Anza D1: The Future of Solana Core Development

The Evolution of Solana Core Development

Solana's core development has undergone a significant transformation since its inception. In the early days of 2021 and 2022, Solana Labs was the primary driving force behind the network's roadmap and decision-making process. However, the landscape has evolved considerably, with multiple entities now contributing to Solana's development and governance.

Top Projects

Lulo

Score: 7,586.13

Bonk

Score: 4,806.12

Sphere

Score: 2,544.42

Moonwalk

Score: 2,026.11

Brennan Watt, VP of Core Engineering at Anza, provides insight into this evolution: "On the culture front, I'm happy to say that I don't think it's really changed a whole lot, even as we've grown. I think there's like a really strong, awesome winning culture that's been set from the top." This culture emphasizes shipping changes and services that positively impact user experience as quickly as possible.

However, the process side has necessarily matured. Watt explains, "The number of like smart people that I see walking around our core dev meetups is absolutely insane. It's awesome. But also smart people don't always agree on everything. And so, you know, we need ways of settling some of those technical disputes and coming to consensus and making sure that we make progress."

This increased complexity in decision-making has introduced more friction into the development process. While this can slow things down, Watt argues that it ultimately leads to better decisions. He cautions, however, against letting this friction ossify the current state of the protocol: "I've seen this happen some places before where like you think doing nothing is the lowest risk thing. And I think that doing nothing is actually the highest risk thing. We have to like continue iterating, continue making things better."

The Role of Jito Labs in Solana's Ecosystem

Lucas Bruder, co-founder and CEO of Jito Labs, provides a complementary perspective on Solana's development. Jito Labs builds software that stands on the shoulders of giants at Anza, focusing primarily on web software. Their goal is to help make the network more efficient and help validators and stakers earn more money.

Bruder emphasizes the importance of Solana's culture of grittiness and hunger: "There's a lot of grittiness and the meme was chewing glass. There's still some element of chewing glass today, but it's gotten much better. And I think that the grittiness is what attracted me to Solana in this green field and a lot of that still exists today."

This culture of continuous improvement and rapid iteration has been crucial in Solana's ability to address challenges and improve performance. Bruder notes, "I feel like we're straddling the line of trying to take long-term long-term views on protocol changes, but also getting a lot of stuff done and trying to find easy ones where you can that have really large improvements on the network."

Solana's Recent Performance Improvements

Despite the increased coordination challenges that come with a more decentralized development process, Solana has made substantial improvements over the past 12-18 months. Watt attributes this to investments made a couple of years ago that are now paying off: "A lot of what enabled that is some of the infrastructure and testing and resiliency and some of those things that are harder to notice like paying those upfront that now allow us to move faster."

The V2.1 rollout of Agave, the client that powers Solana, brought noticeable macro changes to the network. Watt explains, "Some of the important metrics are like the slot and epic times like you mentioned. And so, you know, most people know that Solana targets a 400 millisecond slot or block time, they may not know that it doesn't always mean that it actually achieves that 400 millisecond slot time, like the real world can get in the way of that."

These improvements have led to shorter slot times, which translate to lower latencies and a snappier user experience. Watt adds, "It also can mean tighter spreads on things, you know, the shorter latencies are kind of it reduces some of the information asymmetry. And so I think a lot of people don't appreciate that sort of a thing. It also helps with like better validator economics, you know, the more transactions we're pumping through economics kind of get better there."

Another significant improvement has been the reduction in skip rate. Watt explains, "Skip rate is basically a function of like how many blocks actually got built on the canonical chain versus how many chances did we have to build a block or like how many slots did we have?" This rate has dropped from around 10% historically to now being around half a percent on a bad day, often lower.

The Importance of Client Diversity

The introduction of Firedancer, a second client implementation for Solana, has been a crucial step in improving the network's resilience and performance. Watt envisions an ideal future where "we have four evenly staked clients and any given client could hit some catastrophic failure and the network just keeps humming along and, and no user has to notice."

While this ideal scenario is not yet a reality, the current progress towards client diversity is already yielding benefits. Watt explains, "I do look forward to a day where Frank and Dancer has more runtime, more, more trust and eventually Firedancer and we have like a fairly even stake split on main net. And I think, you know, it's, it's still de-risking, still just having a second client implementation."

The diversity in client implementations and versions helps uncover issues that are hard to discover with a homogeneous client base. Watt likens this to having "more than, you know, just a couple variations and you throw the, the Gito Solana validator in there. It's like, okay, we've got enough diversity now that you're, you're kind of blending all these things together and finding bugs early."

Anza's Roadmap for 2025

Looking ahead to 2025, Anza's primary focus is on increasing block space. Watt emphasizes, "We want to go increase block space. So it's the, you know, IBRL tagline that we love to to me and so often on X, but it really is an awesome rallying force for the entire team."

This focus on increasing block space translates to pumping up Compute Units (CUs) or the total transactions that can be packed into a block. The benefits of this are clear: lower median fees and easier transaction inclusion, especially during peak demand and times of congestion.

