You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is reading the result of the verification of the proof in Ethereum.
@@ -92,7 +93,7 @@ If the proof wasn't verified you should get this result:
92
93
Aligned works in:
93
94
- MacOS Arm64 (M1 or higher)
94
95
- Linux x86 with GLIBC_2.32 or superior (For example, Ubuntu 22.04 or higher)
95
-
If you don't meet these requirements, clone the repository, install rust, and then run:
96
+
If you don't meet these requirements, clone the repository, install rust, and then run:
96
97
97
98
```bash
98
99
make uninstall_aligned
@@ -170,13 +171,13 @@ If you want to run an operator, check our [Operator Guide](./docs/operator_guide
170
171
171
172
## Aligned Infrastructure Guide
172
173
173
-
If you are developing on Aligned, or want to run your own devnet, check our [setup Aligned guide](docs/guides/6_setup_aligned.md).
174
+
If you are developing on Aligned, or want to run your own devnet, check our [setup Aligned guide](docs/3_guides/6_setup_aligned.md).
174
175
175
176
176
177
## Submitting Proofs to Aligned
177
-
For submitting proofs generated by your own project to the network via CLI, see the documentation on [submitting proofs to Aligned](docs/guides/0_submitting_proofs.md).
178
+
For submitting proofs generated by your own project to the network via CLI, see the documentation on [submitting proofs to Aligned](docs/3_guides/0_submitting_proofs.md).
178
179
179
180
180
181
## Integrating Aligned into your Project
181
182
182
-
If you are developing applications using Aligned, we offer a [Rust-SDK](docs/guides/1_SDK.md) for submitting proofs directly to the network within your applications.
183
+
If you are developing applications using Aligned, we offer a [Rust-SDK](docs/3_guides/1_SDK_how_to.md) for submitting proofs directly to the network within your applications.
Copy file name to clipboardExpand all lines: docs/1_introduction/0_about_aligned.md
+45-8Lines changed: 45 additions & 8 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,28 +1,55 @@
1
1
## What is Aligned?
2
2
3
-
Aligned is a decentralized network that verifies zero-knowledge/validity proofs and posts the results to Ethereum. It is designed to provide high throughput, cheap proof verification with low latency.
3
+
Aligned is a decentralized network that verifies Zero-Knowledge/validity proofs and posts the results to Ethereum. It is designed to provide high throughput, cheap proof verification with low latency.
4
+
5
+
{% hint style="info" %}
6
+
If you are unfamiliar with ZK and why this is useful, see [Why ZK and Aligned?](./5_why_zk.md)
7
+
{% endhint %}
4
8
5
9
## Mission
6
10
7
11
Our mission is to accelerate Ethereum's roadmap and the adoption of verifiable computation by enabling fast and economical verification of ZK and validity proofs.
8
12
9
13
## What real value does Aligned bring to the table?
10
14
11
-
Blockchains are verifiable computers. We live in a chaotic world where there will be a huge demand for computation that does not need to be trusted. Current blockchain models are inefficient: each node must re-execute each transaction/operation, making the weakest devices the bottleneck. Besides, adding more hardware does not make the system faster (contrary to what happens in Web2), just more reliable. Alternative L1s to Ethereum make trade-offs, accepting bigger hardware or changing the consensus to be faster but with fewer guarantees. In other words, we reduce costs at the expense of lower security guarantees. Zero-knowledge/validity proofs (ZK) provide a new framework where we do not have to make such compromises. The basic premise is that a party can generate a very short proof of a computation, and the nodes can quickly verify that proof instead of re-executing the computation. The computation can be quite large, comprising many different transactions. This enables the delegation of execution off-chain with the same guarantees, leading to increased throughput and lower operational costs. This led to the rollup-centric roadmap of Ethereum, featuring multiple L2s, but at the expense of fragmented liquidity and complex user experience. Many of these problems could be solved by ZK bridges, but they remain expensive, and the go-to-market time is high.
15
+
Blockchains are verifiable computers.
16
+
We live in a chaotic world where there will be a huge demand for computation that needs to be trusted.
17
+
18
+
Current blockchain models are inefficient: each node must re-execute each transaction or operation, making the weakest and slowest devices the bottleneck. Unlike in Web2, where adding more hardware can increase performance, in these systems, additional hardware primarily enhances reliability rather than speed. Alternative L1s to Ethereum make trade-offs, accepting bigger hardware or changing the consensus to be faster but with fewer security guarantees. In other words, they reduce costs and increase speed at a great expense of lower security guarantees.
19
+
20
+
On the other hand, Zero-knowledge/validity proofs (ZK) provide a new framework where we do not have to make such compromises. The basic premise is that a party can generate a very short proof of a computation, and the nodes can quickly verify that proof instead of re-executing the computation. The computation can be quite large, comprising many different transactions. This enables the delegation of execution off-chain with the same guarantees, leading to increased throughput and lower operational costs. This led to the rollup-centric roadmap of Ethereum, featuring multiple L2s, but at the expense of fragmented liquidity and complex user experience. Many of these problems could be solved by ZK bridges, but they remain expensive, and the go-to-market time is high.
