AtlasPool.io

Attention BTC Solo Miners! Check out AtlasPool.io, the most performant, reliable, and globally available solo mining pool available today. Point your miners to solo.atlaspool.com:3333 for fast access to 100+ globally deployed Points of Presence.

Excited to officially work with @IxTechCrypto as a hardware partner! Their Bitaxe, Nerdaxe, and NerdQX miners now ship preconfigured to point at AtlasPool out of the box — plug in, power on, start hashing. Huge respect for what the Ix Tech team is building for home miners. ⛏️ 🟧 #BitcoinMining #HomeMining #AtlasPool

@tatumturnup atlaspool.io/resources/arti… I wrote an article about this back in April :-). It's a fun read.

I just scanned 950k Bitcoin blocks to find the block with the highest effective difficulty in Bitcoin history. Block 756,951 was mined October 4, 2022 and had an effective difficulty of 50,541,668,626,379,227,136 (50.54 quintillion). This hash is only the second ever to have 24 leading zeroes (the first was mined less than two months prior). At the time, that was 1,611,632x the required difficulty, or 99.99938% below target. This also wins the biggest beat relative to required difficulty which was block 368,527 at 99.99991% under target. Fun fact for you. Gm.

@almsec007 All good! You can rent from @NiceHashMining, @Braiins, or @miningrigrental and configure to point to AtlasPool (solo.atlaspool.io:4334)

@AtlasPool_io How does this works ? Do it rent some hash and add it to my solo miner Or is it separate 🤔 Sorry for the question am learning

🚀 Hash rental day on Wednesday, May 13 on AtlasPool! All rentals welcome… any size, any duration. All major rental outfits supported by AtlasPool (which includes a rental-friendly high difficulty port - :4334). For maximum effect, consider the rental for a 1200 UTC start, but not required. Spend responsibly and have fun. Most of all, good luck and get that block! Who is in?!

@proofofmike Thanks for your kind words and your support! Go get that block!

Atlas has done a lot in a short period of time for this space. Globally distributed infrastructure, work identifying scam pools and bad actors, testing/tuning tools, etc. Let’s help get them their first block today. 🧱🍀

Rental hash day for AtlasPool -- all are welcome!

@CryingLady_ETH Welcome! Go find that block!

I’m putting all my Hash powers into Atlaspool 😭 After researching almost everywhere online, I can see that our Atlas guys have huge knowledge and expertise in their field of work! So let's go hashing 🥺 and find those Blocks 🚧 @AtlasPool_io #solo_mining #btc #bitcoin

AtlasPool users are awesome! We maintain a very friendly and helpful Discord server. All are welcome. Check out the most globally distributed and available (uptime) bitcoin solo pool in the world. AtlasPool rocks.

@Pmaxsd @ypto_pia Remember this? x.com/atlaspool_io/s…

@ypto_pia No. But in general, it’s a good idea not to trust pools blindly. In crypto, scams are everywhere, and I wouldn’t be surprised if some pools mixed jobs with their own payout address into the Stratum job stream. You would never notice.

Lottery Mining Verification is coming to the Nerd*axes 🥳Your device now checks that your address is actually part of the pool’s coinbase transaction payout — and disconnects if the pool fee exceeds your limit. Trust, but verify.🔍Test it now in 1.0.38-alpha1 🔥#nerdqaxe #bitcoin

@FlickyMcDicky More info here. There is a 1.5% fee if a block is found. atlaspool.io/connect-now.ht…

@AtlasPool_io What do I put in the stratum host box to run in your pool is there a fee if I do hit a block or do I get it all?

Attention BTC Solo Miners! Check out AtlasPool.io, the most performant, reliable, and globally available solo mining pool available today. Point your miners to solo.atlaspool.com:3333 for fast access to 100+ globally deployed Points of Presence.

@ckpooldev @infoSTFSTJ Awesome work. How does memory consumption look like for 41k workers?

For interest, here's the CPU and memory usage of the ckpool threads ordered by total CPU time on solo.ckpool.org with 41,000 workers from 24,500 users. It's using just over 1 core's worth of CPU of an Intel Xeon E-2388G CPU in total.

@FlickyMcDicky @huskuldar Higher latency can contribute to a higher stale/rejected share rate. And rejected shares are wasted work that your miner was doing instead of working on something that might yield a real block find. Give AtlasPool a try :-)

@huskuldar @AtlasPool_io Which makes me less likely to hit a block?

Just ran it, that is an amazing tool. @AtlasPool_io

@huskuldar ckpool (used by AtlasPool) is written in insanely fast C code. Public Pool is written in TypeScript, which can't compete against C from the lens of speed.