Watt is emphatic about the scale of these improvements: "We're not improving some BS synthetic workload metrics or anything like that. Like we're doubling real throughput for things that people actually care about like swaps and, you know, like meaningful, complex transactions. And so this will be, you know, noticeable, especially during those times of congestion, we're like meaningfully marching performance forward."

To achieve these improvements, Anza is focusing on several key areas:

1. Aggressively attacking all memory allocations: Watt explains, "Moving data around is the really expensive part. And I know like Kevin Bowers is always being the drum on this that like there's too much magical thinking about how computers work. But moving data is really important."

2. Improvements to turbine: The current bottleneck in doubling throughput and block capacity is getting block data out to the rest of the network. Watt mentions, "Without going too far into the weeds, it's not even so much like a Nick bandwidth problem as it is more of an IOPS problem with the way that agave has implemented this."

3. Eliminating inefficiencies in the validator: Watt describes this as "not doing dumb stuff in the validator." This includes optimizing hash functions, reducing unnecessary serialization and deserialization, and other small improvements that add up to significant performance gains when applied across the entire system.

The Role of Hardware in Solana's Performance

While hardware improvements can certainly help, Watt emphasizes that software optimization is the primary focus for improving Solana's performance. He explains, "Throwing, you know, bigger hardware at it is like a very expensive way of solving this problem. And so, you know, I can tell you about the mindset at ANZA. Like when we look at these performance improvements, like some people try and throw it in some special category of like feature work or its own sort of vertical in terms of engineering work, we like think of these as bugs."

This approach of treating performance bottlenecks as bugs that need to be squashed is more cost-effective and faster than relying solely on hardware upgrades. Additionally, Watt points out that improving the network's performance requires raising the entire minimum hardware specs across all validators, not just upgrading a single node.

The Modular Solana Architecture

Both Watt and Bruder discuss the trend towards a more modular architecture for Solana. Bruder highlights several examples of this modularization:

1. The Jito relay, which modularizes ingress and deduplicates signature-verified traffic.
2. Double Zero, which uses FPGAs and dedicated hardware for transaction processing and low-latency information propagation across the cluster.
3. Multiple schedulers running concurrently, including the priority graph scheduler and the more recent greedy scheduler.

Watt emphasizes that this modular approach doesn't contradict Solana's core thesis: "There's like this lazy argument that I see thrown around all the time on X that like, oh, you know, the Solana thesis will never work or the architecture is broken because the whole world can't run on one computer. And I'm like, well, yeah, I mean, that makes sense, but it already isn't."

He distinguishes between sharding state, which Solana avoids, and sharding compute, which makes sense in a microservice-oriented architecture. While this approach can make system design more challenging, Watt argues that it's worth it to deliver a superior user experience.

Addressing MEV in Solana

Miner Extractable Value (MEV) has been a significant challenge for Solana, as it has been for many blockchain networks. Watt outlines Anza's approach to addressing MEV:

1. Reducing latency and shrinking spreads through the IBRL (Increase Bandwidth, Reduce Latency) initiative.
2. Exposing simple primitives in the validator protocol to help upper layers restrict bad MEV.
3. Implementing multiple concurrent leaders to reduce the monopoly on block space that currently exists in each 1.6-second slot.

Watt emphasizes that it's not the L1's job to introspect transactions or decide what's fair, but rather to "expose the tools for the upper layers to deliver the best customer experience user experience possible."

Bruder adds that increased transparency is crucial in addressing MEV: "Internally, like something we're pushing a lot towards in 2025 is more transparency. So how does our system work? What is it doing? What code is it running? What are the transactions going through it? What are the outcomes of that?"

The Future of Solana Core Development

As Solana continues to evolve, the collaboration between different entities like Anza, Jito Labs, and Firedancer is shaping the future of the network. The focus on increasing block space, reducing latency, and improving overall performance promises to enhance the user experience and solidify Solana's position as a leading blockchain platform.

Watt concludes with a call to action: "IBRL, that's that's the mission statement. Everything kind of falls under that umbrella. You know, you can follow along. We're putting out more on the blogs trying to, you know, to Lucas's point, sunlight's the best disinfectant and, you know, kind of give you an idea of where we're heading with things and shine some light on some of the cool projects that we're working on."

As Solana continues to push the boundaries of blockchain performance and scalability, the ecosystem's commitment to continuous improvement and innovation suggests an exciting future ahead for developers, users, and stakeholders alike.

