Sui vs Solana Speed: Which Blockchain Faster?

Last Updated: June 2025

This SUI vs Solana speed comparison separates theoretical benchmarks from real-world throughput, finality, and reliability. Sui has been benchmarked at a theoretical maximum of 297,000+ transactions per second (TPS), while Solana cites 65,000+ TPS under its own benchmark conditions. The figures are June 2025 references and are not directly comparable unless workload and measurement methods match. For broader context, see the complete Solana vs SUI comparison and the stage-matched analysis of whether SUI is the next Solana.

This comparison focuses on three dimensions of speed: throughput (TPS), transaction finality time, and sustained reliability under load. Ecosystem size and tokenomics are outside this article's scope, as is governance.

Sui vs Solana Speed: Quick Comparison

Note: Theoretical maximum TPS figures are measured under controlled laboratory conditions and do not represent typical mainnet performance. For current figures, consult the Sui mainnet explorer and Solana Explorer directly.

Performance figures reflect data available as of June 2025. Blockchain network performance changes as networks upgrade.

The figures above reflect theoretical maximums. Sustained real-world TPS for both chains is significantly lower, and the gap between theoretical and real-world performance is the most important distinction this comparison addresses. The full breakdown appears in the dedicated TPS analysis section below.

What Does "Blockchain Speed" Actually Mean?

Blockchain speed has three dimensions that must be evaluated together: throughput (how many transactions per second), finality time (how long before a transaction is irreversible), and reliability (whether that throughput holds under real load). Interpreting any benchmark figure without all three dimensions produces an incomplete picture.

Transactions Per Second (TPS): What the Numbers Actually Mean

Transactions per second (TPS) is the number of transactions a blockchain network can process and confirm in one second. Three distinct variants of this metric are routinely conflated in blockchain benchmarking, and conflating them produces misleading comparisons.

Theoretical Maximum TPS is measured under ideal laboratory conditions: optimal hardware, no real user load, controlled transaction types, and no network congestion. These are ceiling figures, not operating figures.

Peak Recorded Mainnet TPS is the highest throughput observed on the live network during a specific event (a stress test, a popular application launch, or a coordinated benchmark). This reflects the network under high but real conditions, though not necessarily sustained ones.

Sustained Real-World TPS is the average throughput maintained over extended periods under normal user activity. This is the figure that matters most for application developers planning transaction volumes.

These three numbers can differ by orders of magnitude for the same chain. The full comparison table for both Sui and Solana appears in the Theoretical TPS vs. Real-World Performance section below.

Transaction Finality Time: The Other Half of Speed

Transaction finality is the point at which a transaction becomes irreversible and cannot be reorganized out of the chain. This is distinct from a first confirmation, which signals only that a transaction has been seen, not that it cannot be reversed.

Two types of finality matter for this comparison. Probabilistic finality means a transaction becomes harder to reverse as more blocks are added on top of it, but it is never technically 100% irreversible in the near term. Deterministic finality (also called instant or single-slot finality) means the transaction is final the moment the consensus protocol confirms it.

For practical applications, finality time often matters more than raw TPS. A DeFi liquidation bot needs to know a transaction is irreversible before acting on the outcome. If finality takes 32 seconds (as it does on Ethereum), a liquidation opportunity measured in milliseconds has already passed. Both Sui and Solana target sub-second finality, but through different mechanisms covered in detail in the Sui vs Solana Finality Time section below.

Real-World vs. Theoretical TPS: Why the Gap Matters

Theoretical TPS and sustained real-world TPS diverge because laboratory benchmarks remove the friction that live networks cannot. On a live mainnet, network overhead from propagating transactions to validators, the coordination time for validators to agree on transaction ordering and validity, the complexity of real user transactions (which are rarely as simple as benchmark transactions), and congestion from competing traffic all reduce effective throughput significantly.

Both Sui's 297,000+ TPS and Solana's 65,000+ TPS are theoretical maximum figures. Real-world sustained performance for both chains is a fraction of those numbers.

