Michael Sutton
@MichaelSuttonIL
Followers
20K
Following
11K
Media
189
Statuses
2K
Computer science, graph theory, parallelism, consensus; taking #Kaspa to the next level
Joined February 2021
“Fast, wide and yet thin” and its relation to “Based ZK rollups over Kaspa” . Kaspa is built on advanced proof-of-work consensus protocols allowing it to scale in blocks (BPS; frequency; fast) and in transactions (TPS; throughput; wide).
23
201
800
170
553
1K
Kaspa Crescendo timeline update:. End of February 2025: Testnet 10 transition activation.End of March 2025: Mainnet version release (activation will be scheduled ~30 days from the release; currently favoring a single, swift transition from 1 to 10 BPS, though this decision will
124
556
2K
Just completed a 6-page mathematical proof of a mass formula which on the one hand prevents dust attacks from bloating the utxo set while on the other hand special care is taken to have minimal/negligible mass cost on normal point-to-point txns. Still a draft. Gn/Gm:)
68
260
941
Mission accomplished. Activation successful. This exact set of consensus rules and consensus constants might be the one applied to mainnet not too far in the future.
The Kaspa Testnet-11 HF including KIP10 (first covenants), finalized KIP9 (dust-prevention) and payloads support has successfully activated at DAA score 287238000. The latest release also includes all the optimizations that we’ve been working on these past few months.
62
278
984
Maybe an unpopular opinion: a cryptocurrency that solves the trilemma doesn't need to convince everyone it solved it (right away or at all). The true properties it has achieved will showcase themselves in time, through challenges and dynamics as the economy evolves over it.
68
256
882
I see a few slightly concerned voices about the #Kaspa rust rewrite and the fact that it's not “done” yet. Well, let me announce it load and clear: the rust rewrite as per the original proposal is completed (long ago I'd say). The node is functioning extremely well, including >.
45
281
861
Leaving the most exciting parts to the end — the full encoding of the Kaspa GENESIS block has finally arrived at the rust codebase
51
271
805
Revealing Kaspa's 10BPS hardfork roadmap — the “Crescendo” Hard-Fork 👇.
50
270
857
Scary stuff ♥️.
🗓️Mark your calendars! .#Kaspa's Testnet 11 (#TN11) is going public this Sunday, January 7th at 8 PM UTC. Testing the new ⚙️#Rust codebase at an impressive 10 blocks per second!🚀. More details coming!. #DigitalSilver #10BPS #CryptoTwitter $KAS
39
202
811
Meet "The Kaspia" (הַכַּסְפִּיה; like "the cafeter-ia"). The renewed attic where Yonatan, Ori, Shai and others come to visit from time to time and work on Kaspa (and prove mass formulas over white boards which are undoubtedly the most important furniture here)
48
171
785
It's a nice opportunity to share this bit about the inception of RK (Kaspa's high-performance implementation in Rust). There's no logical explanation for what drove me into this madness (yeah the text in the image doesn't do), and I have a feeling that if we had waited until it
46
151
804
They are.
I'm gonna do 100 block/s with proof-of-work, who's coming?.
47
238
783
Working on #Kaspa I can find myself thinking/discussing 5 very different topics in a single day. Bouncing back and forth from mathematical proofs and modeling (KIP9, oracles, DK etc), to system design and future architecture (zk opcodes, L2 state handling), and code reviewing.
63
132
742
Technology-wise, Kaspa seems to be approaching a boiling point.
60
147
755
So I'm done with the most heavy research post I had to write. Let's start crescendoizinggg 🎼💃.Are you in for the ride @coderofstuff_ ? we got some stuff to code 😀.
37
128
739
It never gets tiring to go back and read posts by Yonatan. Especially with this 100 millisecond block time thing going on stabely for 8 days straight now. “We want a cryptocurrency to ping and be ponged”. — @hashdag, Dec 28, 2020
21
134
502
FYI me being quiet about #TN11 just means it is running smoothly. We are aware of some performance aspects which can be improved and we are working on that in the background. There are also 2 other big things coming: KIP9 (smart mass formula for dust prevention) >>.
38
189
658
Congrats on amazing three years #Kaspa and looking forward to so much more. I won't let you have too much fun, so here's some geeky stuff so you don’t spend the day only celebrating 😉. The image gives a bird's-eye view over what can be named “A compressed proof of Kaspa's PoW
33
189
688
It takes being slow and steady to build something which operates so fast.
