In these eight years of Solana, Anatoly shares the behind-the-scenes story

Core Viewpoint

The Origins of Solana: From Side Project to Full-Time

Solana originated from a moment of "right time, right place, right people." At that time, a friend and I were working on a startup project, or more accurately, a side project, where we were doing AI-related things, such as deep learning servers, and using these GPUs to mine cryptocurrencies to pay for the GPUs. But a question arose in my mind: why would people spend money on our AI-related products? After a couple of cups of coffee and a beer, my partner and I talked about mining, PoW, Satoshi's consensus, and algorithms, and why electricity is so important in this process.

Most of my career has been spent as an engineer at Qualcomm. Most people should know that Qualcomm is deeply involved in wireless protocols, radio technologies, and mobile phones. Your phone likely uses Qualcomm products, and it may even use products I helped develop.

That day, I stayed up until four in the morning, and suddenly had a flash of inspiration: I thought of encoding the passage of time into a data structure. I recalled the protocol originally used in cellular networks, called Time Division Multiple Access (TDMA). This concept first appeared in the 1960s and 70s and is very simple: divide time into segments and use different time slots to transmit data, which prevents interference and allows more information to pass through. I thought of this because Bitcoin and the PoW mechanism face similar issues.

If there are two block producers, and two miners generate blocks simultaneously, a fork occurs, and the network becomes chaotic, making it impossible to transmit information properly. You have to discard one of the blocks. So, if you can make the two block producers take turns producing blocks, you can avoid conflicts and maximize the bandwidth utilization of the protocol. I roughly calculated that its throughput was 1000 to 10000 times higher than that of Ethereum or Bitcoin at the time.

The idea was born, and maybe I should start a company. I was genuinely interested in smart contract platforms because they provide developers with a brand new application development environment, and these applications are different from anything you build elsewhere, so you can't directly build smart contracts on regular AWS servers. You need the verifiability, cryptographic guarantees, and so on provided by blockchain, which makes it possible to write code that can handle funds.

At that time, many people believed that databases like those on Wall Street controlled the funds, monitored by humans, and many products were just optimizing the work of these people. Smart contracts are completely different; the software itself is responsible for holding funds and is the sole authoritative source of the flow of funds, so in a way, smart contracts disrupt the entire data model.

The Bold Pursuit of Entrepreneurship

When I decided to start a business, I needed to persuade many people, and my wife was the first person I needed to convince. She is an engineer and knows me well; I have always had side projects, always putting some ideas into practice in my spare time. We already had a child, and she said, "Okay, this may work, but you can't be a worker, a father, and a part-time entrepreneur at the same time. You have to choose one: either go all in or give up."

It was this statement that prompted my decision to start a business. I remember she was in Colombia at the time, Facebook was expanding, and she was working at a startup that was a competitor to Facebook, which was still in its early stages. What she learned there was that the market goes through about six months of a boom period, where everyone knows that a product being developed will capture 80% of the market share, and it will have certain explosive characteristics. If you miss that window, you will never catch up. So at the end of 2017, I felt it was the best window to build an L1 blockchain with specific attributes that could scale to cover the globe and truly handle the entire global financial system.

For me, the biggest motivation to create Solana was: first, you have to go all in, and second, you don't want to miss out when the market is booming. I think anyone reading this who is still hesitating about diving into AI or other fields should wait another six months or a year; you will really miss the opportunity. Act now, and if you have already started, that's even better.

Solana: Pursuing Transaction Efficiency Unlike BTC and ETH

Solana is a high-performance blockchain, and the key use case we have always pursued is transactions. If you think of Bitcoin as a means of storing value/digital gold, then building a means of storing value is not an engineering challenge. In fact, ensuring settlement and global availability does require some engineering techniques. Satoshi's PoW algorithm and the Bitcoin white paper do an excellent job in this regard. However, you cannot develop a Bitcoin Plus version; you cannot compete with Bitcoin in this market by adding features or increasing throughput. Ethereum's goal is to use settlement as an application scenario, with the idea that after execution and settlement are completed at the final checkpoint, you can use the Ethereum ledger as a reliable source of truth.

I never thought about competing in the settlement phase; perhaps there is still some room for technical improvements, such as adding an execution layer, but I am more interested in the execution itself. That is, building a global blockchain capable of handling transactions, payments, and everything users need in their daily lives, all of which can be done within one system.

