Solana ICM Roadmap: A Imitation Show for Hyperliquid?

Recently, an eye-catching event occurred in the Solana ecosystem. Key participants, including the Solana Foundation, Anza, and Jito Labs, jointly released a technical roadmap named "( Internet Capital Markets, ICM )". The core idea of this roadmap is "( Application Controlled Execution, ACE )", aiming to enable on-chain applications to gain millisecond-level autonomous transaction ordering rights, creating a decentralized "on-chain Wall Street".

Although Hyperliquid is not directly mentioned in the roadmap, a careful analysis reveals that many of its designs intentionally mimic the advantageous characteristics of Hyperliquid. It's as if Solana is implying: "What Hyperliquid can do, we can do too, and we will do it better!"

It is worth noting that Hyperliquid has dominated the on-chain perpetual contract market, with its trading volume reaching about 65% of the entire decentralized perpetual market at one point. Facing such strong competitors, Solana is clearly unwilling to be surpassed by latecomers, so it has launched this ICM roadmap.

So, what is this "imitation show" all about? Can Solana really catch up to or even surpass Hyperliquid? Let's delve into this topic.

Background and Content of ICM

key players driving change

First, let's take a look at the creators of this roadmap. The participants are all heavyweight players in the Solana ecosystem:

• Solana Foundation/Labs: The "parent" of Solana, responsible for overall coordination and core protocol development.

• Anza: A development company founded by former members of Solana Labs, similar to Ethereum's ConsenSys. In this roadmap, it undertakes many core technological breakthroughs, such as the new consensus protocol Alpenglow.

• Jito Labs: A major MEV infrastructure provider on Solana, with significant influence, almost controlling the "power of life and death" over all MEV traffic on Solana. This time, it leads the provision of Block Assembly Marketplace (BAM) and other transaction ordering solutions.

• Multicoin Capital: A well-known crypto investment institution and an early supporter of Solana. Due to holding a large amount of SOL and equity in ecological projects, it also has considerable influence in the technical direction.

• DoubleZero: A team focused on accelerating network communication, providing dedicated fiber optic network solutions to enhance communication speed between Solana validator nodes.

• Drift: A leading perpetual contract DEX project on Solana. Previously, Drift used an off-chain matching model, which seemed somewhat challenging when facing the fully on-chain Hyperliquid. This time, by participating in the roadmap formulation, it is evident that they hope to turn things around by leveraging underlying upgrades.

core issue

The roadmap focuses on improving the microstructure of the market, mainly addressing the issue of the current on-chain trading mechanisms being unfriendly to market makers. Currently, the Takers who actively initiate trades dominate, while the Market Makers ( are at a disadvantage. This is because Takers often have access to the latest information and can actively increase trading fees to ensure their trades are executed first, while Makers often do not have time to cancel their orders and are forced to trade at unfavorable prices.

Some high-frequency arbitrageurs will exploit this asymmetry to launch "toxic flow" attacks. For example, when the on-chain price has not yet updated, but the off-chain price has changed, arbitrageurs can fill market maker orders at the old price, causing market makers to bear the losses. The result is that makers, in order to protect themselves, either widen the bid-ask spread or reduce the order volume, leading to a decrease in overall market liquidity.

The ICM roadmap aims to balance this pattern and attract high-quality liquidity back to the chain.

) ICM's three-step strategy

Solana divides this grand plan into three phases:

Short term ###1-3 months (: Mainly optimize the existing on-chain trading experience, improve the usability of order book applications, and reduce malicious MEV interference. Specifically including:

• Jito Labs' Block Assembly Marketplace)BAM( module has launched on the mainnet. This module provides a temporary external system that allows smart contracts on Solana to have autonomous transaction ordering rights before the ultimate ACE)Application Controlled Execution( goes live.

• The Anza team optimized the success rate of "trading entering the same Slot", thereby reducing slippage and MEV losses.

These improvements are expected to be implemented gradually between July and September 2025.