39
118
687
To give some context, this is the version meant to “align TN11 participants behind a version which hopefully solves all current processing bottlenecks and provides a smooth operation of the network as a whole” (see Crescendo roadmap post). It contains several significant.
The TN11 HF release is here and link can be found in comment below. Please read the instructions in the release and upgrade your testnet 11 nodes accordingly. The hardfork is set to activate in TN11 at DAA score `287238000` which will be hit roughly on `Dec 3, 2024 1800 UTC`.
28
179
693
I'm in for implementing @hashdag's MEV (kickback via reverse auctions) and Oracle solutions. The idea is to use extreme high BPS and to utilize asynchrony and partial miner knowledge for creating competition (MEV) and voting systems (Oracle) between parallel miners. Who's coming?.
57
157
664
RK #Crescendo update. See the attached image in relation to the blog post timeline. Working hard on this top-notch performance version so that node spec requirements and configuration for 10BPS can be finalized. A lot is also happening in other coding/design areas, including but
Revealing Kaspa's 10BPS hardfork roadmap — the “Crescendo” Hard-Fork 👇.
66
217
667
Update: still running at this pace with remarkable stability. We are now focusing on stabilizing memory management of the node process and on smarter cache policies (pull requests 361 and 351 respectively). Both important milestones for stability in such high processing rates.
> update: still running stable just like this for 15 hours now. 5-6 miners and a slightly wider DAG on average
55
191
636
#Kaspa TN11 update: RAM stability work is merged. The above image was taken now and reflects a DAG with nearly maximum transaction throughput (equivalent of ~2K). Next step: increase TPS pressure s.t. demand is larger than DAG capacity and mempool gets filled
29
173
597
First consequences of this visit to beautiful Ireland and its infinite greens and grays. It was a real pleasure to be hosted by @KaspaKii and to have a chance to share my thoughts about Kaspa in my first international Kaspa adventure. #NewProfilePic #Glendalough
28
75
616
#TN11 some final tests and benchmarks. Here I'm running a controlled experiment where a transaction generation tool (named "Rothschild" if you wondered) is sending ~1200 TPS to the node. The node does not broadcast txs in order to make the test focused >
32
179
592
I can't help but share a technical point of view (I'm not involved in any Kaspa exchange integrations for years now, so this is purely based on speculation). Large exchanges need to operate at scale. Combine this with a new type of technology and you'll realize that it might.
58
115
617
A solution will not appear out of thin air. It will be based on years of multidisciplinary, forward-thinking research by people with sufficient vision who have focused on the right problems. It will be grounded in advanced technologies progressing in parallel across research.
Bitcoin is slow 🫠. Ethereum gas fees are too high 🥵. Solana network gets congested and .transactions are getting failed🤷🏻. What's the solution? 😒
23
158
610
After 4 long months, we are thrilled to finally unveil the initial benchmarks for $KAS on RUST. There is still a long way to go, but they seem to go beyond our expectations. The tl;dr is that 30 bps has never seemed more within reach. The plot shows block processing >>
43
193
553
For all who care for us in Israel, me and my family are safe. Please pray for us wherever you might be.
81
62
566
Our sages say: "wisdom I obtained through harsh times, is that which stands for me". DK has originated through 2020/2021 and during long covid quarantines. I guess its realization migh be done in challenging times as well, but that will be a strength of it, not a weakness.
$kas has no single point of failure. The state of affairs in Israel is hard and sorrowful. But as far as personal safety goes, it still remains very safe. You should be concerned for soldiers on the front lines, as are we. But we and our families are all safely within the.
51
130
564
Rust to Kaspa was like changing the engine of a vehicle in the middle of the ride*. The new engine supports the same spec, so the replacement is seamless, however it is much more powerful, so it can be accelerated (10 BPS hardfork).
30
146
574
Line after line of code apparently does add up at the end.
🎉SAVE the DATE!🔥.#KASPA on #RUST Alpha Releases is April 15th with Testnet & Mainnet Testing binaries! .Beta and stable releases to follow after community testing is complete. $KAS #BeYourOwnBank
42
141
540
$kas people, does this screenshot speak for itself??. (experimental #rustlang node synced live on mainnet 😀)
62
187
545
Dunno how I got into this but I have my blood, sweat & tears invested in this version's effort. Maturing with Kaspa towards fee market dynamics is an exciting milestone in the journey. More detailed release notes can be found in the following discord announcement.
