The Proof of History idea
Anatoly Yakovenko, then an engineer with a long Qualcomm career behind him, drafts a whitepaper describing a cryptographic clock that lets computers agree on the order of events without constantly messaging each other.
Ticks show sections you have finished.
Solana is the network. SOL is the coin that runs on it. This is the complete plain-English guide to both — how they work, who is behind them, what they are used for and what could go wrong.
An illustration, not a live feed: one tick per 200 ms, Solana’s target slot time since 9 October 2026 (it was 400 ms before). Ethereum’s slot is 12 seconds.
Loading live data from CoinGecko…
Six ideas. If you read nothing else on this page, read these.
Solana is a shared ledger run by hundreds of independent computers called validators. Nobody owns it, and anyone can use it or build on it.
SOL pays the network’s fees, is staked to secure it, and is the main currency inside Solana apps. Solana is the network; SOL is the asset.
A new block every 200 milliseconds since October 2026, and fees that are usually a fraction of a cent. It does this on one chain rather than splitting activity across many.
A cryptographic clock (Proof of History) orders events, and the runtime executes transactions that do not touch the same data side by side.
Send dollar stablecoins, trade tokens, lend and borrow, buy tokenised shares, run wireless and mapping networks — and, famously, launch memecoins.
Validators need powerful hardware, the network has a history of outages, there is no supply cap, and much of its activity is speculative.
Mainnet Beta’s first block
Since 9 October 2026; 400 ms before
Early October 2026; no maximum
Active stake, 12 October 2026
Sections tick themselves off as you reach the end of them. Progress is saved in this browser only.
How a wireless-systems engineer’s late-night idea became one of the most used blockchains.
In 2017 the main complaint about blockchains was that they were slow. Bitcoin handled a handful of transactions a second, Ethereum a few more. Most proposed fixes involved splitting a chain into pieces. Anatoly Yakovenko, who had spent more than a decade at Qualcomm working on operating systems and wireless protocols, thought the real bottleneck was simpler: computers on a blockchain have no shared clock, so they spend much of their time messaging each other to agree on what happened when.
Mobile networks solved a similar problem decades ago by giving every phone and tower a synchronised sense of time. Yakovenko’s idea, drafted in November 2017, was to build a clock out of cryptography itself: a hash function run over and over, each output feeding the next, producing a record that proves time has passed. He called it Proof of History.
He recruited former Qualcomm colleagues. Greg Fitzgerald wrote the first working version in Rust in early 2018; Stephen Akridge showed that signature checking could be offloaded to graphics cards. Raj Gokal, an investor and health-tech founder, joined as co-founder to run the business side. The project was first called Loom, then renamed after Solana Beach, the surf town north of San Diego where the engineers had lived.
The company, Solana Labs, raised about $20 million in private sales led by Multicoin Capital, and Mainnet Beta produced its first block on 16 March 2020 — the same week global markets crashed over the pandemic. A public sale on CoinList raised a modest $1.76 million.
In 2021 Solana became the main alternative to an expensive, congested Ethereum. NFTs and trading apps flooded in, SOL rose from under $2 to about $260, and Sam Bankman-Fried’s FTX and Alameda Research became its most visible champions. That association nearly sank it. When FTX collapsed in November 2022, SOL lost about 40% in a day and ended the year near $10 — down more than 95% from the peak — with a reputation for outages on top.
What followed is the part of the story supporters point to most. Developers stayed. The network’s reliability improved. Fees stayed low while Ethereum’s remained high, and in 2024 a wave of memecoin trading — whatever one thinks of it — proved the chain could carry enormous volume. By late 2025, SOL was held in US exchange-traded funds, and by 2026 companies such as Western Union and Morgan Stanley were building products on or around it.
The events that shaped Solana, including the ones it would rather forget. Filter by type.
35 events shown
Anatoly Yakovenko, then an engineer with a long Qualcomm career behind him, drafts a whitepaper describing a cryptographic clock that lets computers agree on the order of events without constantly messaging each other.
Greg Fitzgerald builds the first working implementation in Rust and the first public testnet follows. The project is briefly called Loom, then renamed Solana — after Solana Beach, California — to avoid confusion with another crypto project.
Solana Labs raises about $20 million in private token sales announced in mid-2019, with Multicoin Capital leading.
The first block is produced on 16 March 2020. A public auction on CoinList the same month raises about $1.76 million. The “Beta” label has never been formally dropped.
Inflation starts at 8% a year and is scheduled to fall by 15% each year until it reaches 1.5%.
Solana Labs sells $314 million of SOL to funds led by Andreessen Horowitz and Polychain Capital.
A flood of bot transactions during a token launch overwhelms validators, the network forks and stops producing blocks. Validators coordinate a restart.