Facts + Figures

• Solana targets a 400 millisecond slot or block time, though real-world factors can affect this target.
• Skip rate on Solana has dropped from around 10% historically to now being around half a percent on a bad day, often lower.
• Anza aims to double real throughput for complex transactions like swaps in 2025.
• Solana's development process has evolved from being primarily driven by Solana Labs to involving multiple entities like Anza, Jito Labs, and Firedancer.
• The introduction of stake-weighted Quality of Service (QoS) in April 2023 was an emergency measure to address network congestion issues.
• Anza envisions a future with four evenly staked clients for Solana, improving network resilience.
• Current efforts focus on increasing block space, summarized by the IBRL (Increase Bandwidth, Reduce Latency) initiative.
• Jito Labs started in 2021 and focuses on making the Solana network more efficient, particularly for validators and stakers.
• The V2.1 rollout of Agave, Solana's client, brought significant performance improvements to the network.
• Solana's architecture is moving towards a more modular approach, with examples like the Jito relay and Double Zero implementations.
• MEV (Miner Extractable Value) remains a challenge for Solana, with efforts to address it focusing on reducing latency and implementing multiple concurrent leaders.
• Anza treats performance bottlenecks as bugs that need to be squashed, rather than as separate feature work.
• The Solana ecosystem is pushing for more transparency in 2025, particularly in how systems work and process transactions.
• Current bottlenecks in Solana's performance are more related to IOPS (Input/Output Operations Per Second) than network bandwidth.
• Solana's development culture emphasizes continuous iteration and improvement, summarized by the phrase "the job's not done."

Questions Answered

What is the current state of Solana's core development?

Solana's core development has evolved from being primarily driven by Solana Labs to involving multiple entities like Anza, Jito Labs, and Firedancer. This decentralization has introduced more coordination challenges but has also led to better decision-making and a more robust ecosystem. The development culture still emphasizes rapid iteration and shipping improvements quickly, with a focus on enhancing user experience and network performance.

How has Solana's performance improved recently?

Solana has made significant performance improvements over the past 12-18 months. The V2.1 rollout of Agave, Solana's client, brought noticeable macro changes to the network, including shorter slot times and lower latencies. The skip rate, which measures how often blocks fail to be produced, has dropped from around 10% historically to now being around half a percent on a bad day, often lower. These improvements have led to a snappier user experience and better validator economics.

What is Anza's roadmap for Solana in 2025?

Anza's primary focus for 2025 is on increasing block space, summarized by the IBRL (Increase Bandwidth, Reduce Latency) initiative. They aim to double real throughput for complex transactions like swaps. This will be achieved by aggressively optimizing memory allocations, improving the turbine protocol for block propagation, and eliminating inefficiencies in the validator software. The goal is to lower median fees and make it easier for transactions to be included, especially during peak demand periods.

How is Solana addressing the issue of MEV (Miner Extractable Value)?

Solana is addressing MEV through several approaches. The primary strategy is to reduce latency and shrink spreads through the IBRL initiative, which helps mitigate some of the information asymmetry that enables MEV. Additionally, Anza is working on exposing simple primitives in the validator protocol to help upper layers restrict bad MEV. A longer-term plan involves implementing multiple concurrent leaders to reduce the monopoly on block space that currently exists in each 1.6-second slot. The ecosystem is also pushing for more transparency in how transactions are processed and MEV is extracted.

What is the significance of client diversity in Solana?

Client diversity is crucial for improving Solana's resilience and performance. The introduction of Firedancer as a second client implementation is a step towards the ideal scenario where multiple evenly staked clients could ensure the network continues to function even if one client experiences a catastrophic failure. Client diversity helps uncover issues that are hard to discover with a homogeneous client base and allows for more confident refactoring and improvement of the codebase. It also provides a form of checks and balances, as different teams can challenge and improve upon each other's work.

How is Solana's architecture evolving?

Solana's architecture is moving towards a more modular approach. This includes implementations like the Jito relay, which modularizes ingress and deduplicates signature-verified traffic, and Double Zero, which uses FPGAs and dedicated hardware for transaction processing. The network is also experimenting with multiple concurrent schedulers. This modular approach allows for sharding of compute resources while maintaining a unified state, which is seen as a way to improve scalability and performance without compromising on Solana's core design principles.

What role does hardware play in Solana's performance improvements?

While hardware improvements can help, Anza emphasizes that software optimization is the primary focus for improving Solana's performance. They treat performance bottlenecks as bugs that need to be squashed, rather than relying on hardware upgrades. This approach is seen as more cost-effective and faster. However, to truly improve the network's performance, the entire minimum hardware specs across all validators need to be raised, not just upgrading individual nodes.

How is the Solana ecosystem addressing transparency?

The Solana ecosystem, particularly entities like Jito Labs, is pushing for more transparency in 2025. This includes providing more information about how systems work, what code they're running, what transactions are going through them, and what the outcomes are. Increased transparency is seen as crucial for addressing issues like MEV and for building trust in the ecosystem. Tools and dashboards that provide insights into network operations and transaction flows are being developed to support this goal.

What is the IBRL initiative in Solana?

IBRL stands for Increase Bandwidth, Reduce Latency. It's the core mission statement driving Solana's development efforts, particularly at Anza. The initiative focuses on increasing block space, which translates to pumping up Compute Units (CUs) or the total transactions that can be packed into a block. The goal is to lower median fees, make it easier for transactions to be included during peak demand, and improve overall network performance. This initiative encompasses various technical improvements, from memory allocation optimizations to enhancing block propagation protocols.