The Release Candidate with RBF, Dynamic Fees, Optimized mempool and much more is here!. Wallets can now respond to dynamic and evolving fee markets with the new RBF and FeeEstimate APIs empowering end user to control their transaction inclusion speed. If
31
124
576
PR by @Max143672 & @OriNewman was just merged. This PR is a huge milestone for Kaspa and its script engine in particular. It adds.1. support for introspection** opcodes which allow inspecting the script public key and amounts of inputs and outputs (from.
49
175
566
Rust has hashrate majority now in case you wondered
33
154
535
There's no precedent that I'm aware of in the technology world where huge tech advancements can be dismissed by just anyone with a microphone for 'sociological' reasons or using amateur knowledge and superficial claims about the technology. Seriously? We're discussing a.
67
119
537
What we are witnessing is the remarkable phenomenon that a network encountering a new situation for the first time might face some engineering challenges related to that situation. In this case, if a network becomes congested for the first time, meaning it's entering a fee.
48
108
524
Just got off a really pleasant Kas core dev meeting to kickoff the planing and effort towards mainnet high-bps. Agenda included things that must be perfected on TN11 (as prerequisite), KIPs that require implementation and design thoughts about gradual consensus forking logic.
19
108
527
We made it with an elegant 0.5 hour lag. :) It's becoming interesting pretty fast, my node is already with ~20 peers (to me the extreme P2P dynamics are the most interesting to follow here).
Announcement from Discord:. I am glad to announce that the rusty $kas crew has released the new binaries for Testnet 11*, whereby marking the commncement of the second phase of the 10BPS experiment. If you wish to participate, here are the instructions:.
11
113
515
In the past 2-3 weeks I've been completely all-in in this "esoteric" task of implementing on-the-fly header-pruning for $kas on #rustlang. Now I know this sounds geeky as hell and unimportant, but I'm here today to explain why I see this as our last barrier before higher >>
40
176
467
Hats off 🎩 @CryptoAspect for this tremendous effort and amazing achievement. A broad system already built over #Kaspa on #Rust infrastructure, making the core layers accessible and extendable.
The #Kaspa Next Gen (KNG), the successor to the KDX platform has just been released! #alpha.Please help us test it. You can find more details about this release at The visualizer alone is worth the review!.#RUST #10BPS $KAS
8
118
477
#Kaspa hero @DesheShai spotted on a bus like an ordinary person with a weird 𐤊 on his t-shirt $kas
44
63
463
Doing our best to get back to business in this challenging time, discussing "minimizing peer to peer latency of block processing on 10bps" $kas
47
112
465
So what value should we insert to the left of the green arrow?.$kas on #rustlang, #kaspa_testnet_11 . - with @Tiram_88
126
144
473
Having the pleasure of watching Rusty Kaspa walk on its own, taking the weight and prestige of Kaspa's mainnet on its shoulders, I can't help but recall the following saying:. “A ship is always safe at shore, but that is not what it is built for.”.
15
96
472
$kas on #rustlang update — . All major components are completed; this includes the mempool, script engine, P2P infrastructure, RPC infrastructure, pruning-related components, and more. The core consensus validation engine was completed long ago. From a software development >>.
40
150
442
I ain't going anywhere before completing DK.
$kas has no single point of failure. The state of affairs in Israel is hard and sorrowful. But as far as personal safety goes, it still remains very safe. You should be concerned for soldiers on the front lines, as are we. But we and our families are all safely within the.
41
106
455
Recovering from a bad flu so I'm kind of semi-functional. Ask me anything about #Kaspa that can be answered in a single sentence.
185
54
473
My international Kaspa debut.
Kaspa in Enterprise event video. Michael Sutton - A Masterclass in Scalability on Kaspa . A must watch. Also on YouTube #kaspa $kas #kaspakii.@KaspaCurrency @MichaelSuttonIL
20
97
470
So here's another—rather technical—component of the L1<>L2 design puzzle: The canonical bridge (entry/exit mechanism); see link down the thread. Context. We are designing the bridge to support based zk rollup systems over Kaspa. We focus on the “based” setup as a first-class
10
98
477
"I will end with one aspiration. From the early days of Kaspa, I felt that the community was becoming unique in it's decentralizatio, but also in that more people than average are interested in the technology. I hope you will join the excitement of research".~ @hashdag
11
100
448
Was thinking of pulling an all-nighter on this “canonical” Kaspa research post. I'm probably 50% through, but the night is getting too short, and my eyes are growing heavy. :).What do you know—I'm finding myself writing about “Delegation scripts”, “Authenticity of the primary
21
58
461
They say it ain't official until devs themselves retweet it ;).