SOL finishes a year in which it rose by more than a hundredfold, with a market value around $74 billion at the high.
An attacker mints about 120,000 wrapped ether on the Wormhole bridge without backing. Jump Crypto replaces the funds. The bug was in the bridge, not in Solana itself.
Several outages and long slowdowns hit the network, including halts of roughly seven hours (1 May), four and a half hours (1 June) and six hours (1 October), caused by bot floods and software bugs.
Roughly 9,200 wallets are emptied after a third-party wallet app, Slope, mishandled users’ secret phrases.
Sam Bankman-Fried’s FTX and Alameda Research — among Solana’s biggest backers and holders — go bankrupt. SOL falls about 40% in a day and ends the year near $10. The Serum exchange, controlled by FTX, is abandoned and forked by the community as OpenBook.
A fault in block propagation after a software upgrade stalls the network until validators restart it.
Solana Mobile ships its first Android phone, the Saga. In the same month the Helium wireless network migrates to Solana.
In lawsuits against Coinbase and Binance, the US Securities and Exchange Commission lists SOL among tokens it considers unregistered securities. Robinhood delists it. The Solana Foundation disputes the label.
Visa extends its stablecoin settlement pilot to Solana with payment processors Worldpay and Nuvei.
Pump.fun launches, making it possible to create a tradable token in a minute for almost nothing. Millions of tokens follow. Anza, a new core-development company spun out of Solana Labs, is also founded this month.
A bug in the program loader stops block production for about five hours. As of October 2026 the network has not had a full outage since.
PayPal’s dollar stablecoin launches on Solana, citing speed and cost.
Jump Crypto’s hybrid validator client — Firedancer’s networking with Agave’s execution — begins producing mainnet blocks.
A memecoin launched by Donald Trump days before his inauguration sends activity to records. SOL sets its all-time high of roughly $294.
A proposal to replace fixed inflation with a market-based rate wins 61.4% of votes — short of the 66.67% needed. On 17 March CME lists SOL futures.
The SEC ends its cases against Coinbase and Binance, removing the claim that SOL is a security from active litigation.
REX-Osprey lists SSK, a US fund holding and staking SOL. Canada had listed spot SOL funds with staking in April.
Solana Mobile’s second phone, the cheaper Seeker, begins shipping to pre-order customers.
Validators vote 98.27% in favour of SIMD-326, a complete replacement of Solana’s consensus. Just over half of all stake takes part.
Bitwise’s BSOL launches first, taking in $69.5 million on day one; Grayscale’s GSOL converts a day later and others follow.
Jump’s fully independent validator client goes live — the first time Solana has two complete, separately written clients in production.
Solana Mobile launches SKR, the token of its phone ecosystem, with a fixed supply of 10 billion and an airdrop to Seeker users.
Attackers socially engineer the signers of Drift’s security council, list a fake token as collateral and drain the exchange in about twelve minutes. Analysts at TRM Labs point to North Korean hackers. Solana itself was not breached.
Western Union launches a dollar stablecoin on Solana, issued by Anchorage Digital Bank, to settle transfers with its agents — first in the Philippines and Bolivia.
The Morgan Stanley Solana Trust begins trading on NYSE Arca with a 0.14% fee and staking. A day later the block compute limit rises to 100 million units.
A routing failure at hosting provider TeraSwitch disconnects about 90 validators for roughly 33 minutes. Finality would have stopped at 33.3%; it did not. The episode renews worries about how many validators share the same data centres.
In four steps — 350 ms on 21 August, 300 ms on 28 August, 250 ms on 18 September and 200 ms on 9 October — Solana halves the time between blocks (SIMD-525).
Alpenglow replaces Tower BFT on testnet and devnet on 24–25 September. Developers deny a rumoured 28 September mainnet switch. As of 12 October 2026 no mainnet date has been set.
No maths required. Three hands-on demos show the ideas that make Solana different: a clock, parallel lanes, and the new voting system on the way.
Every blockchain has to answer the same three questions: in what order did things happen, who gets to add the next block, and when is a block final. Solana’s answers are unusual, and they all serve one goal — keeping a single chain fast enough that it does not need to be split up.
The validator whose turn it is (the leader) runs a hash function in a tight loop, feeding each result back in as the next input. Because each hash depends on the one before, the only way to get to hash number one million is to compute all the hashes before it. The sequence is therefore proof that time has passed. When a transaction arrives, the leader mixes it into the sequence, stamping it with a position that nobody can later dispute.
This is not how the network reaches agreement — that is the job of voting, below. It is a shared timeline that lets validators process a stream of transactions without first stopping to negotiate about order.
