Dragonfly partner: How did I miss the opportunity to invest in Solana's seed round?

Missed out on one of the most expensive investment memorandums in crypto history, which would have yielded 3250 times the return.

Author: @hosseeb

Compiled by: Deep Tide TechFlow

Shenzhen Tide Note: At the 5th anniversary of Solana's birth, Dragonfly Capital's partner @hosseeb released a tweet today, reviewing how he missed the opportunity to participate in Solana's seed round investment at a price of 0.04 US dollars in 2018, and missed out on returns of over a thousand times. At the same time, the original investment memorandum is attached in memory. In addition, we have excerpted the discussion between Solana's co-founder Toly and Hosseeb under this tweet.

The following is the original text details:

I turned down the opportunity to invest in @solana's seed round at 0.04 USD in early 2018.

Based on the current price, it is equivalent to missing out on a return of 3250 times.

Solana was one of the first projects I evaluated as a junior VC. At that time, I was still cute, naive, and confident, writing memos for every project I gave up on investing in.

Rereading this memo now is like a 'peak junior VC cringe'. We were all obsessed with finding the 'Ethereum killer', studying consensus protocols, and what technology would replace EVM / eWASM.

So, this is the completely unedited original memo - the worst investment MISS of my career.

Happy Birthday, Solana!🎂

Memo Content

After reading the white paper, my notes are as follows:

Their significant innovation is historically proven (PoH). Essentially, this is a verifiable time-delay function that uses continuous hash operations, similar to sequential proof of work. In other words, a timekeeper is selected who continuously iterates hash operations on a value and discloses all intermediate hash values. Since this process must be executed serially on a single core and cannot be parallelized, nodes should be able to predict the amount of time between consecutive hashes (presumably based on their understanding of hardware performance?).

The PoH node will also mix any current state (such as transactions to be submitted) into these hashes. This allows for the creation of an event history that can be reliably timestamped.

If the PoH node encounters problems or cannot guarantee online, they have proposed a solution to allow multiple PoH nodes to regularly mix states with each other.

A group of validator nodes will replay and verify the operations of PoH nodes (the verification process can achieve more efficient parallelization through the MapReduce architecture). These validators reach consensus using a PoS protocol similar to Casper. If a PoH node is found to have Byzantine problems or behave improperly, validator nodes can elect a new PoH node to replace it.

It looks like they will develop payment and smart contract functions.

They claim to achieve 710,000 TPS and have achieved 35,000 TPS on a single-node test network.

My thoughts:

Their numbers are completely ridiculous. 710,000 TPS is simply ridiculous; even Google's search volume per second is less than 100,000. This data is placed in the most prominent position on their website, which makes me very cautious.

Withdraw the previous praise for the well-written white paper. The high-level content is good, but the technical details are very lacking and vague. As a description of a consensus protocol, the rigor is disappointing.

The team is mainly composed of engineers from Qualcomm. The CEO and CTO are mainly engaged in the work of operating systems, embedded systems, GPU optimization, and compilers. Their background in distributed systems and cryptography is clearly not strong enough, which is evident in the paper. The handling of Byzantine fault tolerance issues is very poor. It reminds me of the white paper of Raiblocks/Nano (they are also low-level engineers).

And the content in the white paper makes me doubt:

[Solana Whitepaper Original Text, Section 5.12]

"PoH allows network validators to observe past events and their timing with some degree of certainty. When the PoH generator produces a message stream, all validators need to submit their signatures on the state within 500ms. This value can be further reduced based on network conditions. Since each validation is input into the stream, everyone in the network can verify if all validators have submitted their votes within the specified timeout period without directly observing the voting process."

This is not a consensus protocol. Assuming that limiting 500ms for consensus in message passing is quite problematic, and it does not make sense to achieve Byzantine fault tolerance. And how do they measure 500ms? Considering that they will estimate the passage of time based on the number of iterative hashes executed, how will other nodes in the system reach a consensus on the passage of 500ms? In addition, how will they address the deviation in clock speed over time caused by hardware improvements, hardware failures, or noise? The issue of time in distributed systems is very complex, and I don't think they realize how difficult it is.

Besides, who cares about time? Is this a big issue in the field of blockchain? Are people not satisfied with the granularity of block time at 15 seconds/1 second (like DFINITY)? I don't think this is a problem. The complexity and confusion they introduce in the protocol do not seem to bring much value.

They have a section specifically discussing the issue of misalignment between attacks and incentives. Their response to the attack is completely unconvincing, and equally lacking in rigor or detailed explanations.

They have a whole chapter discussing proof of replication, just like Filecoin. What's going on? Tell me about your consensus protocol and how transactions, accounts are implemented, what features your blockchain will have. I don't care about data storage proof.

There is another long section that starts to describe smart contracts, but it only says they will use LLVM as the backend to support multiple platforms. But nothing else is mentioned.

A lot of content about GPU and parallelism. This exposes a strange sense of focus — if they need to implement a BFT consensus protocol and an available smart contract platform, they should not be obsessed with the parallel processing of their data packet format. I remember they were like this in the demos I watched — spending most of the time discussing how to use these nodes for optimization, and hardly any time describing their consensus protocol in practice.

Conclusion: I will absolutely not invest in this project

Interestingly, after 5 years, when Haseeb @hosseeb tweeted his blessings for Solana's success in the cryptocurrency world and joked about how he missed a great opportunity when he was young and inexperienced, Solana co-founder Toly @aeyakovenko replied to this tweet: "All your concerns back then were indeed valid. Essentially, this is a gamble - a bet on whether we can maintain the underlying advantages that other teams lack while solving these issues."

Haseeb then replied to Toly: 'I think that's the lesson. Your team's persistence in optimizing the underlying technology and unique attack angles is unmatched by other teams. The ability to maximize strengths and avoid weaknesses is the most important. At that time, I was completely unaware of this.'