What makes Solana unique may be its vision: no need for independent blockchains or hierarchical structures; you can integrate all functions into one massive state machine and coordinate all operations at the fastest speed. Here’s a statistic: the transaction volume completed by Solana in its first month was equivalent to the total transaction volume of Ethereum throughout its lifecycle at that time.

Challenges of Entrepreneurship: Funding and Hiring

There are many challenges in the early stages of entrepreneurship. For any founder, making progress in the first important approval stage may be the biggest obstacle, and the vast majority of companies fail at this stage. I remember having thousands of meetings; around the end of 2017, I listed all the venture capital firms in Silicon Valley that might invest in cryptocurrencies. Fortunately, I was in Silicon Valley at the time, which I think is why it remains a startup hub: you can meet thousands of people in a short time and try to pitch your startup idea.

For founders, being able to effectively pitch the product vision and concept is key; otherwise, you will never be able to hire people, sell products, or guide users, whether you are doing B2B or B2C.

Pitching Solana was a new experience for me and a process of learning and continuous improvement. That’s why I believe you can build a massive list in Silicon Valley, forcing yourself to repeat your efforts thousands of times to ensure you eventually reach the most valuable investors. The more familiar you are with the process, the better your pitch becomes.

For founders, you are trying to convey information in the most concise way possible. In a brief 10-minute conversation, you must figure out how much the other person already knows about cryptocurrencies because you don’t want to repeat what they already know. You also need to explain the specific problems the product is solving and its impact in the shortest time possible, showing them what changes the world will undergo based on the concept of cryptocurrency.

My strategy at the time was (I don’t know if this strategy applies to all founders), first pitch the company, then pitch that partner. Even if the company ultimately passes, I can persuade the partner to make a commitment, and they are more likely to connect me with other venture capital firms they know that invest in this field. Ultimately, this allowed me to attend thousands of meetings and find those focused on the crypto space who were more willing to take risks in the early stages, as the investors were also employees of the company, investing for the company and also making personal investments.

In fact, we had already completed a round of funding and were almost done. It was the first quarter of 2018, and there was no standard, secure investment template in cryptocurrency that could be quickly provided to investors. We spent six weeks having lawyers draft the relevant documents. But during that time, Ethereum started to drop, about 10%, and many funds went bankrupt as a result, which was the first challenge we encountered. Even so, many people were willing to participate; they were not entirely crypto funds and did not invest 100% in cryptocurrencies; their balance sheets held more dollars but viewed this investment as an opportunity. In the end, we completed this round of funding, but the situation was quite unstable at the time.

I was sitting in the 500 Startups (now renamed 500 Global) office with another co-founder, Raj (because one of the investors was from 500 Startups). He said, "I feel like I have to work hard, I have to fight." At that time, I thought that once the product had an investment commitment, it would likely snowball and eventually turn into actual checks, but my advice was to keep raising funds until there was actually money in your bank account.

I think the second challenge was hiring. However, I was lucky; many former colleagues from Qualcomm were eager to do something new, and these people had over ten years of experience in low-level operating systems or protocols. For example, one of the people involved in the development of the Solana protocol had participated in the formulation of the LTE specifications. These individuals had a very deep understanding of networks, operating systems, GPUs, CPUs, and underlying chips, and could understand what I was saying to them: "You are going to change jobs anyway, so consider building Solana as a vacation."

I hired some experts from various fields whom I knew very well, and everyone quickly got into the groove, starting to build what I believe was the most advanced network at the time. It turned out that at launch, Solana was already several streets ahead of all competitors.

From Founder's Alignment to Solana Achieving PMF

Speaking of work partners, the best way to describe my relationship with Raj is that it feels like being in a romantic relationship, requiring full commitment. Raj was introduced to me by a mutual friend, and at that time, I had no impression of him; he seemed like an ordinary person. The mutual friend specifically said, "You are a great engineer, but you have no other experience. You need someone who can complement you. Raj has started companies before and has done well, but he has no engineering experience at all; you two are a good match." We got along very well, and my wife basically referred to us as a "work marriage."

Our decision-making process was indeed exhausting, but in that high-pressure, fast-paced environment, we would repeatedly debate certain viewpoints until we eliminated all obviously bad options, leaving only what I call the Pareto efficient option set (meaning the discussion reached a point where there was no room for further improvement). We could choose A, B, or C, and all trade-offs seemed similar; we had almost discussed all possible directions, and at this point, it was almost a matter of luck.