Each tick hashes the previous line. Drop in a transaction and it becomes part of the chain. Then verify. (Toy 64-bit hash for display; Solana uses SHA-256.)
On Ethereum, transactions run one after another. On Solana, every transaction must declare up front which accounts it will read and write. That one rule lets the runtime, called Sealevel, see which transactions cannot possibly interfere with each other and run them at the same time on different processor cores. Two people paying two different shops never wait for each other. A thousand people fighting over the same new token do.
Leader schedule. Leaders are chosen in advance in proportion to their stake, and each gets four consecutive slots. Because everyone knows who is next, wallets and apps send transactions straight to upcoming leaders (an approach called Gulf Stream) rather than into a public waiting room.
Turbine. The leader chops each block into small pieces and passes them to a few validators, who pass them to a few more, like a phone tree. Extra recovery pieces mean a block can be rebuilt even if some packets are lost.
Tower BFT. Validators vote on blocks, weighted by stake. Each vote commits a validator more firmly to that version of history. A block with votes from two thirds of stake is “confirmed” in about a second; it is treated as fully final once enough later blocks are stacked on top — about 12.8 seconds.
Alpenglow, approved by validators in September 2025, throws out both Tower BFT and Proof of History’s role in consensus. Its voting protocol, Votor, finalises a block in one round if 80% of stake votes for it immediately, or in two rounds if at least 60% does. Votes are sent directly between validators and combined into one small certificate instead of filling blocks as individual transactions. The design target is finality in 100–150 milliseconds. A second part, Rotor, will later replace Turbine with a simpler single-hop relay.
. . Alpenglow is designed to stay safe with up to 20% of stake acting maliciously while a further 20% is offline; classic designs tolerate just under a third of bad actors in total.
Every transaction pays a base fee of 5,000 lamports per signature (0.000005 SOL). Half is burned and half goes to the block producer. On top of that you can add a priority fee, priced per unit of computation, which goes entirely to the block producer.
Because transactions declare their accounts, congestion is priced per account rather than chain-wide. If thousands of people are bidding to trade one hot token, they outbid each other for access to that token’s accounts; someone sending a stablecoin elsewhere on the chain still pays the minimum.
The Solana Virtual Machine is the engine that runs programs. Programs are usually written in Rust and compiled to a compact bytecode. Unlike Ethereum contracts, Solana programs hold no data of their own: code and data live in separate accounts, which is what makes parallel execution possible. Because the SVM is open source, other networks and rollups have adopted it as their execution layer.
A “client” is the software a validator runs. For years almost every Solana validator ran the same codebase, so a single bug could stop — and several times did stop — the whole network. Agave is that original Rust client, now maintained by Anza. Jito-Solana is a variant of Agave with a marketplace for transaction ordering, and has long been run by most of the stake. Frankendancer combined Jump Crypto’s new networking code with Agave’s execution and has produced mainnet blocks since September 2024. Firedancer is Jump’s fully independent client, written in C, live on mainnet since December 2025.
By mid-2026, estimates of the Firedancer family’s share of stake ranged from about 14% to more than a quarter, depending on the source and date. Jump has said it will end Frankendancer support when Alpenglow activates, so operators must move to full Firedancer or back to Agave.
Two capacity changes landed in 2026. On 29 July the amount of computation allowed per block rose to 100 million compute units. Then SIMD-525 cut slot time from 400 ms to 200 ms in four cautious steps (21 August, 28 August, 18 September, 9 October), each behind its own switch so the network could pause if too many blocks were being skipped. Per-block limits were scaled down in line with slot time, so the main gain from the slot change is latency — the wait before your transaction lands — rather than raw capacity.
Side effects: an epoch now lasts about one day instead of two, a transaction’s blockhash expires after roughly 30 seconds instead of 60, and the chain produces twice as many blocks for indexers to store.
Stake. Validators and their delegators have SOL locked behind every vote, and a validator that performs badly earns less and loses delegations. Automatic slashing — destroying the stake of a validator that provably double-votes — has been discussed and prototyped for years but, as far as we could verify, is not yet enforced automatically on mainnet. That is a real difference from Ethereum and is worth knowing.
SOL has no supply cap. What it has is a published schedule, a burn, and an unusually high share of coins staked.
Created at launch in 2020
6 October 2026; total supply is higher because some SOL is locked
Falling 15% a year towards 1.5%
About 439 million SOL of circulating supply
Each epoch the protocol creates new SOL and pays it to validators and the people who delegate to them. The annual rate started at 8% when rewards were switched on in February 2021 and shrinks by 15% of itself each year until it reaches a permanent 1.5%. On the original timetable that floor arrives in the early 2030s.