This gap is the single largest accuracy failure in most competitor articles on this topic, which present theoretical and real-world figures as if they are equivalent. The blockchain trilemma, a widely cited framework associated with Ethereum co-founder Vitalik Buterin, offers useful context: blockchains generally face tradeoffs among security, scalability, and decentralization. Both Sui and Solana have made architectural choices that prioritize scalability, which partly explains their high theoretical throughput. Those choices also have implications for validator hardware requirements and the degree of decentralization each network achieves.

How Fast Is Sui? Architecture Behind Sui's Speed

Sui's speed architecture rests on two structural pillars: an object-centric data model that makes parallel transaction processing the default, and the Mysticeti consensus protocol that achieves sub-second finality. Mysten Labs, co-founded by Evan Cheng (former head of programming languages at Meta) and built by a team of former Meta/Diem blockchain engineers, launched Sui in May 2023 as a Layer-1 network designed from the ground up for high-throughput applications. Full technical details are available in Sui's official documentation.

Object-Centric Data Model and Parallel Execution

Sui organizes its blockchain state as a collection of discrete objects, each with explicit ownership encoded directly in the data structure. An object on Sui is a self-contained data unit: it knows who owns it, whether it is owned by a specific address, shared across multiple addresses, or immutable.

Because each transaction operates on specific named objects with known ownership, the Sui runtime can determine before execution which transactions are completely independent. If transaction A touches object X and transaction B touches object Y, and those objects share no state, the network can process both transactions in parallel without any coordination overhead. Independence is structural, not computed.

A concrete analogy helps here. Imagine a post office sorting letters by recipient. Because each letter is addressed to a specific person, multiple postal workers can sort different letters simultaneously without checking with each other. Sui's object model works the same way: each transaction is addressed to specific objects, and the network routes and processes them independently.

This contrasts directly with how Solana handles parallelism. Solana uses an account-based model, where blockchain state is stored in accounts associated with programs. Solana's Sealevel runtime must perform upfront conflict detection, analyzing each transaction's read/write set to determine which ones can safely run in parallel. Parallelism on Solana is detected; on Sui, it is the default.

Sui Move, a customized variant of the Move programming language (originally developed by Meta for the Diem blockchain project, and distinct from the Move variant used by Aptos), makes the object model work at the language level. Move's resource-oriented design enforces explicit ownership: every digital asset is a resource that cannot be copied or implicitly discarded. Because ownership is encoded in the data structure at the language level, the Sui runtime can determine transaction independence without additional analysis overhead. Move's ownership model is not just a safety feature; it is the mechanism that enables the object-centric parallel execution architecture.

Narwhal, Bullshark, and the Mysticeti Upgrade

Sui currently uses Mysticeti as its consensus protocol, a significant upgrade from its predecessor Bullshark that reduces the rounds required for finality from multiple steps to a single round. Most competitor articles on this topic still reference Bullshark as Sui's active consensus; that information is outdated.

Understanding Sui's consensus requires tracing the evolution across three components:

Narwhal is a DAG-based (Directed Acyclic Graph) mempool protocol that separates data availability from consensus. Rather than combining transaction dissemination and ordering in a single process, Narwhal handles transaction dissemination independently, allowing validators to receive and acknowledge transactions before the consensus round begins. This separation enables high-throughput transaction ingestion without blocking the consensus process.

Bullshark was the consensus protocol built on top of Narwhal that Sui used through 2023. Bullshark provided asynchronous Byzantine fault-tolerant safety, meaning it could reach consensus even under adverse network conditions. However, Bullshark required multiple rounds of voting to achieve finality.

Mysticeti is Sui's current consensus protocol, which replaced Bullshark in late 2023 and early 2024. According to the Mysten Labs Mysticeti research paper, Mysticeti achieves sub-second finality by reducing multi-round consensus to a single round without sacrificing Byzantine fault tolerance. The practical result is approximately 480ms finality time, even under high transaction load. Mysticeti also targeted reliability improvements alongside latency reductions, addressing concerns about network stability under peak demand.

The Mysticeti upgrade is the primary reason why Sui's speed profile in 2025 differs substantially from what earlier articles described. Any comparison that does not account for Mysticeti is working from outdated architectural data.

