Layer N: Hyper Performant and Hyper Composable Execution. An Interview with Co-Founder David Cao

Summary

In this engaging discussion, David Chow, co-founder of Layer n, shares insights on their cutting-edge Layer 2 blockchain solution designed to tackle scalability issues in the crypto industry. Chow discusses the evolution of Layer n, its unique features like the statet communication protocol, and the potential of AI and crypto intersection. Exciting partnerships and the future of Layer n are also highlighted, emphasizing the project's commitment to enhancing on-chain capabilities and user experience.

Takeaways

• 🚀 Layer n is a hyper performant, scalable layer 2 blockchain solution designed to tackle scalability issues in the crypto industry.
• 🌐 David Chow, co-founder of layer n, shares his journey from bioinformatics research at Harvard to building high-frequency trading systems on blockchain.
• 🛠️ Layer n's core innovation is the creation of a unique network of rollups with custom VMs that allow for exponentially more compute for applications while retaining seamless composability.
• 🔍 The platform aims to achieve feature and performance parity with centralized systems, but without sacrificing the core benefits of decentralization, such as permissionlessness and censorship resistance.
• 🤝 Layer n has gained backing from renowned players in the crypto space, including Peter Teal's Founders Fund, highlighting its potential impact on the industry.
• 🔗 The statet concept within layer n enables applications and rollups to share a standardized messaging pipeline protocol, facilitating seamless and instant communication between them.
• 💡 Layer n's approach to scalability involves a modular stack, with layer n focusing on the execution layer, Ethereum providing security, and partnered data availability solutions like IGDA.
• 🚧 The development of layer n is at an exciting stage, with major strategic partnerships and announcements expected in the near future, including a focus on integrating AI into on-chain applications.
• 🔄 The platform's unique Inter-VM Communication (IVC) protocol allows for the movement of assets between rollups and applications without the need for withdrawal periods or third-party bridges.
• 📈 Layer n's testnet has demonstrated the capability to achieve over 100k TPS in a closed test environment, showcasing its potential for high-performance trading and order book applications.
• 🌟 The future of layer n is promising, with a focus on practical engineering solutions, future-proof modularity, and a commitment to trustless, decentralized systems.

Q & A

What is Layer n and what are its core objectives?

Layer n is a hyper performant, scalable layer 2 blockchain designed to tackle scalability issues in the crypto industry. Its core objectives include increasing the surface area of what's possible to build on chain by 10x and enabling more compute for application developers, thereby allowing the creation of complex applications without worrying about computational constraints.

How does Layer n approach the issue of scalability differently from other blockchain projects?

Layer n approaches scalability by introducing a new model of execution layers and virtual machines (VMs). It focuses on providing unbounded compute surface area for application developers and allows for the creation of custom VMs optimized for specific use cases, leading to significant improvements in performance and efficiency.

What is the significance of the partnership with IG-DAO for Layer n?

The partnership with IG-DAO is significant for Layer n as it provides a solution for data availability, which is crucial for the functioning of Layer n's rollups. IG-DAO enables high bandwidth and storage, which are essential for handling the large volume of transactions that Layer n aims to process.

Can you explain the concept of zero-knowledge fraud proofs as mentioned in the script?

Zero-knowledge fraud proofs are a mechanism used by Layer n to ensure security and integrity of the transactions. Instead of running validity proofs on every single transaction, which can be expensive and time-consuming, zero-knowledge fraud proofs only require a validity check when a fault is detected. This method reduces the cost and time required for validation while maintaining a high level of security.

What are the benefits of using custom VMs in Layer n's architecture?

Custom VMs in Layer n's architecture allow for the creation of application-specific execution environments that are highly optimized for their intended use cases. This results in better performance, more efficient use of resources, and the ability to handle complex computations that general-purpose VMs may struggle with.

How does Layer n plan to address the challenge of liquidity fragmentation across different rollups?

Layer n addresses the challenge of liquidity fragmentation by implementing a shared communication protocol known as the inter-VM communication protocol. This protocol allows different rollups and applications within the Layer n ecosystem to communicate and share liquidity seamlessly, eliminating the need for third-party bridges and reducing the complexity for users and developers.

What is the current development stage of Layer n and what can we expect in the near future?

Layer n is at an advanced stage of development with a public test net and subsequent mainnet launch planned. In the near future, Layer n will be announcing partnerships with major liquidity providers and is working on integrating AI use cases into its platform, aiming to provide tangible applications beyond infrastructure.

What are the security considerations for Layer n's architecture?

Security is a key consideration in Layer n's architecture. Despite off-chain data availability solutions introducing a subset of Ethereum's security, Layer n maintains a strong security posture through its use of zero-knowledge fraud proofs and its partnership with IG-DAO. This ensures that the system remains decentralized and trustless while achieving high performance.

How does Layer n's approach compare with other layer 2 solutions like Optimism, Arbitrum, and others?

While Layer n shares the vision of decentralized applications with other layer 2 solutions, it differentiates itself through its focus on modularity, custom VMs, and a shared communication protocol. Layer n aims to provide a seamless and high-performance environment for developers and users, with a future-proof design that can adapt to new technologies and research findings.

What are the key features of the Nord VM developed by Layer n?

The Nord VM is a trading and order book-specific virtual machine developed by Layer n. It is designed to handle tens of thousands of trades per second, offering performance on par with centralized exchanges but with the added benefits of on-chain settlement and trustlessness.

What are some of the strategic partnerships Layer n has announced or is planning to announce?

Layer n has announced a partnership with SushiSwap for the development of a hyper-performance order book. Additionally, they are planning to announce partnerships with major liquidity providers, which will help bring more users and assets to their platform.

Outlines

00:00

🎶 Introduction to Layer n Layer and Guest

The paragraph introduces the discussion with David Chow, co-founder of Layer n Layer, a high-performance and scalable layer 2 blockchain designed to address scalability issues in the crypto industry. The host expresses excitement for the conversation, highlighting the importance of Layer n Layer's backing by prominent figures in the crypto space. David shares his background, from studying at Harvard to his involvement in bioinformatics research, and his eventual entry into the crypto world through building onchain order books. The discussion sets the stage for exploring the challenges and solutions in building high-performance blockchain applications.