12
21
13
-
Still, nodes have to verify those proofs, which can be quite expensive since this is done on-chain: current proof systems can cost between 10 and several hundred dollars (which is strongly dependent on network congestion).
22
+
Anyhow, nodes still have to verify those proofs, which can be quite expensive since this is done on-chain: current proof systems can cost between 10 and several hundred dollars (which is strongly dependent on network congestion).
14
23
15
24
In addition, ZK lets us build other verifiable applications where users do not have to trust the party performing the computation, with impact in areas such as artificial intelligence (AI), the Internet of Things (IoT), and fighting misinformation. This is not clear to most people since proving technologies were not mature enough and verification costs were high.
16
25
17
26
## What limits the development of more complex applications on top of blockchains?
18
27
19
-
The main limitation for building complex applications on top of blockchains has been that the computation can run only a few milliseconds on chain and even then this can be very expensive. You can't have millions of daily active users using Ethereum or any blockchain at the same time.
28
+
The main limitation for building complex applications on top of blockchains has been that the computation can run only a few milliseconds on chain, and even then, this can be costly. You can't have millions of daily active users using Ethereum or any blockchain at the same time.
20
29
21
-
ZK solves this but due to slow and complex-to-use proving, and expensive verification, progress has been limited. In the case of proving, before the development of general-purpose zero-knowledge virtual machines (zkvms), users had to express their computation as arithmetic circuits, making the developer experience something like coding in assembler, error-prone, and complex. Moreover, proof systems depended on trusted setups, adding additional trust guarantees, the need to carry out special ceremonies to initialize parameters, and delaying go-to-market times. Besides, having high verification costs (on the order of 10's to 100's of dollars per proof) meant that only those projects with a lot of capital could afford to build such applications.
30
+
ZK solves this, but due to slow and complex-to-use proving and expensive verification, progress has been limited.
31
+
In the case of proving, before the development of general-purpose zero-knowledge virtual machines (zkVMs),
32
+
users had to express their computation as arithmetic circuits,
33
+
making the developer experience something like coding in assembler, error-prone, and complex.
34
+
Moreover, proof systems depended on trusted setups,
35
+
adding additional trust guarantees, the need to carry out special ceremonies to initialize parameters,
36
+
and delaying go-to-market times.
37
+
Besides,
38
+
having high verification costs (on the order of 10's to 100's of dollars per proof)
39
+
meant that only those projects with a huge capital could afford to build such applications.
22
40
23
41
## Why didn't anybody do it before?
24
42
25
-
To build Aligned, we needed several pieces in place. First, we needed EigenLayer, which allows building services and applications on top of Ethereum without competing for blockspace. In our case, we could bootstrap the economic security for a decentralized network of verifiers, avoiding the limitations of running proof verification on-chain. Second, proving technology had to improve. We currently have general-purpose zkvms (which means we can code in Rust and other high-level languages and prove it), proof systems are faster, and several improvements and developments are on the way. This makes writing applications easier (providing a higher demand for proof verification), and enables faster and simpler proof recursion (for proof compression).
43
+
To build Aligned, we needed several pieces in place.
44
+
First, we needed EigenLayer,
45
+
which allows building services and applications on top of Ethereum without competing for blockspace.
46
+
In our case, we could bootstrap the economic security for a decentralized network of verifiers,
47
+
avoiding the limitations of running proof verification on-chain.
48
+
Second, proving technology had to improve.
49
+
We currently have general-purpose zkVMs (which means we can code in Rust and other high-level languages and prove it),
50
+
proof systems are faster, and several improvements and developments are on the way.
51
+
This makes writing applications easier (providing a higher demand for proof verification),
52
+
and enables faster and simpler proof recursion (for proof compression).
26
53
27
54
## How much can Aligned reduce costs?
28
55
@@ -32,15 +59,25 @@ Aligned operates using two operation modes: fast and aggregation. The cost reduc
32
59
33
60
Aligned has created a new category in crypto: ZK verification layer. It is the only solution for now offering fast and aggregation modes while being linked to Ethereum via EigenLayer. The fast mode provides very high throughput (over two orders of magnitude more than Ethereum), low latency, and lower costs than the aggregation mode. The security is provided by restaking, and this mode is well suited for applications with tight limits on latency and costs. The aggregation mode works using the slower proof recursion strategy, also used in rollups to compress proofs. Another key feature is that Aligned is stateless, simplifying the process greatly.
34
61
35
-
Other solutions focus on building a separate L1 for proof verification (which sets them apart from Ethereum and requires bootstrapping the economic security, which can be lower and subject to volatility, as opposed to that provided by restaking) or focus on the aggregation of proofs from a few proof systems. This last approach adds latency, higher operational costs, and constrains the developer's choice of the proof system.