Sui TPS Benchmarks and Real-World Performance

Under controlled benchmark conditions, Sui has processed over 297,000 transactions per second, a figure that reflects optimal hardware and network configurations rather than typical mainnet load. Mysten Labs conducted this benchmark under conditions designed to demonstrate the theoretical ceiling of the object-centric parallel execution architecture.

On mainnet, peak recorded TPS has reached above 20,000 transactions per second during stress test events, according to data from the Sui mainnet explorer. Under normal operating conditions, sustained real-world TPS for Sui typically ranges from 2,000 to 10,000+ transactions per second depending on network activity levels.

The gap between 297,000 and 2,000–10,000 reflects the friction of real network conditions: validator distribution across geographic locations, the diversity of real transaction types (which are more complex than benchmark transactions), and the coordination overhead that exists even with Mysticeti's single-round consensus. As Sui's user base and application ecosystem grow, real-world sustained TPS will continue to be tested. A full comparison of theoretical versus real-world figures for both chains appears in the Theoretical TPS vs. Real-World Performance section.

How Fast Is Solana? Architecture Behind Solana's Speed

Solana's speed architecture combines three components: Proof of History for transaction ordering, Tower BFT for Byzantine Fault Tolerant consensus, and Sealevel for parallel smart contract execution. Solana Labs, founded in 2018 by Anatoly Yakovenko and Raj Gokal, launched the Solana network in March 2020 as one of the first Layer-1 blockchains built around the goal of achieving high throughput at the base layer. Technical architecture details are documented in Solana's official documentation.

Proof of History: Solana's Timekeeping Innovation

Proof of History (PoH) is Solana's transaction ordering mechanism, not its consensus mechanism. Invented by Solana co-founder Anatoly Yakovenko, PoH creates a verifiable cryptographic record of time passing using sequential SHA-256 hashing, where each hash output feeds into the next as the input. This produces an unforgeable historical sequence that proves a specific amount of computational time has elapsed between events.

The mechanism is a Verifiable Delay Function (VDF): a computation that takes a minimum amount of time to complete and produces a result that anyone can verify quickly. By continuously running this hash chain, a Solana leader validator generates a cryptographic timestamp sequence before the consensus round begins.

This is what enables Solana's speed. In most blockchains, validators must coordinate to agree on the order of transactions during the consensus round, which adds latency. PoH eliminates that coordination overhead by cryptographically pre-ordering transactions before consensus begins. Validators receive transactions that already carry a verifiable timestamp, so they can focus on agreeing on validity rather than order.

The common technical error in most competitor articles is describing PoH as Solana's consensus mechanism. It is not. Tower BFT (Solana's Byzantine Fault Tolerant consensus mechanism) handles actual consensus. PoH is the timekeeping infrastructure that makes Tower BFT fast. Solana uses Proof of Stake for validator selection, Proof of History for transaction ordering, and Tower BFT for consensus. The original Solana whitepaper by Anatoly Yakovenko describes the PoH mechanism in full technical detail.

Sealevel: Parallel Smart Contract Execution

Sealevel is Solana's custom parallel execution runtime, the component responsible for processing thousands of smart contracts simultaneously by identifying which transactions do not conflict with each other.

The mechanism works as follows: before executing any transaction, Sealevel analyzes each transaction's read/write set, meaning the specific accounts it will read from and write to. Transactions that access entirely different accounts have no shared state and can safely execute in parallel across multiple CPU cores and GPUs. Solana validators run high-specification hardware (fast CPUs, large RAM, NVMe SSDs, and high-bandwidth internet connections) specifically to take advantage of this parallel execution model.

Sealevel is the third leg of Solana's speed architecture, operating alongside PoH (transaction ordering) and Tower BFT (consensus). Without Sealevel, Solana's theoretical TPS would be constrained by sequential transaction execution; with it, the network can process many transactions simultaneously, which is why the theoretical maximum TPS is as high as it is.