05:00

🚀 Addressing Scalability and Liquidity in Blockchain

In this segment, the conversation delves into the core challenges of scalability and liquidity in blockchain technology. David explains how traditional blockchains limit complex application development due to computational constraints. He discusses Layer n Layer's approach to removing these constraints by allowing developers to build their own execution environments or virtual machines, thus enabling the creation of complex applications. The discussion also touches on the importance of a shared communication protocol for seamless interaction between different rollups and applications, emphasizing the benefits of Layer n Layer's statet concept and inter-VM communication protocol.

10:00

🔒 Shared Security and the Zero-Knowledge Fraud Proof

This paragraph focuses on the concept of shared security and the innovative Zero-Knowledge Fraud Proof mechanism introduced by Layer n Layer. David elaborates on how shared security assumptions among rollups enable seamless asset movement and communication between applications. He introduces the Zero-Knowledge Fraud Proof as a solution to efficiently handle disputes without extensive validity proofs for each transaction. This mechanism allows for a single-block proof process, potentially reducing withdrawal periods and enhancing the overall security model of Layer n Layer.

15:03

🧩 Positioning Layer n Layer in the Blockchain Ecosystem

The discussion now addresses how Layer n Layer fits within the broader blockchain ecosystem. David explains that Layer n Layer operates as an execution layer, focusing on optimizing performance and compute for application developers. He contrasts Layer n Layer with other rollup types, highlighting its unique combination of optimistic execution and zero-knowledge fraud proofs for proving. The conversation also touches on the partnership with IG-DNA for data availability, emphasizing the importance of high bandwidth and security in the choice of solutions.

20:05

🔑 Trade-offs and the Future of Layer n Layer

David discusses the trade-offs involved in Layer n Layer's approach, particularly the requirement to build within the constraints of the RIS Zero language set, which favors Rust but may limit initial programming language options. He also mentions plans to support a wider range of programming languages and emphasizes Layer n Layer's commitment to assisting developers in this new paradigm. The conversation looks forward to the potential of Layer n Layer to enable novel applications, especially at the intersection of AI and crypto, and hints at upcoming strategic partnerships and developments.

25:08

🌐 Exciting Developments and Partnerships for Layer n Layer

The paragraph wraps up the discussion with insights into Layer n Layer's current development stage and strategic partnerships. David shares about the upcoming launch of Sushi's soua on Layer n Layer, which promises to deliver centralized exchange-level performance with full on-chain settlement. He also teases major announcements related to liquidity providers and expresses enthusiasm for the potential of AI applications within the Layer n Layer ecosystem. The conversation concludes with David sharing his excitement for projects that focus on actual use cases within crypto and AI, emphasizing the importance of trustless systems and decentralized infrastructure.

Keywords

💡Layer N

Layer N is a hyper performant and scalable layer 2 blockchain solution designed to tackle scalability issues in the crypto industry. It aims to increase the surface area of what's possible to build on-chain by enabling more compute power for applications. In the video, Layer N is positioned as a significant leap forward in addressing the bottlenecks that hinder widespread adoption of decentralized finance (DeFi).

💡Scalability

Scalability refers to the ability of a system to handle increasing amounts of work, or its capacity to effectively grow and manage larger workloads without a corresponding degradation in performance. In the context of the video, scalability is a critical issue for blockchain networks, as it directly impacts their capacity to support a broad range of applications and users.

💡Layer 2

Layer 2 refers to a solution that operates on top of an existing blockchain (the 'Layer 1') to improve its performance and scalability. These solutions often involve creating a separate network or protocol that processes transactions off the main blockchain, thereby reducing congestion and increasing transaction throughput.

💡DeFi (Decentralized Finance)

DeFi, or decentralized finance, is a financial system built on blockchain technology that aims to provide alternatives to traditional finance by removing intermediaries like banks and allowing for peer-to-peer transactions. DeFi applications include lending, borrowing, trading, and investing, all conducted in a decentralized manner.

💡Custom VMs (Virtual Machines)

Custom Virtual Machines (VMs) are specialized computing environments designed for specific tasks or applications. In the context of Layer N, these VMs are optimized to provide more computing power and better performance for blockchain applications, allowing for the execution of complex operations that would typically be too resource-intensive for the base blockchain layer.

💡Composability

Composability refers to the ability of different systems or components to be combined or integrated in a seamless and interoperable manner. In the blockchain space, composability is important for creating a flexible and adaptable ecosystem where various decentralized applications (dApps) and protocols can work together and enhance each other's functionality.

💡State Machine

A state machine is a computational model that describes the behavior of a system as a sequence of states and transitions between those states based on inputs. In the context of Layer N, the term 'single shared state machine' refers to a network-wide computational environment that powers the execution of multiple applications, each maintaining a consistent view of the system's state.

💡Zero Knowledge Fraud Proofs

Zero Knowledge Fraud Proofs are a type of cryptographic proof that allows a prover to demonstrate that a statement is true without revealing any additional information about the statement. In the context of Layer N, these proofs are used to secure the system by allowing for the quick identification and correction of fraudulent activity without the need for extensive validation of every transaction.

💡Inter-VM Communication Protocol

The Inter-VM Communication Protocol is a standardized messaging system that enables different virtual machines or applications within a blockchain network to communicate with each other. This protocol is essential for creating a cohesive and interconnected blockchain ecosystem where assets and data can move freely between different layers and applications.

💡Shared Security

Shared security in the context of blockchain refers to a model where multiple rollups or blockchain networks rely on a common security mechanism or a shared set of validators to ensure the integrity and trustworthiness of the system. This approach can lead to more efficient security practices and reduce the duplication of efforts across different blockchain applications.

💡High Performance Computing (HPC)

High Performance Computing (HPC) refers to the use of supercomputers and computational techniques to solve complex problems that require significant amounts of processing power. In the context of Layer N, HPC is leveraged to enable the execution of complex applications on the blockchain, which would otherwise be impossible due to the computational constraints of traditional blockchain networks.

Highlights

David Chow, co-founder of Layer n Layer, joins the conversation to discuss the project's innovative approach to tackling scalability issues in the blockchain industry.