62
+
Other solutions focus on building a separate L1 for proof verification (which sets them apart from Ethereum and requires bootstrapping the economic security, which can be lowered and subject to volatility, as opposed to that provided by restaking) or focus on the aggregation of proofs from a few proof systems. This last approach adds latency, higher operational costs, and constrains the developer's choice of the proof system.
63
+
64
+
The following table contains costs estimates for Aligned, assuming a batch size of 20 proofs.
65
+
66
+
| Proof system | Ethereum | Aligned - Fast mode | Aligned - Aggregation |
67
+
| -------- | -------- | -------- | ---------- |
68
+
| Groth16 | 250,000 | 40,000 | TBD |
69
+
| STARKs | >1,000,000 | 40,000 | TBD |
70
+
| Kimchi-IPA | ?????? | 40,000 | TBD |
71
+
| Halo2-KZG | 350,000 | 40,000 | TBD |
72
+
| Binius. | ?????? | 40,000 | TBD |
36
73
37
74
## Why are we building Aligned?
38
75
39
76
In recent months, we have witnessed the development and enhancement of general proving virtual machines such as Risc0, Valida, Jolt, and SP1. These innovations allow users to write ordinary code in languages like Rust or C and generate proofs demonstrating the integrity of computations. This evolution is poised to transform application development, provided we have verification networks with high throughput and low cost. This is the core vision of Aligned and the reason we are building it: the future belongs to provable applications.
40
77
41
78
Currently, proof verification in Ethereum is expensive and throughput is limited to around 10 proofs per second. The cost depends on the proof system used, and the availability of precompiles. Groth16 costs around 250,000 gas, STARKs, over 1,000,000, and other proof systems are too expensive to be used in Ethereum.
42
79
43
-
Proof technology has been evolving over the last decade, with new arguments, fields, commitments and other tools appearing every day. It is hard to try new ideas if verification costs are high and there is a considerable go-to-market time, as a consequence of development time of new, gas-optimized smart contracts, or the inclusion of new precompiles to make them affordable.
80
+
Proof technology has been evolving over the last decade, with new arguments, fields, commitments and other tools appearing every day. It is hard to try new ideas if verification costs are high, and there is a considerable go-to-market time, as a consequence of development time of new, gas-optimized smart contracts, or the inclusion of new precompiles to make them affordable.
44
81
45
82
Aligned provides an alternative to reduce costs and increase throughput significantly. This is achieved by two different modes: **fast mode** and **aggregation mode**.
This program leverages Aligned's ZK verification infrastructure to run a small quiz. The proof allows any party to check that the quiz was answered right or wrong. If answered correctly, the user receives an NFT.
6
+
7
+
The process is as follows:
8
+
9
+
1. The user runs ZKQuiz and answers the questions.
10
+
2. ZKQuiz generates a ZK Proof of correct answers.
11
+
3. The proof is posted on Aligned.
12
+
4. Upon verification, ZKQuiz mints an NFT via a Smart Contract.
13
+
14
+
The NFT is only granted if the user's answers correctly.
15
+
Incorrect answers or tampering with the ZKQuiz code will result in proof generation failure or mismatched checksums,
16
+
preventing NFT minting.
17
+
18
+
Next, we will see how to execute ZKQuiz to get your own ZKQuiz NFT!
Copy file name to clipboardExpand all lines: docs/1_introduction/3_use_cases.md
+3-3Lines changed: 3 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -8,8 +8,8 @@
8
8
- Verifiable Machine Learning (ML): with general-purpose zkvms we can prove code written in Rust, solving part of the problem of using ML. However, most zkVMs use STARK-based proof systems, which leads to high on-chain costs or expensive wrapping. With Aligned, you can directly verify your proof from the zkVM for much less than Ethereum.
9
9
- Cheap verification and interoperability for Identity Protocols.
10
10
- ZK Oracles: With ZK oracles we can show that we have a piece of information off-chain and produce a ZK proof doing some computation with that data. Aligned reduces the cost of using those oracles. For more background, see the [following post](https://minaprotocol.com/blog/what-are-zkoracles).
11
-
- New credential protocols such as zkTLS based systems.
12
-
- ZK Coprocessor.
11
+
- New credential protocols such as zkTLS based systems: you can create proofs of data shown on your web browser and have the result verified in Ethereum. See the following thread for an [ELI5 on TLS](https://x.com/dabit3/status/1830022029195501799)
12
+
- ZK Coprocessor: ZK allows complex computations to be delegated from the blockchain to a coprocessor. This can retrieve information from the blockchain and perform the computations securely in a more efficient way.
13
13
- Encrypted Mempools using SNARKs to show the correctness of the encryption.
14
-
- Protocols against misinformation and fake news.
14
+
- Protocols against misinformation and fake news: you can generate proofs that an image or audio comes from a given device, and show that a published image is the result of certain transformations performed on the original image.
0 commit comments