Sealevel has a meaningful architectural limitation: transactions that touch the same accounts cannot be parallelized. When two transactions both try to access the same liquidity pool contract in an AMM protocol, they must run sequentially. These are called hot-account bottlenecks, and they become significant constraints during high-demand periods when many users interact with the same popular programs simultaneously. This was one of the contributing factors to the congestion events Solana experienced between 2021 and 2023, a topic addressed directly in the reliability section below.

The contrast with Sui is structural: Solana requires upfront conflict detection to find parallelizable transactions; Sui's object model makes most transactions parallel by default because independence is encoded in the data structure.

Solana TPS Benchmarks and Real-World Performance

Solana's theoretical maximum TPS is commonly cited at 65,000, though benchmark results under different hardware configurations and with simplified transaction types have reported higher figures. The 65,000 figure comes from Solana Labs' own analysis of the protocol's theoretical capacity, as documented in Solana's technical materials.

On mainnet, peak recorded TPS has reached into the tens of thousands during high-activity periods, according to data from Solana Explorer. However, understanding Solana's TPS requires an important distinction that most analyses miss: Solana's total on-chain TPS includes vote transactions, which are internal validator consensus messages, not user-facing application transactions. When analysts report high total TPS figures for Solana, those figures include a substantial proportion of vote transactions. User-facing application transaction TPS is a lower subset of total TPS.

Under normal operating conditions, Solana sustains approximately 2,000 to 5,000+ user-facing transactions per second. This is the figure that application developers should plan around, not the total TPS figure that includes validator vote transactions.

The gap between theoretical and sustained figures traces back to real network conditions: hot-account bottlenecks under peak demand, validator coordination time, and the resource requirements of maintaining a high-throughput network across a geographically distributed validator set. The full three-variant comparison appears in the Theoretical TPS vs. Real-World Performance section, and the reliability context that explains some of this gap follows in the Network Reliability and Congestion section.

Sui vs Solana Finality Time: Head-to-Head

Sui achieves transaction finality in approximately 480 milliseconds using its Mysticeti consensus protocol. Solana achieves finality in approximately 400 to 800 milliseconds under normal network conditions, with that window extending significantly under congestion.

These figures place both chains in a comparable range under normal conditions. The difference becomes meaningful under load. Sui's Mysticeti consensus was specifically designed to maintain sub-second finality even under high transaction volumes, because its single-round design does not accumulate latency as throughput increases. Solana's finality time, by contrast, can extend during periods of network congestion, because hot-account bottlenecks and mempool pressure lengthen the time between transaction submission and confirmed finality.

For DeFi applications where finality time determines profit and loss, this distinction matters. A liquidation bot that submits a transaction needs to know the position is closed before market conditions change further. Under normal conditions, both chains offer competitive finality. Under peak demand, Mysticeti's design gives Sui a more predictable finality window.

The verdict on finality: Sui's Mysticeti consensus delivers more consistent sub-second finality under load, while Solana's finality remains competitive under normal conditions but extends under congestion. For latency-sensitive applications, Sui's finality predictability is a structural advantage.

Theoretical TPS vs. Real-World Performance

Solana's 65,000 TPS figure and Sui's 297,000+ TPS figure are both theoretical maximums measured under controlled laboratory conditions. Neither reflects what an application will experience on mainnet under real user load. This section breaks down all three TPS variants for both chains with source attributions.

Methodology note: Theoretical maximum TPS figures are measured under controlled conditions and do not represent typical mainnet performance. Sustained real-world TPS figures are derived from mainnet explorer data and will vary with network activity.

Sui's 297,000+ TPS figure was measured in a controlled benchmark with optimized transaction types, ideal hardware, and no simulated user traffic competing for resources. The stress-test peak of 20,000+ reflects the network under coordinated high load, closer to real conditions but still not representative of average usage.

Solana's figures require the vote transaction clarification described in the previous section. The 65,000 theoretical maximum includes all transaction types. The 2,000 to 5,000+ user-facing sustained figure excludes vote transactions and reflects the TPS that application users actually generate. Among major Layer-1 blockchains, both Sui and Solana rank at the high end for sustained real-world throughput, though newer benchmarks from chains like Sui continue to push the ceiling.