Layer n Layer is a hyper performant, scalable layer 2 blockchain designed to address the scalability challenges faced by the industry since its inception.

The project has the backing of renowned and prominent players in the crypto space, including Peter Teal's Founders Fund, which co-led the seed round.

David Chow shares his professional background, including his experience with bioinformatics research at Harvard and his initial foray into crypto through building onchain order books.

Layer n Layer's creation was inspired by the realization of bottlenecks in building high-performance onchain systems, leading to the design of a new layer 2 model.

The Tex Tech stack is introduced, highlighting the concept of a single shared state machine powered by a network of custom and optimized rollups.

Layer n Layer enables unbounded compute surface area, allowing developers to build complex applications without worrying about computational constraints.

The statet concept is explained, which allows rollups and applications to share a standardized messaging pipeline protocol for seamless and instant communication.

Layer n Layer aims to achieve feature and performance parity with centralized systems while retaining the core benefits of composability and permissionlessness.

The Inter-VM Communication (IVC) protocol is discussed, which enables seamless asset movement and communication between rollups without the need for withdrawal periods or third-party bridges.

Shared security is a key aspect of the communication protocol, ensuring that all rollups share the same security risks and allowing for the use of zero-knowledge fraud proofs.

Layer n Layer sits at the execution component of the modular stack, focusing on being the best execution layer in crypto for application developers.

The project partners with iG-da for data availability, leveraging their high bandwidth storage to support the project's ambitious throughput goals.

Layer n Layer's approach combines the best of both ZK rollups and optimistic rollups, creating a new category of rollups with zero-knowledge fraud proofs.

The project's use of Rust for building the XVM (eXtended Virtual Machine) is discussed, highlighting the language's benefits for security and performance.

Layer n Layer's Nord VM is introduced as an example of an application-specific VM, designed to settle tens of thousands of trades per second for order book use cases.

David Chow shares insights on the project's development stage, mentioning upcoming announcements and strategic partnerships, including Sushi's soua building on Layer n Layer.

The conversation concludes with David Chow discussing other exciting projects he's following, including the intersection of AI and crypto and the potential for onchain use cases.

Transcripts

[Music]

