Exploring SVM Scalability Solutions With Demand-driven Modularity

Key Points TL;DR

• Solana Ecosystem is growing massively and showcasing strong metrics.
• With this growth, some challenges arose. To address this, Solana has in place compute usage management to solve for and support ecosystem growth.
• As the dApps ecosystem grows more diverse, needs vary, and demand for composable scalability rises.
• Innovative architects propose modular solutions to support growth - Layer-2s such as rollups/rollApps.
• While general-purpose rollups represent a viable solution for Ethereum scaling, the same model might not be applied to Solana due to different challenges faced by the base layer and dApps.
• App-specific rollups on Solana have the potential to offer customized solutions for dApps with specific demands.

Introduction:

The rapid growth of the Solana ecosystem has been nothing short of remarkable. With its high throughput and low transaction fees, Solana has attracted a plethora of decentralized applications (dApps) and a significant capital influx. As per Coingecko Research 2024, the Solana ecosystem emerged as the most popular blockchain ecosystem accounting for 49.3% of global crypto investor interest in chain-specific narratives (followed by Ethereum at second position with 12.7% of investor interest). The research examined interest in blockchain ecosystems based on non-botted global web traffic from January 1 to March 18, 2024.

This interest extends beyond simply blockchain investors. Solana's ability to process high-speed transactions per second (TPS) has established it as one of the most performant blockchains. It saw an unprecedented influx of active wallets in the first few months of 2024 (owing to the memecoin mania) and, on April 6, 2024, touched the highest daily average TPS among blockchains, clocking in at 1,053 TPS (46 times faster than Ethereum mainnet, and 5 times faster than Polygon). Over the first quarter of 2024, Solana's DAUs (active wallets) skyrocketed from 411,582 in January 2024 to 922,611 in March 2024 — a 224% increase. Syncing up to match, the Total Value Locked (TVL) of the Solana dApps soared 23.5% in March 2024 compared to the previous month. These figures highlight Solana's ability to attract substantial liquidity and user activity and showcase an expanding consumer base with heightened interest in the Solana environment.

Additionally, Solana has become a magnet for developers, reinforcing its status as a leading blockchain for innovation. From 2020 to 2023, the number of developers building on Solana has increased tenfold, a testament to its attractive ecosystem and developer-friendly environment. Developer retention (with a 3-month retention rate) rose from 31% to over 50% throughout 2023. This surge in developer interest fueled a thriving dApp ecosystem that spans DeFi, gaming, payments, and NFT marketplaces, driving further growth and adoption. Today Solana is home to 280+ dApps ranging from Web3 wallets like Phantom Wallet, NFT marketplaces like Magic Eden, infra dApps like Neon EVM, and many rising daily.

Solana's growth is rocketing to the moon, and we are bullish! 🚀

Image: On May 12, 2024, Solana generated the highest total economic value in a single day for the first time in history, surpassing other chains. Blockworks.

A huge win.

However, this exponential growth has also led to a suboptimal dApp and user experience as the network's computational bandwidth is being pushed to its limits.

Challenges like transaction congestion, block space competition, and failed transactions have often been discussed by the crypto community.

During times of congestion, more people want to access the limited resource—blockspace! One underlying question that took Crypto Twitter by storm (the first half of 2024) amid the memecoin madness on Solana was: Has the surge in Solana's TVL and transaction volumes translated beneficially for all dApps?

The obvious answer, of course, is yes. User activity and interest are at an all-time high (ATH). And admit it, we simply love ATHs 🐈‍ 🐕

However, dApps are grappling with several critical operational challenges that hinder their widespread adoption and efficiency and need a solution. Prominent projects such as DRiP, Jito Labs, Grass, Zeta Markets, and Neon EVM have highlighted issues related to network congestion, which may be compounded by increasing popularity and activity levels.

Some Shared Thoughts

The beautiful thing about building in blockchain is transparency in dialogue….and open discussions about concerns and wins with the community at large.

Some projects took to Crytpo Twitter to report their issues and solutions and discuss them with the community. Many are also actively working out their own solutions to counter their challenges. Some of these include:

1. Exploits

The rising activities attracted not just the good but also some bad—malicious actors who exploit the network's low fees to attack users. Bot activity also reached ATHs and to combat bot attacks during peak times, Jito Labs shut down its mempool service, aimed at managing transaction prioritization for profit (MEV).

Source: X, Jito Labs Official X

2. Financial losses

For DRiP specifically, which distributes millions of NFTs weekly using Solana as its infrastructure, these issues translated into substantial value leaks, estimated at around $20,000 per week due to rising SOL prices and network congestion.

Source: X, Vibhu Norby, CEO DRiP

3. Inability to customize With many project models getting innovative with new business logic and core architectural differences, especially in burgeoning fields like gaming and AI, modularity and flexibility define implications and potentials in the blockchain-building space.