Medium term )3-9 months (: Introduce dedicated high-speed networks and a new consensus, significantly reduce latency and improve throughput:

• Deploy a dedicated fiber optic network for DoubleZero, providing validators with near-zero jitter and reducing latency by up to 100ms for high-speed communication.

• Launching the Alpenglow consensus protocol, reducing the final confirmation time from approximately 12.8 seconds to about 0.15 seconds.

• Research and Development of Asynchronous Program Execution ) Asynchronous Program Execution, APE (, to reduce transaction execution blocking on consensus.

Long-term )9-30 months (: A revolutionary upgrade to the core architecture of Solana, aimed for realization around 2027:

• Multiple Concurrent Leaders, MCL): Allows multiple validators to propose transactions simultaneously within their respective pipelines, and then merges and sorts these concurrent blocks by priority fees. This weakens the monopoly of a single packager and enhances resistance to censorship.

• Native applications can control execution ( Application Controlled Execution, ACE ) function: Truly empowers on-chain smart contracts with the authority to control the order of transaction execution.

Analyzing this roadmap, it is not difficult to see the story behind it: the established DEX Drift on Solana has been surpassed by the newcomer Hyperliquid with its excellent experience as "on-chain Binance." Drift itself finds it hard to compete and has to seek help from "big players" like Solana Labs, Anza, and Jito. These "big players" proposed an ICM technology transformation plan, claiming to replicate all of Hyperliquid's core advantages for Drift, helping it to re-enter the DEX market. However, they also admit that this technological transformation is extremely difficult, so they have divided the plan into a three-step strategy. In the short term, the only thing that can be provided to Drift is Jito's BAM, allowing Drift to compete with Hyperliquid.

Next, we will analyze in detail which core advantages ICM has imitated and replicated from Hyperliquid.

Imitation 1: Trading Sorting Mechanism

The problem is: As mentioned earlier, the current chain favors takers, and makers suffer from "toxic flow." Users who actively take orders can initiate transactions on the on-chain orders based on the latest off-chain prices and prioritize their trades by increasing transaction fees, while market makers often cannot update or cancel orders in time. As a result, market makers either widen the spread or simply withdraw liquidity, leading to a deterioration in market depth.

( The ultimate solution of ICM: Applied controllable execution ) ACE ###

The ICM roadmap proposes the concept of ACE( Application Controlled Execution ), which decentralizes the transaction ordering rights to various on-chain applications, allowing the applications to determine how to sort and execute transactions related to them. For example, in the future implementation of ACE on Solana, DeFi contracts can achieve the following custom transaction sorting rules:

• Oracle price update insertion: DeFi applications can insert a transaction to retrieve the latest price from the oracle before executing large trades, ensuring that orders are matched at the latest reasonable price and preventing market makers' quotes from being arbitraged based on outdated prices.

• Order Cancellation Priority Execution: The application can set "order cancellation requests" to take precedence over new "taker trades" execution, allowing makers the opportunity to withdraw their orders in a timely manner when market conditions are unfavorable.

• Tail bidding: For example, after a large buy order pushes the price up, the DeFi application will auction the "follow-up" opportunity, where whoever is willing to return the most benefit to the protocol ( or user ) will have their transaction executed right after the large order. The DeFi application can return the auction proceeds to users, thus converting toxic MEV flow into positive income.

( JITO's BAM: Transition Plan

Before the official launch of ACE, Jito Labs introduced a transitional solution called Block Assembly Marketplace )BAM###. The workflow of BAM is:

1. The user sends the transaction to a node running BAM software ( instead of directly to the current Leader ).

2. BAM nodes collect local transactions and run various plugins ( plugin ) to reorder transaction packages ( Bundle ) under privacy protection. The plugin runs in a secure TEE environment, hiding the transaction content externally before execution (. Through the plugin, application developers can customize various sorting rules for their contracts, such as prioritizing order cancellations, updating oracle prices before matching, and even running complex in-app bidding.