Both chains perform closer to their sustained real-world figures during normal operation. The theoretical maximums are engineering targets and marketing benchmarks, not operational guarantees. Real-world performance is also substantially affected by network reliability, which the next section addresses directly.

Network Reliability and Congestion: The Speed You Can Count On

A blockchain's theoretical TPS ceiling means little if the network halts under peak demand. Solana's performance history between 2021 and 2023 makes this the most important reliability question any developer should ask before committing to either chain.

Between 2021 and 2023, Solana experienced more than a dozen significant network disruptions, ranging from severe congestion that caused failed transactions and degraded performance to complete network halts requiring validator-coordinated restarts lasting several hours. These were not minor incidents: some outages took the network offline for four to eighteen hours, affecting all applications built on the chain. The Solana network status history documents these events.

The root causes were architectural and operational. Bot-driven transaction floods overwhelmed the mempool, causing validators to fall out of consensus. The hot-account bottleneck in Sealevel's parallel execution model meant that popular programs like Serum DEX became sequential choke points when accessed simultaneously by thousands of transactions. Validator coordination under these conditions cascaded into full network halts.

Solana Labs responded with a series of targeted improvements. QUIC protocol adoption replaced the UDP-based transaction ingestion that bots could flood without cost; QUIC allows the network to limit and prioritize inbound traffic. Stake-weighted quality of service (QoS) ensures that validators with larger stakes receive priority transaction throughput, reducing the effectiveness of spam attacks. Local fee markets allow transaction fees to rise in response to demand for specific accounts, creating an economic signal that reduces bot flooding by making it costly.

The network has been substantially more stable in 2023 and 2024. No major multi-hour outages have been reported during this period, and the improvements appear to be addressing the root causes rather than symptoms. The trajectory is clearly toward greater stability.

Fee spikes during congestion (using "gas fees" as the colloquial umbrella term for transaction costs across chains) serve as a practical proxy for network stress. During Solana's congestion periods, priority fees spiked significantly as users competed to get transactions included. Sui's gas model, denominated in SUI token, is designed to be more predictable, though the network's shorter operational history means it has not been tested under the kinds of extreme demand that Solana experienced.

Sui launched in May 2023 and has less than two years of mainnet operational history. Mysticeti's design specifically targeted reliability improvements alongside latency reductions, and early mainnet data shows stable performance, but Sui has not been subjected to the sustained demand spikes that revealed Solana's architectural stress points. Sui's reliability record will become clearer as the network scales to higher transaction volumes.

The balanced assessment: Solana's outage history was real and significant, but the trajectory since 2023 is toward stability with specific architectural improvements in place. Sui's stability record is positive but short. Neither chain presents an unambiguous reliability winner.

Speed for Specific Use Cases: DeFi, NFTs, and Gaming

Speed requirements vary by application type. DeFi liquidation protocols care most about finality time; high-frequency trading bots prioritize throughput; NFT minting events need burst TPS capacity; blockchain games need sustained low-latency transaction processing. Here is how each chain maps to each use case.

DeFi applications present the clearest speed differentiation between the two chains. Finality time matters more than raw TPS for liquidation-sensitive protocols. A lending protocol's liquidation bot needs to confirm that a position is closed (fully final, not just submitted) before market conditions shift further. Under normal conditions, both chains offer competitive finality, but Sui's Mysticeti consensus delivers more consistent sub-500ms finality under load.

Arbitrage bots have different requirements: they need high transaction throughput and fast inclusion, not just fast finality. Solana's high sustained TPS and its large existing DeFi ecosystem (protocols like Orca, Raydium, and Jupiter) give it a meaningful advantage for arbitrage applications that depend on liquidity depth and established protocol integrations.

AMM protocols on Solana face the hot-account bottleneck: every trade accesses the same pool contract, creating sequential execution that limits effective throughput for that specific program during peak demand. Sui's object model means individual liquidity positions can potentially be modeled as distinct objects, reducing the bottleneck.

Verdict for DeFi: Sui's Mysticeti finality advantage is meaningful for liquidation-sensitive protocols being built from scratch. For projects requiring existing DeFi ecosystem integrations, liquidity depth, and established partner protocols, Solana's maturity currently outweighs Sui's finality edge.