Because most SOL is staked, inflation works less like a cost to everyone and more like a transfer: people who stake roughly keep their share of the network, and people who do not are slowly diluted.
Half of every base fee is burned. Priority fees used to be half-burned too; since early 2025 (SIMD-96) they go entirely to the block producer. At current activity the burn offsets only a small part of new issuance, so supply keeps growing.
Roughly speaking, a little over a third went to private-sale investors across several rounds, about 13% to the team, about 10% to the Solana Foundation, close to 39% to a community reserve managed by the Foundation, and under 2% was sold in the 2020 public auction. Those early allocations have long since unlocked, but the insider-heavy start remains one of the standing criticisms of SOL.
Illustration of the published formula (8% × 0.85 per year, floor 1.5%). The real schedule is counted in slots rather than calendar days, so the live rate can differ slightly from this curve.
An estimate from your own assumptions, not a forecast. Rates are not fixed, rewards are paid in SOL, and the dollar value of SOL can fall far more than staking earns. The last box is the important one: staking mostly protects your share from dilution rather than creating new wealth.
Solana has no chief executive. It does have a founder whose opinions carry enormous weight, and a set of companies that write its software.
Yakovenko taught himself to program as a teenager and spent the core of his career at Qualcomm in San Diego, working on operating systems and the low-level software that keeps mobile networks in step. That background matters: where most early blockchain designers came from cryptography or finance, he approached a blockchain as a telecoms engineer would — as a problem of clocks, bandwidth and hardware.
By his own account the idea for Proof of History arrived late one night in 2017 after too much coffee. He wrote it up that November, persuaded former colleagues to join, and has been the project’s chief architect and most recognisable voice since.
Three things characterise how he leads. First, a stubborn commitment to a single fast chain: he has consistently argued against splitting Solana into shards or pushing users onto layer-2 networks, betting instead that hardware keeps getting faster. Second, a habit of arguing in public — with Ethereum researchers, with critics, and with his own community — in short, blunt posts. Third, a willingness to let others build the core: Solana Labs handed protocol development to Anza in 2024 and he has welcomed Jump’s rival client rather than resisting it.
He holds no formal power over the network. Changes need validator votes, and he has lost some: he supported cutting inflation, and SIMD-228 failed. But his endorsement moves opinion, as with his backing of SIMD-550 in 2026. When rumours spread in September 2026 that Alpenglow was about to go live, his reply was a single word — “decel”.
Worth knowing: he was named in a 2022 US class action alleging that SOL was sold as an unregistered security; plaintiffs later voluntarily dropped their claims against him and the Solana Foundation, while the case continued against other defendants.
A Wharton economics graduate who worked in venture capital at General Catalyst, founded the medical-device start-up Sano and led product at Omada Health before joining Yakovenko. He ran fundraising, operations and partnerships, and sits on the Solana Foundation board.
A former Qualcomm colleague who wrote the first implementation of Yakovenko’s whitepaper in Rust in early 2018 and published the first official paper with him. He set much of the engineering culture.
Another Qualcomm veteran. He demonstrated that signature verification could be moved to graphics processors, an early proof that the design could reach very high throughput.
Built the original software and raised the money. Since Anza’s creation it concentrates on products and tools. Its subsidiary Solana Mobile makes the Saga and Seeker phones.
Funds development and events, makes grants, and delegates stake to validators — a programme it has been shrinking, removing three subsidised validators for every one it adds. Its president is Lily Liu.
Formed in January 2024 by engineers and executives from Solana Labs. Maintains the Agave client and designed Alpenglow, whose lead authors include Roger Wattenhofer, a distributed-systems professor at ETH Zurich.
Builds Firedancer, the independent validator client, under chief scientist Kevin Bowers. It also covered the 2022 Wormhole bridge loss.
Maintains the Jito-Solana client that most stake runs on, operates the block-space auction that distributes MEV tips, and created JitoSOL, the largest liquid staking token.
Runs RPC services used by many apps and is a major validator. Its engineers wrote the SIMD-411 and SIMD-550 inflation proposals, and it stakes the SOL held by Bitwise’s ETF.
Led the 2018–2019 funding and became Solana’s most vocal institutional advocate. Later a lead investor in the Forward Industries treasury company.
Led the $314 million token purchase in June 2021 that funded the ecosystem’s expansion.
The FTX founder was Solana’s loudest supporter until his companies collapsed in November 2022. He was convicted of fraud in 2023. Solana’s recovery happened without him — and in spite of him.
From stablecoin payments to speculative tokens. Filter by category.
Stablecoins on Solana hit a record of about $17.3 billion on 25 September 2026 (DefiLlama’s stricter count: about $16.6 billion), third among blockchains behind Ethereum and Tron. USDC makes up roughly half.