Source: X, Grass Official X

The operational situation, therefore, underscores a crucial dilemma: while Solana offers high transaction throughput, but also as activity grows, significant barriers arise. Moreover, dApps cannot build custom operational solutions and must fit the general Solana framework preventing them from customizing parameters (such as instructions for specific use cases, setting limits on computation, etc.) per their business requirements. The ongoing debate within the Solana community about the necessity for Layer-2 solutions (L2s), rollups, and modular appchains reflects a broader industry need to balance centralized efficiency with decentralized principles, all while enhancing usability and mitigating operational costs for dApp developers.

What could really help these dApps?

• Dedicated blockspace: So that dApps enable their users to enjoy stable fees and performance, unaffected by congestion from other major happenings on the network (memecoins, airdrops, NFT mints, etc.)
• Scalability: high-speed and cost-effective yet secure transaction execution.
• Flexibility: And importantly, allowing founders and builders the flexibility to maximize their business logic by creating space for customization.

From a broad view, perhaps yes. Let’s explore.

This article will boldly explore the current ecosystem with diverse projects and their commercial needs, and long-term sustainability to illuminate some interesting answers and clarify the debate. It will touch on composable scalability to web3 (Layer-2s and rollups/rollApps) while steering clear of tech-heavy lingo. Here we will take a more market-savvy and commercial approach to dissecting scaling solutions.

Stay tuned to X and register for the power-packed ETHcc Brussels 2024 heavyweight technical panel discussion featuring key speakers (CTOs, founders, devs, and some super hot tech talent) with good insights into upcoming blockchain scaling technology for your dApps. Umm, we kid you not👇

Hosted by Neon EVM. Click here to register.

Background: Importance of Compute Units for dApps on Solana

Compute Units (CUs) on the Solana blockchain are a critical resource that directly impacts the performance, efficiency, and cost-effectiveness of dApps.

A compute unit is a measurement of how much computation is needed to process your instructions. A single Solana transaction can have multiple instructions. By default, each instruction within a transaction has a compute unit limit of 200,000. This can be changed (to a max of 1.4M), bearing in mind that a higher compute unit limit can decrease the chance that your transaction will be included on-chain since it takes more computation for a validator to process it. Also, the compute unit price is the price (in lamports) you pay to the validator per unit of computation.

Why do compute units matter?

Compute usage impacts the performance experience for the users of a dApp, as well as the cost of executing transactions. Each transaction consumes a portion of the total available compute units in a block, so efficient compute usage ensures that more transactions can be processed per block, enhancing overall throughput. This also reduces delays and the likelihood of dropped transactions, providing a smoother and more responsive user experience.

Additionally, efficient compute unit allocation maximizes the number of transactions that can fit within a block, preventing underutilization of block capacity and optimizing network resources. Today the total compute cap on a block on Solana L1 is 48M CU, and during times of congestion, this cap is often reached.

Source: Solana

Solana's current state, solution, and challenges

Solana has continuously made efforts to overcome these challenges at both software and hardware levels, including the introduction of priority fee functionality, improved validator updates, and soon-to-launch Firedancer and Frankendancer upgrades, etc. Some of the latest upgrades by Solana include:

• Priority Fees: Integrating dynamic priority fees on Solana can enhance the user experience by prioritizing transactions effectively. This feature is also being integrated into the command-line interface (CLI) for program deployment.
• Optimize Compute Unit (CU) Usage: By reducing the number of CUs used per transaction, more transactions can be processed within the same block capacity. With CU Budget Request Optimization, developers can specify a compute unit budget for transactions. Currently, requesting more CUs than necessary leads to inefficiencies as the transaction scheduler cannot accurately gauge remaining compute resources in a block. Optimizing these requests to better match actual transaction requirements can improve scheduling efficiency.
• Use Stake-Weighted Quality of Service (QoS): Infrastructure providers should adopt stake-weighted QoS, a feature allowing block builders to prioritize transactions from staked validators. This mechanism will ensure that high-value transactions are processed more efficiently.
• Transaction Scheduler Improvements in v1.18: The upcoming software release (v1.18) will introduce improvements to the transaction scheduler, a component that helps fill blocks efficiently. This new implementation will be tested alongside the existing scheduler and can be enabled or reverted by validator operators based on performance.

Add to these the anticipated Firedancer, a new validator client design underway for the Solana protocol, and the network is ready to scale with bandwidth and hardware. It proposes a complete rewrite of Solana's current validator client in the C programming language, aiming to refine its integrated approach vis-a-vis the current monolithic thesis.

All these vertical scaling paths by Solana will help handle the burgeoning compute load.

But will these be enough?

Eventually, a solution that goes beyond enhancements and upgrades may well be needed. Some dApps are looking for customized solutions that can optimize operations to benefit their needs…. and this would have to be a modular solution.

Layer-2s scaling solutions like rollups and rollApps are proposed architecture solutions from various stakeholders in this direction. They promise to enhance performance and enable a new era of blockchain functionality.

The next part of the article focuses on the latter. Read on to find out whether Solana rollups/rollApps are potential solutions that can benefit the business value of your dApp.

What are SVM Rollups, and are General-purpose Rollups Enough?

