ZKRushTeam Interview | A Deep Dive into Aleo Testnet 3

We conducted our first live team interview on Twitter Space last week to address some of the most commonly asked questions about Aleo Testnet 3. Many of them are concerned with our mining pool upgrade plans, GPU performance, algorithm optimization, and thoughts on the Aleo project. The following is a comprehensive recap of the interview.

Q1: Why do you choose Aleo in the first place? What’s the biggest strength of ZKRush on Aleo mining?

A1: In reality, we’ve been interested in Aleo for a long time, dating back to the first half of 2021, since the blockchain researchers in our team performed some forward-thinking research, and we’ve already made some predictions regarding the Zero-Knowledge track. Aleo is the first novel public chain to put effort to overcome privacy issues while ensuring programmability using zero-knowledge proof. It offers a full stack solution, from the trusted execution environment, zkCloud, to the Zero-Knowledge programming language, Leo, to the SDK and IDE; if all of these components work as planned, the project’s potential is limitless, really attaining the privacy of Zcash and the scalability of Ethereum. With over $200 million in investment led by a16z, the investors perceive the underlying worth of the initiative. Furthermore, we place a high value on the Zero-Knowledge prover market because the process of generating Zero-Knowledge proofs is complex and requires a large number of mathematical calculations, which is precisely our strength, thanks to our extensive mining experience and top algorithm engineers. After the entire switch from PoW to PoS last year, a large number of Ethereum mining rigs was left idle, therefore we believe this is our moment, and we are moving in the correct route. We accounted for more than 24 percent of the network’s computing power in Testnet2, and in Testnet3, we launched ZKRush pool, which can easily enroll Ethereum miners with nearly minimal CPU needs. That, I believe, is our most significant edge.

Q2: Does Aleo mining has a high threshold for hardware?

A1: The hardware requirements for mining on ZKRush are actually very low; in other words, you may mine as long as you have a reasonably new Nvidia GPU. Because our approach is a pure GPU solution, you don’t need a particularly strong CPU; even a Pentium G630 can run it. Our GPU compatibility is also quite high, with support ranging from the newest Nvidia 40 series to the 10 series. Our compatibility will improve as the Aleo technology advances.

Q3: Do you recommend Solo or joining a mining pool to participate in Aleo mining? What are the advantages?

A1: Solo and mining pools both have advantages and disadvantages. In the absence of the two service rates, the advantage of solo mode is that the account, the private key, and the mining address are kept in their own hands, whereas the account is managed and controlled by a third party when joining a mining pool. Furthermore, because there is presently no settlement method, the payout will be dispersed by the mining pool, which may pose some risk. However, similarly, solo has downsides. In the background of this incredible spike in computing power, if you are a little or medium-sized miner, your expected time for mining out a block will be rather lengthy, perhaps a week or even a month. As a result, your revenue distribution will be quite unequal; also, the extension of the block generation time has an effect on personal decision-making. However, if you join a mining pool, your revenue distribution will be reasonably even, and you will be able to monitor your revenue change each day, which is beneficial for personal decisions. So, I still recommend joining the mining pool for small and medium-sized miners.

Q4: Why did you focus on GPU ONLY algorithms from the beginning, instead of the CPU+GPU approach in the market?

A4: The reason we turned to a GPU-only approach in the first place is that several repositories of GPU implementations are already available for FFT and MSM to refer to. Furthermore, as compared to the combination technique, the GPU solution is more cost-effective during the rest of the proving process. Given the changes in the Ethereum mining market, the GPU-only approach is more appealing to the bulk of ETH miners and can be implemented fairly smoothly; even for small miners, it remains a more cost-effective alternative. We don’t know why there are fewer GPU-only algorithms on the market, however, developing GPU algorithms isn’t difficult for our team.

Q5: ZKRush has always focused on business-level miner access, so why did you choose to launch a mining pool now?

A5: Miners with high computing power, or major business-level clients, are and will continue to be our primary service focus. As I previously stated, following the conclusion of the Merge last year, a massive quantity of power was left idle, which does not rule out the possibility that many miners have already left the market. The traditional PoW mining projects led by Bitcoin and Ethereum may have exited the historical stage, but in the future, we still have new opportunities in this industry, which is the Zero Knowledge prover market. Now the ZK Rollups on Layer2 and zkEVM have already gained a high degree of popularity, although there is still a long time away from the practical stage, but it is never too late to start on that track now. Aleo may have been the first to take aim at the prover market. We conducted the argument analysis in conjunction with the market demand. First and foremost, high-volume clients prefer to go solo; their computing power is huge enough to compete in the network producing blocks, and the private keys are kept in their own hands, giving them the initiative to assure security. As the project grows in popularity, the network mining difficulty will increase, and the block production of small miners will become highly unstable, to the point where they will be unable to mine a single block in a few months. By then, the solo mode has no market; the only option is to join a mining pool to aggregate the computing resources and earn a consistent portion of daily revenue. This, I believe, is what small and medium-sized miners are pleased to see. The introduction of the mining pool responds to market demand while also improving testing of our algorithms’ compatibility with varied hardware, which is advantageous to our overall testing and adaption updates as well as our product itself. Furthermore, running a mining pool has a very excellent publicity impact; as long as our mining pool functions well, the spread of ZKRush will be extremely rapid.

Q6: A lot of firms claim they have Aleo-customized mining rig, what do you think of this?

A6: Nowadays, the market is quite concerned about the mining rig problem; there are also many so-called customized Aleo mining rigs on the market, and I believe we still need to examine three aspects:

1. A mix of hardware and software. We built really powerful hardware, but without the right software to effectively utilize the computing capabilities, many of them will be idle, so simply looking at the mining hardware is not a smart idea.

2. Cost-benefit analysis. When it comes to mining rigs, the input and output ratio is always important. We are now more concerned with the increased performance of a single GPU, and there are numerous methods to make this a reality, such as how much CPU input. However, the cost ratio is frequently the deciding element.

3. The timing. The current focus of customization and optimization is on GPU; the next step might be FPGA or ASIC. All of this is predicated on Aleo having a healthy and long-lasting ecosystem.

Q7: How competitive is your algorithm in the current market?

A7: Our team’s objective from the start has been to retain our top-tier status, and we have so far succeeded. Simultaneously, we have been monitoring market movements and continually refining our algorithm. For example, the enrollment of ZPrize a couple of months ago inspired us a lot, and we then made a big upgrade to the algorithm. We wish we could preserve our competitive advantage at all times.