@__spekulator__ Yeah I'll post it below. Its one sample conducted in late 2025. Its findings are interesting in that though the execution of transactions may be by bots, they are directed by humans to do so.
https://t.co/ttkbudP1YO
Solana continues to dominate real usage.
β’ Live TPS: 3,900β4,200+ (β2,000 non-vote/user)
β’ Block time: ~365β375 ms (heading to 200 ms)
β’ Daily non-vote txs: regularly 100M+, recent ATHs 112β172M
β’ Daily active addresses: 2.0β2.8M
Fees stay ultra-cheap:
Median β $0.0004β0.0005 | Average β $0.002β0.003
Ecosystem strength:
DeFi TVL ~$5.4β8B
Stablecoins ~$16B
24h DEX volume ~$2.8B
Reliability: 30 months without a full outage.
~680β700 validators, 433M+ SOL staked, Nakamoto coefficient ~ 18
Upgrades (slot time cuts, higher compute, Firedancer, Alpenglow) keep pushing performance higher.
Solana isnβt just fast on paper β itβs processing more real user activity than almost anyone else, at fractions of a cent. This is what high-performance infrastructure looks like in production.
π₯ Top Crypto Chains Right Now (Aug 2026)
Daily Active Users
π₯ BNB Chain β 3.5β4.3M
π₯ Tron β 3.6β4.0M
π₯ Solana β 1.9β2.2M
Daily Transactions
π₯ Solana β 50β90M+
π₯ BNB Chain / Tron β 12β20M
π₯ Tron / Aptos / Sui β 10β15M+
BNB Chain & Tron dominate real users (USDT + retail).
Solana still owns pure speed & volume.
Mass adoption is already here β just not where most people are looking.
I see $SOL eventually dominating across all metrics. The winner takes most!
#Crypto #BNB #Tron #Solana #Onchain
@BrandonWealth There's a way for those of us who are disciplined and focused. It's a path few take bcz its means foregoing the now. I'm talking about investing in stocks and crypto. Despite the best efforts of boomers and bureaucrats we can prevail.
@fenibelal Definitely, it still isn't their major driver of revenue but it will only continue to grow. Hyperscalers will increase capex yoy and I expect Mvrl to capture a sizable share of it.
π΅ What I'm into π΅
Marvell Technology ( ticker: $MRVL ) is a semiconductor company that designs specialized computer chips.
These chips help move, store, process, and protect data quickly and reliablyβespecially inside large data centers, cloud systems, and AI infrastructure.
Think of them as the high-speed βplumbingβ and networking gear that lets servers, storage, and AI systems talk to each other efficiently.
Marvell focuses on things like custom chips for big tech companies.
Fast networking switches and connections.
Storage controllers
Optical and high-speed interconnects
It designs the chips but doesnβt manufacture them itself (a βfablessβ company).
The future is bright for companies like Marvell. The AI infrastructure build out will deliver great returns over the next few years.
BOT VS HUMAN TRANSACTIONS
π£ Solana π£
Solanaβs non-vote transactions are mostly bots.
70β90%+ of daily non-vote txs come from automated activity.
One major Dune analysis found ~87% of transactions by bots β while nearly 99% of wallets were human.
A tiny group of high-frequency addresses (thousands to 100k+ txs per day) drives the majority of volume. Addresses doing >100k txs/day alone have hit ~40%+ of total transactions in sampled windows. Low-frequency βhumanβ wallets make up 90%+ of unique addresses but only a few percent of actual txs.
DEX volume is even more extreme: 70β95% bot-driven in recent periods, with some reports near 95% by early/mid-2026.
Failed and reverted txs? Heavily bot-dominated β mostly failed arbitrage and MEV attempts.
Low fees + high speed = perfect playground for bots, market makers, and agents.
Humans still matter for unique users and real direction.
But the raw transaction count? Overwhelmingly machine.
βͺοΈ Ethereum βͺοΈ
Ethereum plays a different game.
Higher fees keep pure bot spam in check. Overall automated/bot share of total transactions usually lands in the 20β50% range β far below Solana.
MEV and sandwich bots are still active and extract real value, especially in DeFi. One well-known operator has dominated a large slice of sandwich attacks. Total MEV remains massive. But the cost of high-frequency probing limits the extreme transaction-count dominance seen on cheaper chains.
On some L2s the picture gets closer to Solana, with speculative MEV eating big shares of gas. Mainnet itself stays more balanced.
Key takeaway:
Higher fees = fewer pure spam and arbitrage bots flooding the count.
Human and intentional activity make up a larger relative share than on Solana.
π‘ BNB Chain π‘
BNB Chain runs hot on automation.
Precise βbot vs humanβ splits matching Solanaβs are rarer, but the pattern is clear. Low fees, fast blocks, and strong AI-agent infrastructure push heavy bot, MEV, and agent activity.
BNB has hosted a disproportionately large share of on-chain AI agents β at times ~60% of tracked deployments across major networks. Sandwich and sniper bots have long been active, especially on PancakeSwap and other DEXs.
Like other cheap, fast chains:
Most unique wallets are still human.
Most transactions and volume? Machines.
BNBβs combination of distribution + low cost keeps attracting both retail users and relentless automation.
π΅ Base π΅
Base looks a lot like Solana on bot share.
Dune-style frequency labels (addresses >500 txs/day = bot) regularly put high-frequency wallets at 70β80%+ of total transactions. Low-frequency βhumanβ addresses form the big majority of unique wallets but only ~15β20% of the actual transaction count in sampled days.
Power users fill the middle ground.
Cheap blockspace fuels continuous market-making, arbitrage, trading bots, and fast-growing agentic payments.
Stablecoin and trading flows show particularly strong automation. System-triggered and agent activity keep rising.
Humans matter for unique participation and consumer use cases.
The daily transaction numbers, though, are dominated by bots and automated systems.
Low fees + Coinbase distribution = high machine activity.
@__spekulator__ That's a great question and the answer is vague. Bots make up the majority roughly 86% of non vote transactions according to dune. But in the age of agentic Ai this is the norm.