hi this is uncut Gems by Bitcoin Swiss

where we get into conversation with

leading crypto Founders that will enable

the next Leap Forward in our industry

and ahead of the curve uncut gems

provides you with insights on major

emerging narratives and offers valuable

one-of-a-kind discussions with

pioneering minds and today I'm very

excited to be joined by David Chow who

is co-founder of layer n layer N is a

hyper performant very scalable layer 2

blockchain that is designed to tackle

the scal scalability issues that our

industry deals with since the very

beginning and the scalability issues

that also hinder the widespread defi

adoption basically remarkably I want to

to state that layer and posts the seat

backing of very renowned and prominent

players like Peter Teal's Founders F who

co-ed the seat round David I'm very

stoked to have you today and thanks for

joining us thank you so much Dominic

it's a pleasure and honored to to be on

the show nice okay let's kick things off

and start maybe with your professional

background and maybe you could also

share with us the story behind The

Animated visuals on your website I think

they are pretty dope I guess like quick

personal background prior to coming into

crypto used to study at Harvard and used

to do bioinformatics research first got

into crypto with my same co-founders at

ler and actually but we first got into

crypto building onchain order books on

the salot of blockchain actually

initially and that was our first

introduction to the whole world of high

frequency not only high frequency

trading system but high frequency sort

of systems in general right especially

ones on chain and building that order

book and having that opportunity to

interact with Traders and market makers

and other system thinkers and designers

made us realize that even if we were

building on the plat the fastest

blockchain at the time and still today

there were still immense bottlenecks at

the infr layer that prevented us from

building what building essentially

something that could compete with modern

Financial Networks like NASDAQ New York

Stock Exchange Visa coinbase Etc so

basically once we hit that sort of wall

of okay if we want to build something

that can actually compete we can't do it

with the current stack we spent more and

more times going lower and lower down

the stack researching okay like how can

we actually scale up this onchain order

buck and that eventually led us down the

path of accidentally designing a new

layer 2 model and so that was like the

first story of the first Sparks of how

ler end started to become what it is

today and then after multiple iterations

and research Cycles it eventually became

what L is today which is this more this

sort of unique network of rollups with

custom VMS that allow exponentially more

compute for applications all while

retaining seamless composability

so really it was this like firsthand

experience of trying to build something

hyper performant on chain failing and

spending the time to realize how we

failed in do that and fixing the problem

that essentially blocked us from trying

to build that first thing so it was a

very natural sort of experience for us

and I think that's also something that

hopefully also makes us understand

Builders a lot more given that we've

also been through the whole process of

trying to build a protocol long chain

that's a nice background and so as I

understand basically the boundaries you

dealt with back then drove you towards

layer n and now you're aiming to solve

all of these the bottlenecks you

described like basically aiming to to

achieve a feature and performance parity

to centralized system systems but as I

also understood you still value like the

core features of the centralized systems

very high which is permissionless like

censorship resistance so you're trying

to combine the best of both words that's

a very nice story let's start with very

high level unpacking the teex stack bit

by bit like what's new in layer n why

should we be excited about it I know you

co the term statet a very powerful word

which is this single shared State

machine that is powered by that Network

you already described of course custom

and optimized Road UPS so before we dive

deeper into the Tex Tech maybe give us

some insights very high level on what

are the perks of such a system and why

should we be excited about it for sure

absolutely so I think there's really two

major unique unlocks right number one is

basically removing the bounds removing

the computational constraints for

application developers so historically

if you were to build any sort of

complex application arm chain it would

be really hard so that's sort of part of

the reason XY K became a thing right

because we had to find a simple way of

computing like a market making formula

without blowing out the gas constraints

on the EPM but basically it's a it tells

the tale of how hey if you're trying to

build anything more complex like you

just can't do that on chain at the

moment you can't do that on things like

the evm even the svm like I remember

back in the day when we were building on

salana we had to build a lookup table to

find a square root of a number right and

this is yeah like you should be able to

just find a square root of a number this

is basic math computations and so the

first core unlock that we enable by

allowing people to build their own

basically execution environments or

virtual machines as we like to call it

is that you have this like basically

unbounded compute surface area and you

can build very complex applications

without needing to worry about okayy I

need to fit all these computational

constraints in the that the EVN has

right so that's number one you're now

able to build very complex stuff we can

even build all the way to things like

incorporating like AI models too right

which I'm sure what will'll talk about

it some point on this show given how

much AI is of Interest now that's the

first unlock the second unlock is none

of this would be useful if it can't

share liquidity and communicate with

everything else right so like we've seen

like this Paradigm play out of everyone

building their own individual rollups

the whole problem with that is

everything then gets fragmented right

liquidity gets fragmented the user

experience is fragmented you need to

work with third party Bridges to move

from one roll up to another it's just

overall not a good experience right and

so the second core unlock that we have

is basically the statet concept which

allows each of these rollups and

applications to share a standardized

messaging pipeline protocol that we call

the inter VM communication protocol that

basically allows applications en roll s

to seamlessly and instantly communicate

with each other right and so now you no

longer need to worry about 7day withdraw

periods times you don't need to worry

about interoperability and bridging

protocols and stuff like that all of

that just like works right so like me as

an application developer I either build

my smart contracts on one of the evm

rollups or I build like my I launch my

own rollup with my own custom VM and

everything just works right these just

compose just as if I were building on a

monolithic blockchain so really like the

goal is hey like how can we like you

said earlier how can we accomplish

feature parity and performance parity

with centralized systems all while

retaining the benefits of composability

that you're used to on the monolithic

system so that's really yeah so

hopefully that was clear yeah perfect

answer and what you described earlier

very much reflects what I observe or

experience as a user myself right like

it's it's painful if you're like stuck

on different rollups and seriously

sometimes I find myself questioning like

how should I describe or explain that to

something somebody that is new in the

space while avoiding to get fished or

something like that on the path towards

briding from all these it's a thing and

it's

interoperability and all of this

liquidity fragmentation and and so forth

remains one of the core pain points

within the industry in my opinion and

it's it's awesome to see projects like

layer and tackling these very challenges

of our ecosystem so layer n kind of

strives to enable this Melting Pot of

VMS As I understood and rollups on the

back of this like you said shared

communication but also liquidity layer

maybe we can dive into BM communication

protocol which you call IVC and maybe

could explain like how does it work and

how does it compare with other Solutions

are there for instance Concepts like

shared security or something like that

please guide us through the very concept

of this liquidity and communication

layer so at

the base of the communication protocol

is what you said the concept of shared

security right so we wouldn't be able to

accomplish

a seamless the the features of the

seamless communication protocol without

some elements of shared security

assumptions so basically what that means

is all of these rollups or all of these

rollups all settled to the same sort of

big Global like State machine right so

that's also where the word State comes

from it's this idea thaty you have this

single big state machine that's

separated between this like networks of

applications and that's a really

important assumption right it's like you

need shared security otherwise you have