Visa has settled with processors in USDC on Solana since 2023. PayPal brought PYUSD in 2024. In May 2026 Western Union launched USDPT, issued by Anchorage Digital Bank, to settle with its agents.
The trading front door for most users: an aggregator that routes a swap across many exchanges to find the best price, plus perpetual futures and lending.
The main decentralised exchanges where tokens are actually pooled and traded. Solana’s exchanges handled about $284 billion in the first quarter of 2026.
Lending and borrowing against SOL, staked SOL and stablecoins. Total value in Solana DeFi was about $5.5 billion in spring 2026, roughly half its August 2025 peak.
Deposit SOL, receive a token that earns staking rewards and can still be used elsewhere. Convenient, with added smart-contract risk.
Backed Finance’s xStocks put more than 60 US shares and ETFs on Solana from June 2025. Solana venues handled an estimated 95% or more of on-chain stock trading in mid-2026.
Real-world assets on Solana passed $4 billion in August 2026, nearly triple January’s figure, including about $1.2 billion of tokenised US Treasuries. Ethereum remains far larger at around $17 billion.
Launched in January 2024, pump.fun lets anyone create a token in moments. It has earned hundreds of millions of dollars in fees and faces a US class action calling it an unlicensed casino. The overwhelming majority of tokens launched there become worthless.
The TRUMP token of January 2025 brought record traffic and the all-time high in SOL. It also drew criticism, including from US lawmakers, about conflicts of interest. Treat this corner as gambling, not investing.
A community-built wireless network that moved to Solana in April 2023. People earn tokens for running hotspots that carry mobile data.
Render (GPU computing) moved its token to Solana in 2023. Hivemapper pays drivers for dashcam mapping. These networks need cheap, frequent micro-payments.
Solana Mobile’s second Android phone began shipping in August 2025 with a built-in hardware key vault and a crypto-friendly app store. Its SKR token launched in January 2026.
Solana was the second home of the 2021–2022 NFT boom. Compressed NFTs, introduced in 2023, cut the cost of issuing millions of items to almost nothing, useful for tickets and loyalty schemes.
A leading perpetual-futures exchange, and a cautionary tale: it lost about $285 million in April 2026 after attackers tricked its security signers.
Since late 2025 you can hold SOL exposure in an ordinary brokerage account. Here is what exists, and what it does and does not give you.
Early October 2026
9 October 2026, SolanaFloor tracker
Bitwise Solana Staking ETF
Forward Industries, 21 Sept 2026
The path took about two years. In June 2023 the SEC was calling SOL an unregistered security in court. By 2025 those cases had been dropped, CME had listed SOL futures (17 March 2025), and Canada had approved spot SOL funds that stake (April 2025). In July 2025 REX-Osprey listed SSK, the first US fund to hold and stake SOL, using a different legal structure from a standard spot ETF.
The SEC then adopted generic listing standards for crypto funds, and on 28 October 2025 Bitwise’s BSOL began trading, taking in $69.5 million on its first day. Grayscale converted its existing trust to GSOL the next day, and VanEck (VSOL), Fidelity (FSOL), 21Shares, Franklin Templeton and others followed. On 28 July 2026 Morgan Stanley listed its own MSOL trust on NYSE Arca with a 0.14% fee.
Unlike the first Bitcoin and Ether ETFs, most SOL funds stake some or all of their coins and pass rewards to shareholders, after fees. BSOL stakes its holdings through Helius; Morgan Stanley says it keeps none of the rewards for itself. This makes SOL funds behave a little like a share that pays a dividend in more of itself.
About $1.7 billion is real money but small next to Bitcoin ETFs, which hold well over a hundred billion dollars. One fund, BSOL, has taken roughly three quarters of all inflows; several others have seen long stretches with almost no movement. A 14-week run of weekly inflows ended in early October 2026.
A second channel is listed companies that hold SOL on their balance sheet, modelled on Strategy’s Bitcoin approach. Forward Industries — backed by Galaxy, Jump Crypto and Multicoin — is the largest, with about 8.16 million SOL in September 2026, roughly 1.4% of circulating supply. DeFi Development Corp held about 2.49 million. Upexi and others run similar strategies. These shares can trade far above or below the value of the SOL they hold, and they add company risk on top of coin risk.
Solana has no single official roadmap — upgrades are proposed, argued over and voted on one at a time. This is the state of play on 12 October 2026.
Slot time halved in four steps under SIMD-525. Transactions land sooner; epochs last about a day.
More computation per block, after earlier rises from 48 to 50 and then 60 million units.
A second full client in production. Next milestone: enough stake on it that no single codebase can halt the chain.
Finality in about 100–150 ms and no more on-chain vote transactions. Live on testnet and devnet since late September 2026; mainnet window described as October 2026 but unconfirmed.