SVM rollups are a layer-2 (L2) scaling solution. They operate by bundling multiple transactions into a single batch, which is then processed off-chain before being settled on-chain. Rollups simply work by batching/rolling many small transactions into a single compressed transaction, which is then submitted to the roll-up smart contract on the Layer1 chain (for example, Solana). When this is done, the State for all the transactions is updated. This can be considered similar to how zip files work, where multiple files are combined into a single file to save space.

How different rollups perform computations and prove themselves largely determines their type. Without going into massive detail, here is an overview:

Optimistic Rollups: Rely on the assumption that transactions are valid by default, with a dispute resolution mechanism in place for incorrect transactions. However, they face challenges related to dispute resolution timeframes and complexity. Therefore, until someone challenges the execution, whatever is submitted by optimistic rollups is assumed to be correct, hence the name optimistic.

ZK-Rollups: These utilize zero-knowledge proofs to validate transactions while assuming all transactions are false until proven valid, offering greater security. Here, rollup nodes must provide validity proofs for all transactions to ensure the transaction data is valid.

With rollups, moving transaction computation and execution off-chain alleviates congestion on the L1 chain, they can process much higher volume than on the L1, resulting in faster and more cost-effective operations.

However, a general-purpose rollup does not fully solve the underlying issue—blockspace management. When a general-purpose rollup reaches the same activity as the L1, it is also subject to transaction bottlenecks.

Further, general-purpose rollup would essentially have its own set of problems:

• Fractured UX: From a macro viewpoint, thousands of Web3 users move assets between networks every day in order to be able to interact with various dApps on different blockchains through bridging, swapping, wrapping, and other such mechanisms. The result is complex blockchain UXs causing headaches for everyone and making it difficult to onboarding new users.
• Liquidity Fragmentation: An accessible consolidated market is important for any dApp, whether gaming or decentralized finance (DeFi) apps. The distribution of liquidity (or available trading volume) across multiple blockchains and networks results in fragmented pools of liquidity rather than a consolidated, easily accessible market. For DeFi, this often implies less efficient trades due to high slippage.
• Reduced social and community interaction: Not all your users are willing to jump chains and play Spider-Man all day long and then you need to make additional marketing efforts and expenses to indulge this.

(Many, therefore, associate rollups and L2s as fragmented horizontal scaling solutions)

Thus, general-purpose rollups have their unique advantages but do not necessarily solve the network congestion and compute management problems that dApps on Solana L1 are looking to solve.

This raises the question of customized or app-specific rollups, better called rollApps.

What are App-Specific Rollups or RollApps?

Think application-specific blockchains minus the consensus overhead, leaving just a highly scalable execution environment. RollApps leverage modular architecture and are application-specific chains.

If general-purpose rollups aim for permissionless scaling for the Solana ecosystem, then rollApps aim to scale business ideas and allow for the seamless deployment of app-specific rollups.

• High level of customization: Users pay fees to the rollApp in their own token or any token of choice, offering greater flexibility in economic models. dApps can tailor their environments to specific needs, enabling them to optimize performance and cost efficiency. For example, an order book DEX needs an execution speed of <100ms to compete with TradFi. RollApps can customize this parameter and allow them to unlock such speed. RollApps can allow dApps to customize parameters that favour their prime operational metrics, such as throughput and transaction confirmation time. For instance, larger block sizes and shorter block times can increase transaction throughput for a gaming dApp, while faster block times can reduce transaction confirmation time.
• Dedicated resources: Unlike general-purpose rollups, rollApps provide dApps with dedicated throughput and blockspace. This eliminates competition with other applications and ensures stable performance even during high network activity.
• Enhanced security: With the consensus layer being handled by the underlying Solana L1, rollApps inherit the robust security and decentralization of Solana, while offering customizable execution layers for dApps.

Are rollApps for everyone?

Suffice it to say, it is not one over the other; but rather identifying the ideal approach for a blockchain project. Enterprise chains or dApps that demand more control over how things work (security, customization, no heavy lifting of consensus-related responsibilities, etc.) would be better suited to explore their app-specific RollApps. In contrast, novice projects with limited resources, a niche customer base, and the ability to achieve their goals within Solana’s rules would be better suited to stay on Solana L1 and explore its upgrades like Firedancer and more.

What is Possible Now, and What is Coming in the Future?

While Solana's current vertical scaling solutions, such as priority fees and transaction scheduler improvements, address some immediate concerns, modular solutions like rollApps may support the future of dApp scalability and performance.

These offer customization and dedicated resources potentially unavailable in a monolithic architecture. As Solana continues to evolve, the adoption of rollApps will likely be an alternative approach to sustaining its growth and maintaining its competitive edge in the blockchain ecosystem.

Future Outlook

We are excited about the possibilities opening up for the infrastructure growth on Solana. Our team is dedicated to continually enhancing infrastructure and tooling to ensure that developers can experience the full potential of Web3 technology. Stay tuned for more insights in Neon EVM upcoming articles. The debate between vertical and modular scaling solutions is just beginning, and we will continue to explore the best paths forward.

Are you building?

If you are building an SVM L2 solution or RollApp and would like to connect with us, contact Ramiro Gamen, Partnership Manager, at [email protected] for an insightful conversation. Don’t miss a beat: follow Neon EVM latest updates on Discord and Twitter.