problems if someone were to run their

own let's just say op stack roll up and

someone want to run another optimistic

or even ZK rollup because they both have

different security assumptions you need

that kind of 7-Day withdrawal period

from the optimistic rollup or you need

some kind of third party bridging

provider to take on the risk and the

liquidity providers take on the risk of

moving assets between the rollups right

and so all of that is solved once you

have the shared security assumption

right so all the rollups share the same

security risks which means that okay

like if one rollup has an issue you

simply so we used this new thing called

zero knowledge fraud proofs I think we

were one of if not the first to to push

this out but basically it's the idea of

instead of doing validity proofs on

every single transaction like current

zero Dodge rollups do and instead of

doing like interactive fraud proofs like

the current optimistic rollups tried to

do and I say tried do because I don't

think anyone has like public

implementations of interactive fraud

proofs yet but basically the Z non fraud

proofs is very elegant way of basically

only running the validity proof when

there is fault so that allows us to do

the proof in a single block and it's

also a lot more straightforward than

thinking around all the game theory

economics of the interactive fra proof

so anyhow so basically yeah like if one

rollup has an issue any valid leader can

replay the transaction history from that

they get from the da layer identify on

which R up that issue occurred and at

which transaction then submit that ZK

fraud proof basically to be like hey

like there's an error here and if there

is an error then we would go into sort

of the state roll back process now to

get back into the sort of the

communication layer so now that we have

this sort of shared security

assumption we can now seamlessly move

assets in between rollups and

applications without needing to about

those like withdrawal periods and those

bridges right and the key towards doing

that is to have some kind of

standardized way of assing and

pipelining

messages from one application slash

rollup to another and basically that's

what we're building in house so think of

IBC on Cosmos but then like just remove

all of the consensus components right so

basically have you have a very

straightforward problem of okay like I

have one roll up I have another roll up

like how do I create a standardized

format of sending one message that can

be any sort of arbitrary bites to

another rollup right so then that if you

like distill that problem to to the just

that it's like a very similar just like

Web Two communication problem and at

that point once you have a standardized

way of doing it it's very seamless for

everything in your network to

communicate with within each other like

on a very fundamental level it's I mean

you say that it already it pretty much

reminds me of the early concepts of

Cosmos and paulod do at least from the

idea and the approach like pretty dope

that you manag to to to abstract away

the the different VMS and pro provider

somewhat like shared security shared

liquidity layer but also have

composability like solve the comp

composability Problem by you already

touched base on the whole proving

concept which was very new to me I know

the the XVM We Touch based on the VMS

and what layer n enables within that

session as well but the XVM that you

call XVM are built in Rust and are

enabled as I stood As I understood by

risk zero and if we look this modular

stack periodic table with the different

rollups if we think about layer n maybe

help us a little bit if we look at this

modular St table how would you classify

layer and is it a settlement rollup is

it a sovereign rollup and then the

second aspect or component I'd like to

cover here is the very fraud prooof CK

fraud proof life cycle you said because

until now I was used to either I have a

validity proven system or rollup or a

fraud or I use fraud proof proofs but

now you combine these and create some

kind of magic with huge benefits being

that you don't have to to prove

everything as you mentioned maybe you

could yeah help us understand that

concept a little bit better and where

you would classify layer n like how can

we think of layer n in that regard for

sure so in terms of the modular stat

right we sit at the execution component

so I like I would say that we're like an

execution layer right so we're purely

focused on hey how can we make how can

we be the best execution layer in crypto

where application developers can come

and just purely focus on building the

best applications right so if you think

about it in the modu periodic table like

you said there's we have the execution

component which is us the security

component is on ethereum and then the da

layer we're currently partnered with

igen da to use in this sort of Validia

model right so that's like the the

general component with regards to the

fraud proving question so that's like

the really cool thing about what we're

doing right we don't really fit in the

camp of like fully ZK rollups we don't

really fit in the camp of the typical

optimistic rollup we combine The Best of

Both Worlds to create this new category

right and I wouldn't we're more akin to

optimistic rups in the sense that we

still run execution optimistically but

when we talk about proving it's really

this sort of new category that we called

the zero knowledge fraud proofs right

and then the basic idea is really hey

doesn't make sense to run validity

proofs on every single transaction if

you're trying to optimize for

performance because ZK proofs are

expensive they're timec consuming to run

and if you're trying to do anything like

what we're doing which is like hundreds

of thousands of transactions per second

you're not going to be able to do that

and if you were to do that you're

probably paying millions of dollars to

AWS to run your proving costs right or

something like that on the other hand

you have the interactive fraud proof

which

have historically been really hard to

implement and then we've seen that in

practice if you look at any of the

current optimistic rollups I don't think

as of the date of this interview that

any of them have public implemented

fraud proofs right so optimistic has no

fraud proofs so you're basically running

on the that you trust op Labs which is

not a bad assumption but it's it that is

the Assumption arbitrum has wh listed fr

proofs right so it's not public so it's

only like a wh list set of validators

that can run so that's if you think

about it from a game the pers or I guess

from a theoretical perspective it has

the same properties as not having fraud

proofs at all because basically like

only insiders can run fraud proofs which

is the same thing as if the Insiders can

like did have fraud proofs but again

it's like not a terrible assumption if

you trust arbitro but what I'm trying to

get at is interactive problems are

really complicated right because they're

multi processes it depends on this like

interaction game basically between

theover and and the verifier instead of

doing all of that complicated stuff

right we say hey like validity proofs

are great but validity proofs are only

good if you can minimize the amount of

time that you actually use it right so

instead of doing it at every transaction

we'll only do it when a fraud occurs

right so when a fraud occurs a validator

simply needs to rerun that state

transition on like they can run it on

their own zkv on their own MacBook if

they want right Y and then basically

through the RIS Zer zkm they get a hash

that they can then use to submit on to

an onchain smart contract that verifies

the legitimate the legitimacy of that

hash and if it is legitimate then hey

you've got a valid proof right so it's

that straightforward and the cool thing

is hey like now we can do this within a

single block as opposed to a multiblock

process and did theoretical interactive

fraud proof which also means that you

can theoretically reduce your minimum

withdrawal period by like a significant

amount right so like the whole problem

with the 7-Day withdrawal period 7day is

arbitrary but it's also based on the

idea that like hey if it takes like x

amount of blocks to prove fraud and each

block has some probability of being

censored right what's the probability of

every block being censored like multiple

times and how much time do we need to

ensure that all of these blocks go in

right now if you only have a single

block that you need to worry about the

risk decreases significantly right and

so that's why we say hey theoretically

like you can decrease the withdrawal

period Times by quite a significant

amount while decreasing the actual risk

of censorship itself so hopefully that

gave you like a an overview of of where

we land in terms of the security

Spectrum

there yeah yeah it definitely did but it

also like to be frank it sounds like

almost too good to be true that you get

like this massive

kind of bump in in withdrawal periods

but also reduced risk at the same time

and what you stated earlier I'm on the

same page like it I I would be a little

bit more critical and say it's even a

problem that maybe optimism and other

rollups

still do not have these fraud proofs

implemented yet I guess arbitrum is on a

