mobile phones we had 2mb of RAM and 16 bit 200mhz arm processors. When I left Qualcomm we had 4GB 8 core 64 bit arm chips. Hardware scales at an exponential pace. All you need to scale is just not get in the way. The challenge has been designing consensus and network protocols that can do that. We have been able to adopt a lot of the optimization techniques used by wireless protocols to build solana. The result is a network with 400ms block times that can handle 50kt/ps.

With Ethereum 2.0 launching perhaps later this year, what do you think are the possible implications for Proof of Sake networks and Solana in particular?

Eth 2.0 has been a fantastic source of research and development in the space. I don’t think we would have been able to build as good of a product without all the hard research done by the ethereum community. I hope that with 2.0 we will see a more projects launching and gaining traction and discovering new sustainable business models.

So far Solana has been focusing more on the tech side and less on the community and marketing. Are you planning to change this, and how?

We are hosting meetups and events across the world and working with our validator partners to help us reach developers and grow our community worldwide. I would love to see more core contributors to the core protocol and developer tools.