This was tiring and required strong endurance. It also required mutual trust, believing in each other's judgment. I believe that CEOs and the initial employees or co-founders need this character; they can argue fiercely based on mutual trust but still feel that everyone respects each other. This is difficult; I enjoy arguing, and I don’t mind losing. Many of the CEO's shortcomings or personality traits ultimately affect the company culture, and in the early stages of the company, any factor can trigger a debate.

Work hard to build the product and complete development as quickly as possible, but you cannot anticipate all possible failures. Should you assume you will succeed and invest funds to develop some auxiliary features to consolidate success and better launch the product? Or should you focus on developing the product well, proving you can do it, and then work on other enhancements? In the early stages, especially when developing complex products, you must make many such decisions.

For example, entrepreneurial books like Peter Thiel's Zero to One contain many excellent pieces of advice, and the best advice you can get is to build a Minimum Viable Product (MVP), which is to create the smallest product that can validate your idea, but this is actually very difficult to define. So you must find your niche market. We spent some time doing this, and it was almost forced, probably in the second year of our development cycle.

At that time, we had about 12 months of funding left (a total of 24 months of funding), and the product still could not operate normally. We had to cut all other features except for the existing ones, release the product as soon as possible, and minimize the changes needed. This allowed us to seize the market opportunity and launch a product that was completely different from all other products on the market.

To some extent, in the first year of developing Solana, I wanted to take on as much product risk as possible and aimed to create a top-notch product. This was indeed part of our vision; by the end of that year, we had developed a series of features and taken on about eight technical risks. If you only take the risk of trying one technology, the probability of success is 50%. But if you try eight technologies, the probability of all eight succeeding is only 1/256. So, the chance of failure is high, and various problems arise, and then you have to find ways to fix them and make adjustments repeatedly to push it to market.

But it was precisely because of these decisions that we took on these risks early on, which gave us a series of differentiated features that were effective to varying degrees. They were not perfect, but we did expand capacity and reduce latency, and the development experience based on Solana was completely different from any other platform.

At that time, Ethereum used a PoW mechanism, and the block generation time was about 12 seconds, but you had to wait for at least two blocks to confirm the finality of a transaction. So, users had to wait 30 seconds to confirm a transaction, which was definitely a poor user experience, and the processing capacity of 7 or 11 transactions per second was too low for any scale of application.

We achieved final confirmation of thousands of transactions in just 400 milliseconds, and including all round-trip times on the server side, it was only one to two seconds. So users or developers who saw Solana's performance were amazed because Solana was so different, even though the product itself was still quite imperfect. But it could run, although it would crash after about an hour.

Then came the timing of stabilizing its market launch, which was also the most stressful thing. You need to cut some things, such as supporting EVM, or supporting a certain programming language, or needing a high-level browser, or launching your own wallet stack, etc. Strip these away and push the most basic version to market as quickly as possible. But I think defining a Minimum Viable Product (MVP) that can achieve Product-Market Fit (PMF), which is ultra-high capacity, low latency, and removes all other features, is very difficult because you have no idea how much you should sacrifice, nor do you know what developers really care about. We were lucky because we basically made most of the right choices based on our previous experience in developing operating systems and developer platforms, which greatly helped the final outcome.

But I think the hardest part is the product's sustainability; cryptocurrencies can bring many deceptive viral effects. Your token price may skyrocket, but in reality, there are no users, and you are disconnected from them. At that time, we didn't have much of a user base, but the SOL token price was rising, and we needed to take advantage of this opportunity to accumulate as many actual user cases as possible. If we missed this opportunity, it would be hard to recover.

We were lucky at the first hackathon; many people submitted works, but the applications they created were all kinds of messy things. By the second hackathon, I felt, "Wow, we seem to have found direction," because the works from the first hackathon, after three months of continuous improvement, presented a very complete product with full features that truly aligned with our overall vision for finance, trading, and DeFi.

During the second hackathon, while judging the entries, I found that there were huge differences in quality, usability, business models, and actual entrepreneurial capabilities (such as whether they could raise funds and survive). Seeing these companies secure funding during the hackathon made me feel that we now had Product-Market Fit, and it was part of our core business with a path to profitability.

So I think this was the biggest change since Solana's launch. I mean, considering all factors, to reach this stage within a year of product release is incredibly fortunate. Most companies take several years to explore before finding the best product-market fit; I think building a real company takes ten years.

From High Spirits to Sudden Setbacks: Surviving the Crisis of Solana

Then came one of the worst lows we experienced in the industry—the FTX incident. As everyone knows, FTX was one of our largest investors and partners. At that time, we were hosting the third Breakpoint conference, which was massive, attracting about 1600 developers. Our tickets were sold out, and then, on the return flight, FTX collapsed.