good path here and optimism all of the

of these rollups still have training

wheels on let's be honest and I hope it

this will get better because we want to

build like these sound decentralized

permissionless immutable systems what

now I would like to know you stated that

you partnered with I da for the data

availability kind of aspect would you

mind giving us some insights on why you

chose IG da and not other Solutions like

aale or Celestia and yeah you touch Bas

on it but I would like to know what to

understand it maybe a little bit better

so layer n is only the execution part I

da a you use for data

availability settlement is still

happening on ethereum is that

correct yes that's correct so basically

layer Ed is still in L2 right because

we're at the layer two in terms of the

hierarchy chart so like our core

function is how do we make execution

better right execution better in the

sense of performance and compute as well

as that's the Mac feature ex and then

execution better also in the sense of

composability right and then as any

rollup does We Roll Up the state and

then put it back to ethereum where it

can be verified and challenged in case

there are problems right and then the

main difference is instead of Hosting

data availability on ethereum which a

lot of the current rollups do we

actually decide to do it with an off d

solution like i d right and the core

unlock there is really like a math a

math problem right it's like Ian

da will be able to enable up to 8 to 10

like megabytes per second that's crazy

storage bandwidth right yeah and the ma

there is just hey yeah you look at

something like Celestia it's currently

doing 1.5 megabytes and it just if

you're trying to reach the scill at

which we're trying to do which is tens

of hundreds of um tens thousands of

transactions per second you need

something that has high bandwidth right

that's just a prerequisite and so it it

really just boiled down to that I think

the other really cool thing about I da

forces the whole reaking component to it

right one of the core things that people

are worried about with these offchain da

Solutions is like hey like now we're

losing some security there right which

is true if you're not storing your data

on ethereum you're inherently going to

inherit like a smaller a smaller subset

of the security that you would assume on

ether let's just assume ethereum is like

the most secure decentralized system

right like any other system is probably

like a subset of that right and then

even I da which is like resters right

it's still going to be a subset right

unless like the whole set of e theorum

validators are restak into 8 right let's

just assume it's like a sub said so yes

that is like the tradeoff that you would

be taking right but from our perspective

it is a strong enough tradeoff and a

beneficial enough tradeoff that it makes

a lot of sense right if you're getting

like a hundred times better performance

but just the midle school fraction

amount less security and even then we

can argue as in really less security and

I would say not really I think the

tradeoff is very valid and I think we

can play this out and see what types of

products people will use more often but

I think until we're able to reach

performance parody like we still have a

long ways to go from a prod Market fit

perspective if we're really if we don't

reach that performance parody threshold

quicker

so that was I lost track of the original

question oh I think you were asking

about the a DA stuff yeah so basically

basically it just came down to a math

problem right okay like we just want

wherever has the most amount of

bandwidth yeah and then two where had

where it has the where it's the most

secure as well so the nice thing about I

is you still have some some strong

subset of each security as opposed to

some of these other systems who need to

bootstrap their own set of validator

nodes which could be a bit less secure

right yeah great take thank you because

you touch based on tradeoffs what is

your opinion on the your approach now is

using these like or rather novel I'm not

sure if they are battle tested yet novel

CK fraud proofs what is a potential

trade of you using that kind of proving

system in your opinion or aren't there

any yeah for sure the main tradeoff

would be we need to

build the VMS and the rollup component

such that it's it is it is basically it

that what's the word I'm blinding it it

basically works with since we're using

RIS zero right which requires the risk

Live Language set basically we need to

build within those constraints right so

we wouldn't be able to out of the box

build something in like JavaScript or

something right so that's a major

constraint because it means that hey

like the initial set of rollups and VMS

we're going to have to build a rust with

that being said also huge rust Maxi and

I think the world should move towards

rust but that's a separate topic with

that being said we are thinking of of

implementing wasum as well at point

which will allow us to support a much

wider breadth of program programming

languages for people who would who might

want to build applications in different

languages and then the other note I'll

make because this is a common point of

confusion when people ask me hey like

what programming languages is it is

layer and in like we build everything in

Russ right but that doesn't necessarily

mean that the application developers

need to build in Russ so depending on

what type of application the application

developers build wants to build it could

be different right so we will have our

own evm rollup as well that's going to

be built in R East which is the rust

implementation right but if you're

deploying a smart contract on that evm

rollup that's in solidity right so all

of your pre-existing code all of your

pre-existing sort of solidity and smart

contract knowledge you can still use

that right with that being said if

you're trying to build a an XVM and your

own your own rollup that's where you'll

need R and at the current stage we're

willing to partner really deeply with a

lot of the people that we work with in

that process to get them started just

because we know that okayy this is a

completely new way of Designing things

right one that maybe most may not be

familiar with and so we're happy to

spend that time deating the initial

course set of developers and helping

them out a lot in that process nice so

as I understand while you are

constrained in the programming language

for the XVM which is which kind of build

that network of different virtual

machines right you on the other hand of

being constrained by the programming

language you get a huge design space By

by allowing crazy optimizations of these

generalized but also application

specific VMS and on the other hand you

still provide As I understood with the

NM still provide a VM environment for

developers who want to build in solidity

which is super nice because doing that

you at one hand you wipe out the

boundaries of operating within the evm

but you also allow developers to build

within solidity because you also have

the NM so with these X VMS what design

freedom and benefits do you achieve

exactly maybe we you can give us some

insights on that and what are the perks

what are the perks between the

because you also allow for generalized

VMS but also application specific ones

maybe you have some cool examples to

share I think like the

easiest

mental model that I like to use to think

of XVM and more generalizable smart

contract VMS like the evm VM Etc is okay

like imagine you're trying to build some

kind of you're trying to build a car

right the evm is basically a scenario

where like someone comes and gives you

all these generalizable Lego blocks

right you can use these Lego blocks to

build a house you can use these Lego

blocks to build a table you can use

these Lego blocks to build a car right

because they're generalizable right

these are Lego blocks you can use it to

build anything right they might not be

the best right if you're trying if

you're trying to build a really solid

house or a really fast car like it might

not be the best right but it it does the

job

right and it's generalizable and that's

why people like it and it's

straightforward to use the XVM is

basically like we don't have any Lego

blocks for you but we have all of these

primary resources right such as wood Etc

you go build like whatever car you want

from the ground up in the way that you

want to right and that really is the

core thing right like people typically

think about scaling as this like as only

two things right like throughput and

latency right but people often forget

the third thing which is compute right I

mean comp is really important when

you're trying to build anything more

complex right and that's really

constrained on what the sort of the

virtual machine environment allows you

to do right so when you have very preset

ways of doing things on the EVN that

makes it really hard to build things

that are very computer intensive so

imagine basically the XVM as this place

right where you can pretty much just

build an application like without

thinking about these like blockchain

constraint right like just build the

application as you would build it if you

were building it on web to right and one

of the examples of this is actually the

Nord VM that we built ourselves in house

