Now, blockchain technology is taking a keen interest in ZKP protocols, which may add yet another layer of technological advancement to blockchain's evolution.
At the center of this particular development stands DZK, a company leveraging ZKP to create a blockchain-based proof platform.
DZK is two very bright men: Rami Akeela and Weikeng Chen, both Ph.D. holders with deep expertise in the blockchain space. Blockleaders recently spent an evening chatting to both, to learn more about their work, and why it matters.
This is our conversation:
Blockleaders (BL): What prompted you to start DZK?
Weikeng: We built this company because, in recent years, ZKP technology hasn't seen a lot of adoption, particularly in the blockchain area. A big issue with ZKP is that, although the technology is very powerful, it can consume a lot of computation resources. If there are not enough computation resources, then ZKP is basically not practical. Now, we feel we can use hardware and hardware-friendly algorithms to make ZKP practical for the real world. This is what DZK is all about.
Rami: Continuing with this idea, Weikeng is the cryptographer, and I'm the hardware engineer. We're applying the lessons that we learned by working in previous industries, including IoT, communication -and more recently, Machine Learning-, and we can actually see the same patterns. To make anything practical, we need to look at the underlying platforms, which include hardware and software. This is what motivates us. We saw what happened before, and we can now see what's going to happen in this industry (blockchain). We decided to start DZK to address these issues.
BL: Just out of curiosity, what does DZK stand for, if anything?
Weikeng: It's distributed zero-knowledge. Distributed in the sense of distributed computing, that it uses different hardware devices to produce proof.
BL: Let's talk a bit about your backgrounds.
Rami: I'm a hardware guy. I've basically been doing hardware design and engineering for most of my professional life.
I worked on logic level, architecture level, and digital level. I got my Ph.D. from Santa Clara University, and the topic was actually field-programmable gate array (FPGA) acceleration of multiple applications. One of them was ZKP algorithms. So I've worked in several companies doing hardware design, and about three years ago, I was introduced to blockchain and crypto by someone who was interested in my work. So I started exploring this field, back in 2019, working on FPGA acceleration applied to blockchain. Some basic algorithms, and soon after I started looking at ZKP. I created a start-up on my own to work on this, but then the pandemic happened, so everything sort of stopped.
But I'm still committed to this field, I feel there's a lot to be done, and that's why I'm currently working on blockchain.
Weikeng: I recently graduated from UC Berkeley with a PhD in computer science. During my time in graduate school, I focused on cryptography, mostly building practical systems for zero-knowledge proof and secure multi-party computation. To make these systems practical, I need cryptography to run fast, so I have always been interested in hardware acceleration.
My background is mostly using cryptography to build this system, and now is the time for me to look at the hardware side to build this working platform with real-world usability.
BL: Just as a personal curiosity, how did you two meet?
Weikeng: We actually met in Palo Alto, California. I was introduced by someone in the blockchain industry that Rami is doing some very cool stuff about zero-knowledge proofs.
Recommended for You
BL: Where do you see the future of blockchain?
Weikeng: My feeling about blockchain is that the technology will likely go in different directions. Firstly, privacy. People are starting to do a lot of financial transactions and playing with NFTs on blockchain, and they are looking for privacy.
Another one is scalability, because there are a lot of applications running on blockchain, the decentralization nature of the system limits its throughput. This is what we want to solve.
I would add that interoperability is another important topic today—different blockchain ecosystems interacting with each other as if they were the same blockchain.
So I see blockchain developing in these three strands. And I think that all three aspects confluence in ZK: privacy, scalability, and interoperability. ZK enables proving things without disclosing secret information, that's privacy. It allows blockchain to verify whether something has been processed correctly, without needing all the validators to check it all, that's for scalability. And lately, we have started to see the needs for one blockchain to interact with many other chains. ZK can also be applied here, to this interoperability.
Rami: I think large corporations are beginning to realize that ZKP is something they can incorporate into their existing platforms and services, because of the unique position of ZKP and blockchain technology. And to capture this interest, and the wave of applications that are coming, one needs to think about how to address issues that will definitely be surfacing soon.
I mean, blockchain is still in its infancy, really, and people got caught in the excitement of developing applications without answering some basic questions first. How will the app play out in terms of processing speed or cost, for example. I think the industry will do well once those foundational questions are addressed. We firmly believe that blockchain and ZKP have a very, very interesting future. We just need to take care of the basics before we move forward. Otherwise, we'll get stuck at some point. If we try scaling too fast we'll hit the wall.
Weikeng: That's a very good point. Actually, for all these directions, privacy, scalability, and interoperability, the problem has always been about computation. Some people might want to use ZKP to generate proof, but the proof generation is way too slow. So the actual hardware infrastructure for the blockchain is a crucial consideration to use ZKP efficiently and realistically.
BL: DZK is collaborating with AMD-Xilinx to architect the field-programmable gate array (FPGA) division in the upcoming ZPrize competition. The project seeks to accelerate the Number-theoretic Transform (NTT) operations for zero-knowledge proofs. I wanted to know what this collaboration meant for DZK, both personally and professionally.
Weikeng: I think the blockchain community is starting to realize that they have a strong need to have a robust hardware infrastructure to run ZKP efficiently. The ZPrize is originally focused on CPU and GPU, so we feel it is a good time to promote awareness of FPGA. We're doing this as the next stage of blockchain acceleration. Basically, we want to use this opportunity to push the industry forward.
Rami: AMD is very passionate about this concept too. They believe that ZKP will be the 'killer app' for FPGAs, for several reasons. FPGA will be vital in the processing of ZKP-based apps. The work that we're doing with ZPrize, we're actually the architects of the FPGA division, and, as Weikeng said, we're taking this opportunity to promote awareness of FPGAs to get people to start looking into this.
We got a lot of attention and support from several top ZKP companies and investment firms, including Aleo, Polygon, and Jump Crypto. The prize pool for the NTT FPGA division is $725k, and the prize pool for MSM GPU/FPGA division is $910k. This is the biggest prize for ZKP research and development to date.
So this collaboration is an indication that the industry is a) acknowledging the need to look into this technology, because they believe that this is the next generation of blockchain technology. And b) this is the perfect environment for innovation. If we onboard not only university students, but also companies and developers, we have a great incentive. It's the perfect environment, and it's good for us as a company, because it validates that the work we're doing is going to be the future and the answer to all the problems we currently have in the ZKP and blockchain communities.
Weikeng: Just to add that our partnership in DZK, that is, a cryptographer and a hardware engineer, has long been needed in the industry, and the prize money in ZPrize also serves to create this sort of novel partnerships between individuals with expertise in seemingly disparate fields, but can come together to foster innovation.
BL: As the last question, if you were to define DZK's mission in one sentence, what would that be?
Weikeng: Make the impossible possible.
Rami: Well, I can't top that, really, so we'll go with Weikeng's statement!
BL: Just as we were wrapping up, Rami wanted to make an official announcement. You're hearing this information here first, at Blockleaders.
Rami: DZK has developed a set of end-to-end, FPGA-accelerated privacy algorithms. We have successfully created systems capable of efficiently generating proof built on FPGA servers. This is the first, and so far only, end-to-end FPGA accelerator system for ZKP.
BL: Interesting revelation, and an interesting future for this enterprise.