NFT applications require burst TPS during minting events, when thousands of users attempt to mint simultaneously within seconds. Both chains have active NFT ecosystems: Solana has mature marketplaces including Magic Eden and Tensor, while Sui has Bluemove and native Mysten Labs marketplace infrastructure. Sui's object model provides native NFT support at the protocol level (each NFT is a distinct object), which offers composability advantages beyond raw speed. For new NFT projects prioritizing protocol-level elegance and lower-latency drops, Sui's architecture is a better fit. For projects targeting the largest existing collector base, Solana's established ecosystem provides a head start.

Blockchain gaming requires sustained high-frequency, low-value transactions: item trades, in-game actions, and leaderboard updates where TPS matters more than finality time. Sui's object ownership model maps naturally to game asset ownership, each in-game item can be a distinct object with a clear owner, and Sui Move's resource model prevents the duplication exploits that have affected asset economies in other blockchain games. Both chains are actively targeting gaming, but Sui's architectural alignment with game asset semantics gives it a structural advantage for developers building new gaming applications.

Speed Roadmap: What's Coming for Sui and Solana

Both Sui and Solana are actively developing speed improvements, and the current performance gap may shift materially within the next 12 to 24 months. Understanding the trajectory matters as much as the current snapshot.

Solana: Firedancer. The most significant speed development on the Solana roadmap is Firedancer, a complete reimplementation of the Solana validator client developed by Jump Crypto in C and C++. This is a validator client, not a protocol upgrade or new blockchain; Firedancer implements the existing Solana protocol with highly optimized, low-level code targeting maximum performance.

Jump Crypto demonstrated Firedancer processing over 1 million transactions per second in testnet conditions. This figure is from a controlled testnet demonstration and should not be interpreted as a mainnet performance guarantee. Mainnet performance under real conditions will be lower, though the optimization discipline behind Firedancer is genuine and significant.

Firedancer's importance extends beyond raw throughput. As a second independent validator client, it improves Solana's resilience: a bug in one client implementation does not halt the entire network if validators running the other client can continue. A partial implementation called Frankendancer (combining elements of both clients) was deployed to mainnet validators in 2024, with full Firedancer still in active development. When fully deployed, Firedancer is expected to increase Solana's mainnet throughput substantially, though the specific mainnet impact has not been benchmarked yet. The Firedancer validator client repository tracks development progress.

Sui: Post-Mysticeti Optimization. Sui's development team has indicated continued consensus optimization beyond the Mysticeti upgrade. The object-centric model theoretically has more throughput headroom than account-based models as hardware improves, because the parallelism is structural rather than computed. More concurrent transactions can be processed by adding capacity without architectural changes to how independence is determined.

Sui's shorter operational history (under two years at the time of writing) means less battle-testing than Solana's five-year track record. Speed improvements being developed on a younger network carry more uncertainty about real-world impact than they would for a more established one.

The trajectory: Solana is closing any finality gap with improvements to its mempool handling and Firedancer's eventual full deployment. Sui has structural parallelism headroom that may allow continued throughput scaling. Neither chain's current speed lead on any single dimension should be treated as permanent.

Which Is Faster: Sui or Solana?

As of 2025, Sui leads on theoretical maximum throughput (297,000+ vs. 65,000+ TPS) and delivers more consistent sub-second finality (approximately 480ms) through its Mysticeti consensus upgrade, while Solana maintains advantages in ecosystem depth, five years of mainnet battle-testing, and the potential step-change that Firedancer may deliver when fully deployed.

Breaking this down by dimension:

Throughput (TPS): Sui leads on theoretical maximum TPS by a wide margin. On sustained real-world TPS under normal conditions, both chains operate in a comparable range (2,000 to 10,000+ for Sui; 2,000 to 5,000+ user-facing for Solana). Neither chain has a decisive real-world throughput advantage at current adoption levels.

Finality Time: Sui leads on finality consistency. Mysticeti's single-round consensus delivers approximately 480ms finality under both normal and high-load conditions. Solana's finality is competitive at 400 to 800ms under normal conditions but extends under congestion. For latency-sensitive applications where finality predictability matters, Sui has the structural advantage.