which is basically the first order book

specific and exchange optimized virtual

machine and really what it does is it

only does one thing right and it does

that one thing really well which is

basically settle tens of thousands of

Trades per second at something else I

can lay and seas on this like massive

like order book right and and basically

the reason it's able to do that is

because it was purposely built for that

so it was actually modeled off of elmax

digital which is I think one of if not

the fastest current like institutional

centralized exchange and we're able to

do that because we're not building

within the constraints of the evm and

you get to just build whatever you want

and so that's like uh that's basically

like how I like to think about the

difference between xvn and and a more

generalizable virtual machine and I'll

also think and the other thing I'll not

about the XVM is like the XDM is also

not trying to be like a generalizable

platform right so you don't need to

think about a lot of the problems and

inherent challenges that comes to

building a generalizable platform so you

can really just be super application

specific focused and not worry about a

lot of the other constraints that the

EDM needs to worry about right like how

to create basically generalized way of

doing everything you don't need to worry

about any of that basically yeah did

that kind of answer the question yeah

fully fully satisfied here David yeah

it's very cool I imagine as a developer

it's awesome to have a blank canvas and

build whatever you want because what

tailor your solutions to towards your

application that's really nice by

coincidence in the previous episode of

uncut gems we had Keon from from Monet

who kind of shares a similar background

stemming from Solana slh high frequency

trading and as better than me probably

parallelization is a very powerful tool

to scale integrated chains and there are

some more areas worth tweaking to boost

the performance such as database

consensus or pipelining like monut is

doing and I feel to me it's very

interesting to observe the developments

and Industry from the sideline and

observing that there there are these

teams choosing the modular Direction and

then there are the other teams like

fully committed to the integrated

approach and clearly ler n went with the

modular narrative right so would you

mind providing some insights on your

decision making progress and why you

eventually went with modula instead of

integrated so please guide us through

some pros and cons of these two very

exciting exciting developments

the best way to think about our decision

making as a company and as a team is we

have one mission in mind and that is you

have the current sort of surface area of

what's possible to build on chain how do

we take that and increase it by 10x

right that that's not 10x you get like

how do we take the current surface area

of what's possible to build on chain and

vastly increase that right not just we

don't just want to increase it by 1.2x

1.3x 2x we want to increase it by 10x

and to us like the key towards unlocking

that really just comes down to compute

right like you enable more compute for

people that's going to allow people to

do so much more and to do that we need

to think in a completely different

Paradigm and a completely different way

about building applications on chain

right and the truth is yeah like I think

what what a lot the other teams are

doing in terms of improving the evm is

like super core and super important

right because there's a whole bunch of

people who are still building on the evm

the problem is hey like that's not going

to take us to a 10x in terms of compute

right really that's what we're doing

there is we're updating an existing

model but we're not thinking outside the

box in terms of like how can we create

an entirely new model that's just so

much better right so it's the same thing

as being

like it's a horse to the race car right

so there's this like really thing that I

think Ford said at some point in

time maybe it was someone else but I

think it was Ford and then he said

something along the lines basically of

hey ask people what they want today

they'll tell you that they want like

better horses right or like Faster

Horses and stuff like that right but

really what they want is like this new

thing that's called the car right so

it's like that that that's essentially

how we think about this dichotomy okay

nice so your take is to approach you are

taking is the superior one and with less

kind of less constraints like basically

the next chapter not only an iteration

but the next chapter that's cool let let

me try to consolidate a little bit what

we what we talked about until now you

combine I da which gives you like an

insane level of throughput with the stat

net which is a network of custom VMS and

that very statet is powered by shared

communication and liquidity layer only

stating that is already very exciting

because we obtain this super highly

performed L2 with also as I understood

like very low

latencies and I read in January you

announced that layer ends S engine which

is also a custom rust VM As I understood

that you manage to achieve like 100k TPS

on a close test net and I know it's a

test net but I still think it's beyond

remarkable to achieve like 100k TPS and

now I wonder where these TPS where is

full-fledged

transactions including kind of smart

contract logic or was it vanilla value

transactions and then maybe you can

provide some insight on insights on this

n engine that I just mentioned because

it's a new trading optimiz optimized

rollup engine and maybe one of the the

first examples on how highly optimizable

rollups will be on layer n for sure so

the Nord engine or the Nord VM as we

like to call it now we went through

different iterations of the name but but

they all mean the same thing right

basically it's this custom execution

environment custom VM custom engine that

just does one thing really well which is

like the order book use case and as you

brought up we ran this close test set

with the IG team actually were all we

were all live on the CLA at that time

and then we were running it and and

everyone got super excited when the

numbers came out but basically yeah like

it it was basically it wasn't just empty

messages it was trades right real trades

on the order book and I think we went up

to 120 something thousand number I

forget that it was 120 and something but

yeah so it's and it's a very meaningful

number right because I think like when

we think about scaling

the history of scaling on ethereum we've

always just dealt with hey how do we go

from 10 to 100 or how do we go from 100

to 700 how do we go from 700

to or I think the ma I think even with d

sharting the math comes out to about

you're only supposed to reach like

around 800 transactions per second

across all rollups or something like

that I think I don't remember the exact

MTH but some low

number full d Proto de sharting with

full sharting that might be different

but with Proto I think it's still just

800 or something but basically my point

is we're in like the ballpark R we're in

like we we just we haven't dreamt bigger

right and so I think this is very

meaningful because even salana does I

think 6,000 TPS or something like that

and then I think 6,000 that's accounted

for the vote transaction so not I think

without the vote transactions like

actual real TPS I would need to check

what it is but I think it was like 500

or like a thousand or something like

that yeah it's been a while since I

checked those stats but it's definitely

not 10,000 not even 100,000 we are

talking about an entirely different

Ballpark and the elegant thing to me is

while you innovate so much on this kind

of VM layer but also on the liquidity

layer and the composability aspect you

still incorporate what is very important

to me good amount of security and

decentralization from ethereum so that

is is pretty appealing to me at least I

would like to as we inch closer to the

end of this nice conversation David I

would like to your take on as I see it

layer n kind of builds what also

arbitrum orbit and stylus sets out to do

what optimism sets out to do with the

super chains what CK sync does with

Hyper chains but with a few like very

interesting and very elegant like

implementation differences so how do you

see yourself compared to these I will

not say competitors but other projects

that kind of aim to for a similar end

game and how do you compare when it

comes to decentralization when it comes

to security when it comes to training

wheels when it comes to

orifices aspect o ification for rollups

because just now with the tenun upgrade

coming it it proved that like imagine

there's a rollup that was aifi already

and not upgradeable like they could not

Implement blobs so sorry maybe that's a

very comprehensive question but maybe

would be nice to get your take on kind

of comp

landscape yeah for sure I think everyone

has this shared vision of eventually

things are going to move onto their own

rollups it just makes a lot of sense

right like the congestion that you face

from a single model and doesn't make a

lot of sense I think a lot of it just

ultimately comes down to implementation

details right because a lot of these

other projects they' started with a

certain fion in mind it's a lot harder