That was the situation; on the plane, I felt everything was going smoothly when FTX collapsed, and the crypto market plummeted, leading to a market downturn that could potentially destroy the entire ecosystem. Solana was founded at the beginning of the 2018 bear market when Ethereum was dropping 10% weekly. So we were very cautious; we never overhired, and the company had sufficient funds to develop and improve the product.

I was very scared. Many Solana ecosystem projects that raised funds on FTX actually left their funds on FTX because if their funding chain broke, it would be over; there would be no way to replenish funds, and all funds would be completely exhausted.

Fortunately, we conducted a large survey, and the results showed that 85% of the companies were doing well, while 15% were completely finished. Among these companies was a promising one, Armani's Backpack, which was developing a wallet. They had just completed a round of funding, around $10 million, and all their funds were stuck on FTX and could not be withdrawn. They were left with only a few million dollars and were planning to double their team size to build a product and complete the remaining seed round funding; at that time, they had only about six people. I thought most companies would go bankrupt, but they managed to survive.

Despite losing a significant portion of their funds, Backpack doubled down on their efforts and focused on the product. I believe they turned the situation around by launching the Mad Labs NFT series and establishing an exchange. I think Armani's anger towards FTX and the desire to build a better exchange contributed to this transformation. It was like the energy of a founder driven by anger; I felt that when they launched Mad Labs, they captured the attention of the NFT market and the entire industry, lasting for a full two weeks. It felt like a complete turning point, and you could see many companies doubling down and bouncing back.

It was like the return of a bull market. One of the biggest lessons I learned from this is that building a company during a bull market is actually very difficult, especially in the cryptocurrency space, because the signal distortion is severe. You don’t know who your core users are, nor do you know which features are truly important for your product and growth.

But during a market downturn, if you have 10 to 20 loyal users who frequently use your product, especially in the financial sector, if you understand what value your product brings to them and continuously optimize it, making it better every week, then during a bull market, you will see tremendous growth because, first, these users will become your biggest advocates, and second, your product will be highly optimized for specific use cases.

The product already has Product-Market Fit, and the financial industry is very cyclical; during a bull market, timing risk can generate enormous transaction volumes and revenues, so you need to have your product highly optimized and ready to scale, regardless of what your business model is.

So it was interesting to see the companies I interviewed after the FTX collapse; they were basically saying, "We will continue to optimize the product. We have enough funds. Let’s see what happens next year." All these companies succeeded and did exceptionally well.

The most severe issue was that the price of SOL dropped 97% from its peak, and most people thought SOL was dead.

Now I think it's great to have a co-founder who loves crises; some people are naturally better suited to operate in crises because your decision-making is constrained, and you must act quickly. What we did most was communicate with the founders of those companies that continued to grow, trying to help them grow, achieve Product-Market Fit, and clear obstacles as much as possible. But at that time, we couldn't provide financial support because funds were completely exhausted.

Regarding the FTX incident, I was very surprised by Sam; as you can see in interviews, he is that kind of super nerd, a quantitative analyst from MIT, a geek. They completely went bankrupt. But thinking about the potential losses caused by that chaos is truly unbelievable.

Will There Be More Chaos in the Future of Crypto with Improved Regulation?

I believe that the frequency of engineering-related hacks has significantly decreased, largely because innovations in smart contracts have reduced, and many uses of blockchain have already been explored. Smart contracts are beginning to commoditize; once deployed, you only need a certain number of CPMM automated market makers, and there’s no need to take on huge engineering risks to build another one.

Similar cases include Bonding Curves, lending protocols, etc.; you will see the attack surface for hackers has shrunk. Whenever there is a surge of innovation in the smart contract space, it is always accompanied by many risks. Besides that, I think there are now better tools, formal verification, better testing, and a deeper understanding of relevant attack vectors, and people are doing better in deploying these aspects. Risks have significantly decreased, and with the launch of new financial systems, their risks are lower, simply because they rely more on on-chain technology.

As for regulatory issues, this is a major problem faced by many exchanges or institutions. If regulation is too strict, it will take too long and cost too much. For example, obtaining a license might take two years, but you can't wait two years to gain market share. Projects will choose to move their business to overseas locations with less regulation and use banking infrastructures that are not as robust as those in the U.S. to build their businesses, resulting in various problems. I believe many failures in the last economic cycle fundamentally stemmed from this.