The second half of Alpenglow: a simpler, faster way to spread blocks, replacing Turbine. A separate proposal that follows Votor.
SIMD-550 (faster decline) and SIMD-547 (a burned flat fee) are both in discussion. Neither has been put to a final vote.
Alpenglow removes vote fees of roughly 1 SOL a day and replaces them with a smaller burned admission fee. Helius estimates break-even stake could fall from about 4,850 SOL to about 450.
Projects such as Jito’s BAM and proposals for multiple simultaneous block producers aim to make transaction ordering fairer and attract professional trading venues.
The August 2026 near miss showed how much stake sits with a few hosting providers. Spreading validators across more data centres is a stated priority without a simple fix.
Mostly certainty. Transactions already appear within a second, but exchanges, bridges and payment companies wait for full finality before treating money as received — roughly 13 seconds today. If Alpenglow works as designed, that wait shrinks to a fraction of a second. It also frees the block space now used by validator votes, and cuts validator running costs. None of this is guaranteed until it has run on mainnet under real load.
Replacing consensus on a live network carrying tens of billions of dollars is the riskiest kind of upgrade. After the August 2026 routing incident, developers were explicit that they would not rush. One Anza researcher’s summary of the mood: “No Alpenrush.”
Five networks competing to be the place where applications run. Figures are approximate and move over time; each has trade-offs.
| Solana | Ethereum | Sui | Aptos | BNB Chain | |
|---|---|---|---|---|---|
| Launched | March 2020 | July 2015 | May 2023 | October 2022 | 2020 (as Binance Smart Chain) |
| Design approach | One fast chain; parallel execution | Conservative base layer plus many layer-2 networks | One chain; object-based parallel execution | One chain; parallel execution | Ethereum-compatible chain with a small validator set |
| Smart-contract language | Rust (SVM) | Solidity (EVM) | Move | Move | Solidity (EVM) |
| Block time | ≈ 0.2 s | 12 s | Well under 1 s | Well under 1 s | Under 1 s |
| Time to finality | ≈ 13 s today; ≈ 0.15 s targeted with Alpenglow | ≈ 13–15 minutes | Under 1 s | Under 1 s | A few seconds |
| Typical fee | Fraction of a cent | Cents to dollars on the base layer; cents or less on layer 2s | Fraction of a cent | Fraction of a cent | Cents or less |
| Validators | Several hundred | Roughly a million validator keys run by thousands of operators | A little over 100 | A little over 100 | A few dozen active |
| Supply policy | No cap; inflation declining to 1.5% | No cap; issuance partly offset by fee burn | 10 billion cap | No hard cap | Shrinking through burns |
| Notable outages | Several, 2021–2024 | No base-layer halts | Some short halts | Some short halts | Rare halts, including after a 2022 bridge hack |
| US spot ETF | Yes, since October 2025 | Yes, since July 2024 | Check current status | Check current status | Check current status |
| Main strength | Speed, low fees, large user base | Security, decentralisation, deepest DeFi | Fast finality, modern language | Fast finality, modern language | Exchange distribution, low cost |
| Main criticism | Hardware demands, past outages, memecoin dependence | Cost and fragmentation across layer 2s | Younger ecosystem, concentrated supply | Younger ecosystem, lower activity | Centralisation around Binance |
Scroll sideways on small screens. Sister guides: What’s Sui? covers Sui in the same depth.
It is a disagreement about where to put the work. Ethereum keeps its base layer slow and cheap to verify so that almost anyone can run a node, and moves everyday activity to layer-2 networks that inherit its security. Solana keeps everything on one layer and asks validators to run serious hardware. Ethereum’s approach scores higher on decentralisation and has never halted; Solana’s gives a simpler experience — one network, one set of apps, one pool of liquidity. Both are credible bets and many developers build on both.
Sui and Aptos descend from Meta’s abandoned Diem project and use the Move language, which was designed to make asset-handling code safer. Both already offer sub-second finality — the thing Solana is trying to add with Alpenglow. Solana’s advantage is scale: far more users, stablecoins, trading volume and developers. Whether better technology or a bigger network matters more is the open question.
A wallet stores the secret key that controls your coins. Filter by what you need. Listing here is not an endorsement — always download from the official site or app store.
10 options shown
The most widely used Solana wallet. Clean interface, built-in swaps and staking, transaction previews that warn about known scams, and support for several other chains.
One of the oldest Solana wallets, with especially detailed staking controls and strong hardware-wallet support.
A wallet from the team behind the Anchor developer framework, linked to a regulated exchange of the same name.
From the Jupiter trading platform; built around fast swaps and tight integration with its exchange tools.