to easily and really quickly progress

towards this very scalable model that

I've described under the state net right

if you look at the super change stuff

super cool super it's great what they're

doing the problem is like there there's

like a lack of I guess like

implementation detail right so

everything that we've talked about in

terms of the communication protocol the

shared liquidity system like those

things are lacking right and the reality

is it's not it's just a really hard

problem right when you you have all of

these like independent rollups being

deployed and you don't there isn't this

pre-existing Network that brings them

all together right so you have this kind

of like situation where like you already

have a bunch of rollup that you're

trying to figure out like afterwards

like what are pipes that we could put

together but it's really hard because

they all have their own security and

they all want to do things their own

ways and so it's very hard whereas like

I think like where we come into play is

like really at making like the

implementation like Flawless and making

it right from the start and also making

it modular enough such that it's very

future proof so like think that you

talked about the con of tication is

really interesting but you also don't

want to do it in such a way that like

hey like two years from now some new

technology comes across some new

research unlocks some some new

technology you want to be able to

implement that so I think I think that's

really important that's like part of the

technology cycle right and so a lot of

our philosophy is based just on that

right what is the most practical

engineering solution towards achieving

trustless and that's really like how we

think about designing Solutions rather

than okay like rather than taking this

maybe this more like philosophical

approach of saying okay like there's

this one concept that I just really and

I'm just going to try to build things

around that instead of being like that

we like to think about what are the

problems and how do we build the best

solution out those problems so yeah

hopefully that answered the question but

basically I think it really just boils

down to

implementation and how we think about um

piecing all these all these pieces

together in a way that actually works

really well yeah it's it's certain does

thanks for the explanation and now that

we are all hyped up including the

audience maybe you can share some

insights on the current stage

development stage of ler n and can you

share some exciting strategic

Partnerships already with us I saw that

Sushi's soua is building on layer in

already so that's pretty cool but maybe

there's more to share I don't know for

sure yeah so number one

is this coming month is going to be

packed with a lot of exciting stuff so

if you're not following us already on

Twitter if you're not in our Discord and

telegram group chats already make sure

to join cuz we have a lot of exciting

stuff that's going to be announced Susa

is the first major announcement and

we're super super excited about it

because it's not just any other decks

it's the first hyper performance order

book that matches centralized exchange

performance but not only that it also

fully settles every single trade on

chain right so if you think of any of

the current offchain order book besides

like AO vertex hyper liquid etc

etc they're fast but the problem is they

only settle maxed trades on trade right

so all of the trades that you place like

Place cancel order the vast majority of

the trades aren't actually settled right

so it's not a fully onchain and

trustless system whereas Susa is not

only going to be like the fastest it's

also going to settle everything on shap

meaning you can actually fully verify

everything right down to every single

place order cancel trans action that you

make which is going to be pretty

gamechanging from a security perspective

but also from a feature perspective

there's a few things that will be major

unlocks in the deck space that are

uniquely that will be uniquely

implemented by their design and their

utility of this unrestricted compute

surface that we're enabling for them by

being their own XVM so we're super

excited about that announcement and

there's a few other announcements with

major liquidity providers that are

coming up as well when we talk to

developers and being a developer before

as well one of this not the most asked

question is just hey like where is

liquidity going to come from who could I

work with to get liquidity and where are

the user come from so that's all going

to be solved there's going to be major

partnership announcements coming up

ahead and we have yeah like huge

Partners to help bring in liquidity on

that front so I don't want to tease too

much but basically lots of exciting

announcements coming up in the next next

few weeks make sure to be following

along to stay up to dat and

yeah very cool very exciting and happy

to hear that such a kind of pioneering

project is close to launching some and

announcing some big stuff maybe last

question be before we end this cool

conversation outside I hope I don't

catch you off guard but outside of leer

n what is the most exciting project you

are following and why oh

that's a good question you're asking me

about yeah no that that's a really good

question there's a few really exciting

things I think modad for one is has a

very different approach to how we're

doing things but I think the sheer focus

on engineering and a lot of the

improvements that they're bringing to

the evm are actually really exciting so

that's the part that that I'm really

excited about modad I think the next

generation of

is also really exciting because up until

now I know there's a lot of dexes that

constantly get launched but up until now

like there really hasn't been a DEX that

truly Rivals like a centralized Exchange

in terms of being able to provide the

same level of user experience the same

set of assets the same set of of

features right and then even providing

more than the current set of features

there really isn't anyone that even

holds a candle to like coinbase Finance

Etc right so I'm really excited about

someone to do that wink on layer end and

the other thing I'm really excited about

is actually the intersection of AI and

crypto I think it's gotten like a bad

rep because of a lot of I think as

Anything Grows enough hype right there's

always going to be like a strong amount

of sort of noise around it but if you

think about actual use cases and what

actual

onchain the centralized inference can

enable it gets really exciting so one

thing that we're thinking and we're

working on behind the scenes or not one

thing a few things there's a few AI

partners that we're working with behind

the scenes as well to enable some actual

onchain use cases and I don't want to

spoil the sort of the announcements but

when we think about Ai and we think

about the intersection of trading and

gaming that's where I see like a lot of

the really interesting products and

applications and there's a few partners

that were working on making that happen

and that can uniquely happen thanks to

the fact that we can enable a lot more

compute allowing you to call these

basically AI functions right into these

like AI models and you can work with

with a zkl team to do all the imprint

proving and so on and so forth and now

suddenly you have ai running on chain

fully trustless right so that's actually

something we're super excited about as

well nice I really love your take

because I seriously believe like after

all of these years that we are in now in

crypto and digital assets it's also it's

time it's finally time for some actual

use cases right maybe we at some point

move a little bit away from having that

major focus on infrastructure and maybe

instead provide some very actual use

cases as you said with gaming or Ai and

with a with the current Bonanza around

AI it's very exciting to hear that you

have some Partnerships going on and I

know it's hyped it's hyped up for the

and the reason basically for me is like

two industries that are highly

speculative but don't have too many

tangible use cases yet are clashing

together which is AI crypto and the nice

thing about about that occurrence is

again that it shows us like as with many

instances say deep fakes or cbdcs that

blockchain Solutions and decentralized

systems are insanely important for

future use cases such as AI being a risk

mitigator being an enabler for

democratizing AI or things like that and

with that David thank you I really

enjoyed hosting you today I guess our

discussion ranged from very complex but

also to to digestible topics it was

super insightful on behalf of Bitcoin

Swiss and I also think the entire

audience thanks for sharing your

valuable time with us and also best of

success obviously with the challenges

ahead and a smooth public test net and

subse subsequent mayet launch David

thank you so much Dominic this was a lot

of fun and I hope I I didn't bore your

audience too much with my

ramblings not at all and finally to our

viewers if you're interested in learning

more about layer n be sure to visit

their social streams follow them on

Twitter read their blog posts to make

sure that you don't miss the

announcements I will provide links in

the description below and yeah until

next time thanks for watching

he