Interesting that the CKpool/Bassin -port 3456 on Umbrel has a lower stratum time than Public Pool port 2018. I used your --runs 3 command and did it many times, same results. The miners show only ping in the UI, so not sure that helps, but all of them are low, 3-6ms, no matter which pool they mine to. The other thing I found is that if you use umbrel.local the stratum time jumps massively and is very inconsistent vs using the actual Umbrel device IP address. All my miners will be using the IP address from now on; most were already. Your tool helps in more ways than I thought!

@SoloPoolBTC @altair_tech I will add you.

@altair_tech @AtlasPool_io Left us out :) SoloPool.com

Ever wondered which mining pool is actually fastest for you? ⚡ @Atlaspool_io built a tool to test: 📡 Latency to any pool ⛏️ Stratum response time Make a data-driven decision on which pool is least latent from your home network Try it 👇 atlaspool.io/speedtest/

@KarmaCpu @huskuldar @DrBudd42014 The stratum connect time is representative of what your miner does when connecting to a pool. Agree with you that ping alone is not the right measurement.

@huskuldar @DrBudd42014 @AtlasPool_io Check the ping on the actual mining device these tools dont give real honest answers you will see when you actually connect your device on real life latency

@FlickyMcDicky @huskuldar Post here with the results if you switch to AtlasPool :-) Curious to know the difference. Bonus: better uptime too.

@huskuldar @AtlasPool_io Mine are sitting around 105 I probably need to switch it up

@philip_dath @ckpooldev Thank you for insisting on the highest standards, Philip. This is awesome! AtlasPool will fold these changes into its stratum servers very soon! Nice work!

I hope, for your sake, that I don't have any other good ideas you like. :-) To everyone, I think this has massive consequential impacts for solo miners. Massive. I will be very interested to see in 12 months' time the average time to solve a block on the bigger ckpools compared to before this code change. Having 40,000 workers start every new block 10ms sooner over a year (starts writing on napkin) amounts to an *additional* combined mining time of 34 weeks each year! Blocks per year = (24*60/10)*365 = 52,560 Combined Extra mining time (s) = 52560*40000*.01 = 21,024,000s Seconds per week = 7*24*60*60 = 604,800s Divide = 21,024,000/604800 = 34 weeks!!! ckpool often has 100PH/s to 200PH/s of hashrate. That's an additional 34 weeks at 100+PH/s of hashrate. This must result in a long-term increase in the number of blocks solved by ckpool. And that is with just the new block template processing. That's ignoring all the messaging improvements that @ckpooldev is still testing.

Ckpool rework for yyjson. Whilst ckpool is very low overhead and scalable due to being coded in multithreaded C, one of the largest overheads within the code is the json parsing library. Libjansson has been included since 2014 and it touches approximately 5,000 lines of code with ckpool. Since that time a number of higher performance json libraries have come out, with yyjson proving to be the fastest - 10-100x faster depending on the workload. I've resisted reworking the code for a long time, but the benefits are substantial - not only would ckpool be much lower overhead and even more scalable, it would lead to much faster block template changs which are a critical performance measure in mining pools. I've embarked upon a major rewrite of all the performance critical components of ckpool to use yyjson. It is unfortunately far from a drop-in replacement and has required some significant surgery. Some of the safer code has already been merged into the master branch on git, but the massive rewrite has been committed to a yyjson branch. I'm currently testing the existing code on testnet4, and will continue refining it further before eventually merging it into the master branch and deploying it on the solo ckpools. The solo ckpool currently would probably safely handle 10x its current load (which is 40,000 clients). I don't know how much of a real world improvement it will work out to, but I'm hoping for good things. At the very least, faster block changes are always worth pursuing. Watch this space, as I'll be announcing when I'm deploying it on solo. And no, this isn't a modification any existing AI has a hope in hell of vibe coding, so it's all done by hand. Current code: bitbucket.org/ckolivas/ckpoo…

@ckpooldev Props for kicking it old school and writing code by hand. Have you considered using AI as a second set of eyes for ferreting out any possible bugs, security issues, or bottlenecks that you otherwise might miss? Just curious what your view is of AI's place in coding.

@ckpooldev Nice!

I've successfully replaced all the performance critical json in the main pool code and the bitcoin daemon parsing, leaving other modes (proxy, passthrough, node, etc.) to use the old code, and have addressed all the potential memory leaks. The code's looking very solid right now and I'm hoping to get it up in a day or two on solo. I'll deploy it in a staggered timeframe over the 3 pools as each proves stable. It's the most substantial code change to ckpool in many years: 15 files changed, 1124 insertions(+), 540 deletions(-)