Reliability: Solana has the longer operational track record and has demonstrably improved its congestion resistance with QUIC, stake-weighted QoS, and fee markets. Sui's shorter history means less stress-test data, even if the early record is positive. For projects where operational risk tolerance is low, Solana's established improvement trajectory provides more evidence.

Use-case trade-offs:

• DeFi liquidation-sensitive protocols: compare finality consistency and shared-object contention
• DeFi protocols requiring existing integrations: compare liquidity depth and partner availability
• NFT platforms: compare object-level architecture with the size of the existing collector market
• Blockchain gaming: test asset ownership, concurrency, wallet support, and developer tooling
• General high-throughput applications: benchmark the actual workload instead of relying on headline TPS

For current mainnet performance data, check the Sui mainnet explorer and Solana Explorer directly; figures change as both networks evolve and Firedancer's deployment progresses.

Frequently Asked Questions About SUI vs Solana Speed

Is Sui faster than Solana?

Sui leads on theoretical maximum TPS (297,000+ benchmarked vs. Solana's 65,000+) and delivers approximately 480ms transaction finality through its Mysticeti consensus protocol. Solana's finality is comparable (400–800ms) under normal conditions but extends under congestion. For raw theoretical throughput and finality consistency, Sui leads; for ecosystem maturity, Solana maintains an advantage.

What is Sui's TPS?

Sui's theoretical maximum TPS has been benchmarked at over 297,000 transactions per second under controlled conditions. On mainnet, sustained real-world TPS typically ranges from 2,000 to 10,000+ during normal network activity, significantly below the theoretical ceiling. For current figures, consult the Sui mainnet explorer at suiscan.xyz.

What is Solana's TPS?

Solana's theoretical maximum TPS is commonly cited at 65,000+ under ideal conditions. Total mainnet TPS is higher in some measurements but includes vote transactions (internal validator consensus messages). User-facing application TPS on mainnet is approximately 2,000 to 5,000+ transactions per second under normal load. For current data, see Solana Explorer.

What is Proof of History in Solana?

Proof of History (PoH) is Solana's transaction ordering mechanism, not its consensus mechanism. It operates by running a continuous SHA-256 hash chain where each output becomes the next input, creating a verifiable cryptographic timestamp sequence. Validators receive pre-ordered transactions, eliminating coordination overhead during consensus and enabling Tower BFT to focus on validity rather than sequence.

What consensus does Sui use?

Sui currently uses Mysticeti as its consensus protocol, which replaced Bullshark in late 2023 and early 2024. Mysticeti reduces consensus from multiple voting rounds to a single round, achieving approximately 480ms transaction finality. Narwhal handles transaction dissemination at the mempool layer. Articles referencing Bullshark as Sui's active consensus are working from outdated information.

What is Sui's transaction finality time?

Sui achieves transaction finality in approximately 480 milliseconds using its Mysticeti consensus protocol, according to the Mysten Labs Mysticeti research paper. This sub-second finality holds under high load conditions because Mysticeti's single-round design does not accumulate latency as throughput increases. Real-world figures may vary slightly depending on network conditions.

Does Solana still have network congestion problems?

Solana experienced more than a dozen significant network disruptions between 2021 and 2023, including complete halts lasting several hours. Since then, Solana Labs deployed QUIC for transaction ingestion, stake-weighted quality of service, and local fee markets. The network has been substantially more stable in 2023 and 2024, with no major multi-hour outages reported, though full stress-testing at higher adoption levels remains ahead.

Which is better for DeFi, Sui or Solana?

Sui's approximately 480ms Mysticeti finality may matter for liquidation-sensitive workloads, while Solana's longer operating history and established integrations may matter for applications that depend on existing liquidity. Teams should benchmark their own shared-state behavior, finality requirements, and integration dependencies before selecting a network.

This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency investments carry significant risk. Always conduct your own research before making investment decisions.

Related Reading

• SUI vs Solana: Layer-1 Blockchain Comparison
• Is SUI the Next Solana?