Hello everyone, good afternoon! It's that time of year again—the Double Eleven shopping festival. I didn't expect that even a single person like me could have an exclusive holiday. In previous years, I would spend a lot of money during this period.

Tonight, I will spend a lot again. I just checked my expenses this year and found that I have already spent so much on 🍑. I'm close to reaching the Black Diamond membership level, but I suddenly thought of a question:

When I check out during shopping, the page shows the final payable amount. I want to confirm whether this price is the correct amount after stacking discounts, red envelopes, and coupons, but I don't want to expose my privacy details—such as my membership level, coupon usage records, shopping allowances, etc.—so what should I do?

❙ Zero-Knowledge Proof provides an answer, but it also brings a more headache-inducing problem: it’s very computationally intensive.

▰ First, let's talk about the power of Zero-Knowledge Proofs: "One calculation, permanent verification."

Generating a Zero-Knowledge Proof can confirm the amount in just a few milliseconds, but the process of generating this proof involves re-running the calculations according to mathematical rules, which can be hundreds or even thousands of times more work than the original calculation.

▰ What's more troublesome is that the industry currently relies on GPUs to handle this computational load.

GPUs can indeed process many repetitive calculations simultaneously, barely meeting the demand, but they were originally designed for image processing, not the "finite field mathematical operations" needed for Zero-Knowledge Proofs.

Many functions inside GPUs, such as modules for rendering game lighting and shadows, are completely useless for proof generation. Instead, they consume resources and energy unnecessarily, resulting in a very low cost-performance ratio.

▰ At this point, @cysic_xyz's approach becomes particularly crucial. They are not researching on general-purpose hardware but focusing on "custom dedicated chips."

❙ So, how do we determine that the metrics for Zero-Knowledge computations differ from those of regular computers?

It depends on how many hash operations, mathematical constraints per second, and the time and energy consumption for proof generation. Dedicated chips can optimize these aspects by removing unnecessary functions and concentrating all performance on the core needs of Zero-Knowledge Proofs.

I believe this is not only a technological breakthrough but also a clear path for the entire industry. Zero-Knowledge Proofs are gradually moving from theoretical concepts to practical applications.

In the future, the demand for "verifiable computation" will only grow. I have to say, @cysic_xyz has made a very insightful choice in this field.
#cysic