34
106
421
The fact that a public node I'm running has more than 40 incoming connections 2 hours after launch is astonishing for me and demonstrates the technical strength of the $kas community. Testnet 11 is now using maximal block mass capacity and is approaching 3K tps over 10 bps.
The $kas 10BPS testnet has been running smoothly for over 2 hours now. Even though we don't stress load transaction yet, some naughty participant decided to generate some anyway, whereby we are already witnessing more than 2000 TPS!. For the first time EVER a pure proof-of-work
17
138
423
Based zero-knowledge rollups over the appropriate L1 are indeed end game. The way I see it, this boils down to the following two significant points:. (i) Layer 1 can focus on scaling its throughput (data-wise) rather than scaling its computational expressiveness (I have.
Your number 1 goal this holiday season should be to learn everything you can about based rollups. Based rollups are the most egalitarian thing we've ever seen. It's insanely ambitious, and if it works, can really reshape the fabric of society.
17
129
481
I wrote a thread yesterday about the motivation for designing based zk rollups-like systems over Kaspa and linked my new research post at the end which explains how to design such a system over Kaspa. Perhaps there's need for a short intro to the how as well. The key for me
13
94
442
I know the main contribution of KIP9 is the quadratic cost of storage inflation, however here I'd like to put the spotlight on a conceptual discussion which highlights interesting relations between the UTXO model and account models, and their relation thereof to micropayments🧵
15
129
422
AMA about the current #Krc20-related network load and I'll do my best to answer (at least for the coming hour).
55
80
428
Kas start rolling the word Exa on your tongue. 0.5EH/s (0.5 quintillion hashes per second).
13
93
421
With @hashdag's determination for clarity on an algorithmic complexity question, we just hit a promising and important mempool optimization. Computer science trees to the rescue once again.
20
93
418
How curious are you to read the bits of this construction? .I think I'm about 60% through with the post 😃
16
71
432
You know that feeling when an optimization you designed and implemented is adopted by several entities or a network and you know that hundreds/thousands of computers are not wasting millions of CPU cycles anymore and you're all happy about it but no one else really cares? Oh you.
48
75
422
Imagine being a Kas follower with no sense of humor.
33
38
394
QED.
With @hashdag's determination for clarity on an algorithmic complexity question, we just hit a promising and important mempool optimization. Computer science trees to the rescue once again.
19
93
400
A day in the life of a Kaspa researcher/developer. - analyze heap profiles of a rust kaspa process overflowing OOM due to excess notification handling. - investigate a formula for tx mass calculation for encountering dust/QoS attacks (hint in pic; with @OriNewman & @hashdag). >>
15
105
393
What topic should the next workshop be about?. Reachability, Incremental Ghostdag, Superblocks and PoW history proof, Prunality rules and data pruning, DAA, Block reward scheme, IBD & P2P, Mempool and transaction selection, are all some topics that come to mind.
If you can read Rust and you find UTXO DAG processing interesting, please join us. Note that basic prior knowledge and some familiarity with the RK codebase is assumed
31
105
396
I'm planning to write a document in the coming days to summerize Kaspa on Rust stability progress and estimated timelines for testnet and mainnet. Committing here so that you can stone me if I don't. In short we are completing the first stability improvements cycle, including >>.
23
112
378
> update: still running stable just like this for 15 hours now. 5-6 miners and a slightly wider DAG on average
21
78
377
Imagine being an American post 9/11, but instead of mourning your people you are busy convincing the world that the twin towers actually existed, that airplains literally crashed into them, that actual people were jumping down. This is how I feel wandering around here these days.
48
105
388
A sparkle of thought, one night 3 years ago, coming into full existence .@hashdag.#dagknight
24
99
383
Fwiw, all miners of a cryptocurrency agreeing to a consensus rule change (aka hardfork), is just as secure as them constantly--every single second--agreeing to not change the rules.
20
72
373
Apparently someone is working on getting this talk ready for publication 😀 I couldn't thank you more @fishtuna, the work on this seems tedious and technically complicated and possibly took way more effort than me preparing it in the first place
11
83
378