3. The sorted transaction bundle is then sent to the Solana Leader for packaging and on-chain.

BAM can be seen as a testing ground before ACE goes on-chain, functionally very close to the ultimate ACE, but it operates in an independent network off-chain, rather than being built into the Solana main chain protocol.

It is worth noting that Jito has previously provided infrastructure aimed at MEV extraction, such as the Jito Block Engine ), with a business model that creates opportunities for arbitrageurs by optimizing transaction ordering and sharing profits. This is somewhat like a "spear" standing in opposition to ordinary users and those being arbitraged. However, Jito closed the public memory pool ( mempool ) feature aimed at arbitrage bots in early 2024 to reduce negative externalities like sandwich attacks. This move indicates that the Solana community tends to suppress harmful MEV to maintain fairness for users.

The launch of BAM is in line with this idea: it essentially transforms the ordering mechanism originally used for MEV arbitrage into a "shield" to protect liquidity providers such as market makers. This includes prioritizing forced order cancellations to avoid losses for market makers and introducing bidding rebates to reduce front-running profits. Original MEV seekers must change roles if they want to make money; they need to write BAM plugins to serve DeFi protocols and earn profits from plugin fees.

( Learning from HYPERLIQUID

The above ACE/BAM approach can actually be seen as a pursuit of the on-chain matching mechanism of Hyperliquid. Hyperliquid is a dedicated chain )Appchain(, inherently designed to serve DEX. Moreover, the HLP Vault operated by Hyperliquid is actually one of the largest market makers on the platform, so it is not difficult to understand that the chain rules of Hyperliquid are more inclined towards liquidity providers, having already implemented many designs on the chain layer to protect market makers, such as:

• Order cancellation priority protection: Cancel orders and maker-only orders are prioritized, preventing market makers from being adversely executed without their knowledge. The "cancellation priority execution" mentioned by Solana ACE has been practiced by Hyperliquid for many years.

• Latest Price Guarantee: Hyperliquid's liquidation and matching process emphasizes the "double check" using the latest feed price and margin status. For example, when an order is matched and executed, the system will pull the latest oracle price again to assess both parties' margins, ensuring that risks do not arise due to price lag. This is similar to ACE inserting oracle updates before trade execution.

• Self-trade protection: If a buy and sell order matches from the same address, Hyperliquid will automatically cancel instead of matching, to prevent wash trading or unnecessary fees.

Solana ICM's ACE/BAM is undoubtedly "learning from" Hyperliquid. As a leading on-chain CLOB, Hyperliquid has implemented various mechanisms favorable to market makers using a dedicated chain. Solana now hopes to replicate this effect with a universal chain and modular plugins------that is, to allow each application to have control over trade ordering similar to that of Hyperliquid.

Mimicry II: Instant Finality

) Current Consensus Comparison

Solana currently uses Tower BFT, where confirmation and finality are probabilistic and gradual: a block is considered "confirmed ### Confirmed (" once it receives 2/3 of the votes, but it requires the accumulation of about 32 subsequent blocks ), usually taking about 13 seconds ### to be anchored as "finalized ( Finalized )". For certain applications (, such as high-frequency trading ), a final confirmation time of over ten seconds is still too long.

HyperBFT is a consensus algorithm developed by Hyperliquid, inspired by the HotStuff consensus, which adopts a two-round voting process to confirm blocks, achieving "instant finality".

• Round 1: Prevote ( Prevote ): After the validators receive the candidate block broadcasted by the Proposer, they will conduct a quick verification. If the verification passes, each validator will cast a "prevote" ( Prevote ) vote for this block and broadcast it to the entire network. This vote represents: "I have initially reviewed this block, and it is fine."

• Round Two: Precommit ( Precommit ): Once a validator has collected Prevotes from more than two-thirds of the validators for the same candidate block, it gains sufficient confidence that the majority of the network members endorse this block. Therefore, the validator will cast a more significant "precommit" ( Precommit ) vote and broadcast it. This vote represents:"