Keeps your key on an offline device; you approve each transaction on its screen. Works with Phantom, Solflare and others. A sensible step once amounts matter.
Open-source hardware wallets; recent models support SOL, Solana tokens and staking.
Signs by scanning QR codes, so it never connects by cable or Bluetooth. Pairs with Solflare and Phantom.
The default wallet on Solana Mobile’s Seeker, keeping keys in a secure hardware area of the phone and approving with a fingerprint.
Requires several people to approve a transaction. The standard for teams, treasuries and anyone who wants more than one key between them and a mistake.
Leaving SOL on Coinbase, Kraken, Binance and the like is simplest, but the exchange holds the keys. Fine for small amounts and trading; FTX is the reminder of what can go wrong.
A Solana address is a string of 32 to 44 letters and numbers with no “0x” at the front. Tokens are held in separate token accounts attached to your wallet, which is why receiving a new token for the first time can cost a tiny refundable deposit. You cannot send SOL to an Ethereum address or the reverse — use a bridge or an exchange, and send a small test amount first.
A plain step-by-step. This is practical information, not a recommendation to buy.
Delegating directly, as above. Simplest and closest to the protocol. Your SOL is locked until you deactivate.
Swap SOL for a token such as JitoSOL or mSOL that grows in value as rewards accrue and can be sold at any time. Adds smart-contract risk, and the token can briefly trade below its backing.
ETFs such as BSOL and exchange “earn” products stake for you and take a cut. Convenient; you rely on the provider and do not choose the validator.
The protocol does not currently take stake away automatically for validator misbehaviour, so the main risks are different: a poor validator earns you less; a liquid staking contract could be exploited; and above all, the price of SOL can fall. Staking rewards of a few percent do not offset a large drop in price.
Fake wallet apps and browser extensions; search adverts leading to copycat sites; “airdrop” tokens that appear in your wallet with a link to claim more; people in chat groups offering help; and anything promising guaranteed returns. If an offer needs your recovery phrase, it is theft.
An honest guide has to include the case against. Each card gives the criticism first and the counter-argument second.
Solana fully stopped several times between 2021 and February 2024, for up to 19 hours. It has not halted since, but in August 2026 a single hosting provider’s failure took 28.8% of stake offline — within five points of stopping finality.
In fairness: More than two and a half years without a full halt, a second independent client, and staged upgrades with kill-switches are real progress. The concentration in data centres is not solved.
Validators need high-end servers and fast connections. The number of voting validators has fallen from over 2,500 in early 2023 to several hundred, and about 18 control a third of stake.
In fairness: Much of the fall was the Foundation ending subsidies for low-quality validators. Alpenglow should cut running costs. But a smaller, more professional set is still a more concentrated one.
SOL inflates every year, and attempts to reduce issuance have so far failed or stalled. Holders who do not stake are diluted.
In fairness: The rate declines automatically and most holders do stake. New proposals are live. But “scarce” is not an honest description of SOL.
A large share of fees and activity has come from memecoin trading, which is cyclical and leaves most participants poorer. When it fades, revenue falls sharply.
In fairness: Stablecoin payments, tokenised stocks and institutional products are growing and are less faddish. They are not yet the majority of activity.
Wormhole (2022), Mango Markets (2022), Slope wallets (2022) and Drift ($285 million, April 2026) were failures of apps and their operators, not the base chain — but users lost money all the same.
In fairness: The core protocol has not been hacked. For a user the distinction is thin: the risk sits in whichever app holds your funds.
Fast blocks and scheduled leaders make it profitable to reorder trades. “Sandwich” attacks have cost ordinary traders significant sums.
In fairness: Validators caught sandwiching have been removed from delegation programmes, and new ordering designs are in development. The problem is reduced, not gone.
Alpenglow replaces the heart of the system. Any flaw could mean a halt or worse, and the removal of Frankendancer forces operators to switch software at the same time.
In fairness: It has formal proofs, a year of testing, an overwhelming validator vote and a visibly cautious rollout. Risk remains until it has run on mainnet for a long time.
The legal status of SOL in the US rests on dropped lawsuits and fund approvals rather than a court ruling or statute. Early investors and insiders received a large share of the original supply.
In fairness: Regulatory acceptance has moved a long way since 2023, and initial allocations are years old. Policy can change with governments.
SOL fell more than 95% from its 2021 high to its 2022 low, and in October 2026 trades well below its January 2025 record near $294.
In fairness: Volatility cuts both ways, but nobody should hold SOL with money they need.
Ten things people commonly get wrong, in both directions.
“Solana goes down all the time.”
It did, between 2021 and early 2024. The last full outage was on 6 February 2024. There was a serious near miss in August 2026.