Now the U.S. has a stablecoin bill, and the SEC has also made significant changes, making it much easier to start a business here. But the U.S. is indeed lagging behind; Japan, France, and the UK have already introduced cryptocurrency-related laws that make cryptocurrency development easier. Japan may be the best place, as people from developed countries are getting into cryptocurrency. This is why projects like FTX Japan have been so successful; they are actually far ahead, but the market size in Japan is indeed smaller compared to the U.S.

Future Outlook: Solana's Vision to Dominate Financial Services

There are no engineering or technical reasons preventing the development of Solana. Solana's grand vision is that it can handle payments, transactions, contracts, IPOs, and all other businesses, all of which can be completed through a single execution engine on one chain. Accelerating the circulation of dollars, participating in the IPO market, and completing any transaction globally is a massive engineering effort that requires a lot of hard work and time to optimize and perfect, but from an engineering perspective, there is no reason to prevent its existence.

So this is what we truly want to build; if this system exists and has PMF, and everyone is using it, then you can actually reduce financial costs to the lowest level, equivalent to physical costs, which can be described as the ultimate state of software eating the world (i.e., the financial world).

The Solana ecosystem has many advantages because it is a more mature, faster-growing, and still-growing market. But I believe that achieving this vision will be highly competitive. I am not sure if there will be a blockchain as large as Google that can handle 99% of important transactions. The main reasons are twofold: first, countries with unique regulatory systems and firewalls may have their own blockchains; second, everyone wants a piece of the pie.

Even Google has launched its own chain. What about fintech companies and related businesses in the future, such as guiding retail investors to which platform, etc.? How will these integrations happen? It is still uncertain, but I believe Solana is that platform, so we will wait and see.

In this direction, what I truly want to see in the future is that companies in the U.S. and Silicon Valley that want to go public can do so through a simple method I call "Linux IPO from scratch," at a faster speed and lower cost. Founders like me, if they want to do this, can use on-chain immutable smart contracts that can be written into the S1 documents submitted to the SEC, stating that you are using this contract to go public on this open commercial blockchain, which has auction properties. I can directly list my equity on-chain, which will become the true source of the equity structure table and allow the public to access this information at any stage of the company's establishment without paying any fees to any investment banks, incurring no indirect costs, and all incentives and any fees you usually pay to banks can be used to incentivize AMMs to provide liquidity.

This would be my ideal way of operation because once this happens, it will greatly change the way companies obtain capital and how the public accesses early-stage companies.

I believe one of the most important components of the American dream is the free market. You know, I came to the U.S. from the Soviet Union in 1982, when the internet was just emerging, and companies like Microsoft and Amazon were growing rapidly. They were like building the future, and now these companies have become trillion-dollar giants. I think in the 90s, people could buy Amazon stock, which was undoubtedly a huge gift from America or a significant value proposition of America. And now the number of publicly listed companies in the U.S. is probably the lowest since the 1970s, or the lowest in terms of IPO numbers. So if we can provide founders with tools to complete IPOs at the lowest cost, fastest speed, and least legal fees, I believe this will greatly change the entire industry landscape.

This is a part of a very cool sci-fi future, where everyone globally can access financial services at the lowest possible cost and at speeds comparable to the speed of light. I think this is one of the coolest projects I can be involved in.

Epilogue: The Future of Crypto, A World of Stablecoins

I see cryptocurrencies being effectively adopted by Wall Street and some global institutions, with stablecoins being the main driver of this institutional adoption trend. The Genius Act passed by Congress creates a framework for issuing stablecoins and starting to achieve product-market fit, which is far better than any funding interface that traditional banks can provide. Even building all fintech products on traditional banks is far less effective than using stablecoins. So this will be a major driving force, with people expecting $10 trillion worth of stablecoins to be issued in the next 5 to 10 years. Currently, the issuance of stablecoins is about $250 billion (note: it has actually exceeded $300 billion), which is equivalent to a growth of several dozen times, and this liquidity will flow into all financial-related industries you can think of.

If you are a founder passionate about fintech or want to build a fintech company, I might suggest you build your business around stablecoins, either by integrating with existing stablecoins and managing various stablecoins or by creating your own stablecoin for specific purposes.

Translator's Reflection

From concept to action, Solana has experienced peaks, valleys, and rebirth over nearly eight years. The co-founder of Solana is one of the most passionate founders I have seen in the industry; they have advanced technology, understand operations and risk mitigation, have faced crises and emerged unscathed, and are full of confidence and execution power for the future vision. They are true crypto builders. At this moment, the heart of a SOL guardian is gradually warming up again.

Original Link