Solana Client Agave Smashes 1.1 Million TPS, Matching Firedancer’s Record
Solana’s Rust validator consumer Agave has matched Firedancer’s lab report, briefly peaking at 1.1 million transactions per second in an artificial, single-node benchmark. The consequence was disclosed by Solana core engineer Andrew Fitzgerald, who wrote: “Hit a burst of 1.1m TPS on agave this morning. Single-node artificial check with easy transfers. On a department with a number of modifications not but merged,” earlier than itemizing PoH recording and status-cache enhancements alongside new “scheduler-bindings,” with block/shred limits disabled for the run.
Pressed on caveats, Fitzgerald added a plain-English qualifier: “disabled block/shred limits, and it’s a burst not sustained. have a number of long-term perf enhancements squashed on prime of scheduler-bindings to make this occur.” The alternate underscored that this can be a laboratory datapoint, not a network-wide throughput quantity.
Solana’s Core Shopper Smashes 1.1M TPS
The milestone instantly fed a broader narrative about client-level competitors on Solana. Helius co-founder Mert Mumtaz framed it this fashion: “final 12 months Firedancer hit 1.1M TPS on an artificial check — now, Agave has finished the identical. There’s an outdated notion that Solana will turn out to be quicker provided that Firedancer. That is from a time when Agave wasn’t as aggressive — however it’s now. The competitors between the 2 consumer groups will enhance the chain like by no means earlier than.”
Anatoly Yakovenko, Solana’s co-founder, poked enjoyable on the victory lap, quipping: “Pls no extra. Simply ship ag and decrease the timers to 150ms.” The comment tracks with the community’s ongoing push to scale back consensus latencies. That push is presently crystallized in SIMD-0326 (“Alpenglow”), a consensus overhaul now in neighborhood voting that targets ~150 ms block finality by remodeling how and the place validator votes happen.
Below the hood, the Agave check hints at the place efficiency headroom is being unlocked. “Scheduler-bindings” — a forthcoming extension that lets validators plug in customized block-packing logic with out forking core — has been on Anza’s public roadmap since Could. Current Agave 2.3 literature additionally particulars a revamped TPU consumer (“tpu-client-next”), AccountsDB I/O reductions, a grasping scheduler, and snapshot/gossip enhancements, all of which lower real-world overhead even when they don’t present up in artificial peaks one-for-one.
The apparent query is what the 1.1M TPS burst really means for customers. Artificial single-node checks measure uncooked execution and scheduling throughput with some guardrails quickly lifted; they don’t translate linearly to mainnet capability, which is bounded by community propagation, signature verification, scheduler coverage, and financial constraints. Nonetheless, the quantity is directionally in line with the community’s trajectory. Earlier this month, unbiased experiments noticed six-figure TPS bursts on mainnet beneath heavy program-call load — a separate datapoint that, taken along with Agave’s lab determine, reinforces the tempo of optimization throughout each consumer and protocol layers.
Two broader takeaways stand out. First, Solana’s consumer range is not hypothetical: Agave (Anza) and Firedancer (Jump Crypto) at the moment are buying and selling blows on the identical artificial leaderboard, with completely different codebases stressing completely different elements of the system — a wholesome signal for resilience and future efficiency. Second, the product focus has shifted from trophy numbers to latency and predictability: the 150 ms goal, if adopted, compresses user-perceived finality in ways in which matter for funds, buying and selling, and real-time apps even when headline TPS fluctuates.
At press time, SOL traded at $207.86.