“Proof of History is Solana’s consensus mechanism.”
It is a clock. Agreement comes from proof-of-stake voting (Tower BFT now, Alpenglow’s Votor in future).
“Solana Labs controls the network.”
Protocol changes need validator votes, and core development is spread across Anza, Jump and others. Influence is concentrated, but control is not held by one company.
“SOL has a fixed supply like Bitcoin.”
There is no cap. New SOL is issued every epoch; only the rate declines.
“Transactions are free.”
They cost 5,000 lamports per signature plus any priority fee — tiny, but not zero, and higher in a rush.
“Solana died with FTX.”
It lost more than 95% of its value and many projects. It also recovered to set a new high in 2025 and gained ETFs.
“Solana is only memecoins.”
Memecoins are a large part of its trading. So are $17 billion of stablecoins, tokenised shares and payment pilots from Visa, PayPal and Western Union.
“Alpenglow is live, so finality is instant.”
Not on mainnet as of 12 October 2026. Finality is still about 12.8 seconds.
“Staking SOL is risk-free income.”
Rewards are mostly new issuance, are paid in a volatile asset, and depend on your validator.
“Owning a SOL ETF is the same as owning SOL.”
A fund gives price exposure. You cannot send, spend or use those coins on the network.
Every term used on this page, in plain English. Type to search.
36 terms
12 questions, each with an explanation. Your best score is saved on this device and counts towards the Graduate badge.
Everything asked here is covered on this page. Nine or more correct earns the final badge.
Short answers to the questions people ask most.
Solana is a public blockchain: a shared database run by hundreds of independent computers rather than one company. It is designed to process a very large number of transactions quickly and for a fraction of a cent each. People use it to send dollars as stablecoins, trade tokens, run apps and, increasingly, to issue traditional assets such as shares and treasury funds.
SOL pays transaction fees, is staked to help secure the network and earn rewards, covers the small refundable deposit (rent) needed to store data, and is widely used as collateral and trading currency in Solana apps.
Anatoly Yakovenko wrote the original Proof of History whitepaper in late 2017 and co-founded Solana Labs with Raj Gokal. Greg Fitzgerald and Stephen Akridge, former Qualcomm colleagues of Yakovenko, were among the founding engineers.
On the base layer, yes by design. Solana now targets a block every 200 milliseconds against Ethereum’s 12 seconds, and fees are usually well under a cent. Ethereum takes a different route, keeping its base layer conservative and pushing most activity to layer-2 networks. Solana’s speed comes with heavier hardware demands on validators.
There is no cap. The network started with 500 million SOL. In early October 2026 about 588 million were circulating. New SOL is issued as staking rewards at a rate that began at 8% a year and declines 15% a year until it reaches 1.5%.
Yes. It suffered several full outages between 2021 and early 2024, the longest lasting 17 to 19 hours. The most recent full halt was on 6 February 2024. In August 2026 a data-centre routing failure took 28.8% of stake offline for about half an hour without stopping the chain.
Alpenglow is a new consensus protocol intended to finalise blocks in about 100–150 milliseconds instead of roughly 12.8 seconds. Validators approved it in September 2025 and it runs on Solana’s test networks. As of 12 October 2026 it is not on mainnet and no date has been confirmed.
In the United States, spot SOL exchange-traded products have traded since 28 October 2025, and several stake their holdings. Nine such products held about $1.73 billion in early October 2026. An ETF gives price exposure through a brokerage account; you do not hold SOL yourself and cannot use it on the network.
You delegate SOL from your wallet to a validator. Your coins stay in an account you control and earn a share of newly issued SOL plus some fee income, minus the validator’s commission. Stake activates and deactivates at epoch boundaries — an epoch now lasts about a day. Rewards are not guaranteed and SOL’s price can fall by more than you earn.
This site does not give investment advice. SOL is a volatile asset that fell more than 90% in 2022 and has also risen sharply at other times. Read the risks section, understand what you are buying, and never put in money you cannot afford to lose.
More than a company database, less than Bitcoin or Ethereum by most measures. There are several hundred voting validators, and about 18 of the largest hold a third of the stake between them. Running a validator needs expensive hardware, and many sit in the same handful of hosting providers. Client diversity has improved since Firedancer launched.
There is no single best. Phantom and Solflare are the most widely used apps; Backpack and Jupiter’s wallet are popular alternatives; a Ledger or Trezor hardware wallet adds an offline key for larger amounts. Whatever you choose, the recovery phrase is the wallet — never type it into a website or share it.
The references used to write and check this guide in October 2026. Where sources disagreed we give a range or say so.
This guide is part of The Coin Galaxy, a network of independent explainers. These are the ones that bear most directly on Solana.