You are right, lets keep this about Monero. I'm only interested in the ASIC mining aspect / discussion.
Some see ASIC development as a logical next step in a ongoing hardware evolution within crypto (specifically proof of work crypto).
Kinda like The Borg : highly adaptable and futile to resist...
With regards to centralization : i think that will happen at some point anyways, even with networks with only GPU's.
It will just be in the form of centralized GPU hubs, created by wealthy miners.
There are fundamentally different kinds of centralization. ASICs are extremely special hardware, and their manufacturers NEVER sell them while the mining process by their own ASICs is profitable. So, actual degree of centralization equals to number of ASIC manufacturers in this case. Furthermore, because ASIC manufacturing is a very narrow niche, typically just One ASIC manufacturer of ASICs, designed for specific mining algorithm, remains on the market after short competition wars. Thus degree of centralization become One. Might you know, Bitcoin is heavily centralized, because number of Bitcoin ASICs manufacturers (that ASICs are still effective for mining against current price) ~ 1. We, the Monero community, the Ethereum community, and many other coins' communities, DENY such a degree of centralization. We want every single man can enter the mining market with relatively cheap COMMON equipment to be bought in almost ANY local store right now. Today, common equipment is CPU and GPU. That's all. Of course, low degree of centralization will take place with common equipment too. But the nature of this centralization is conceptually different. That degree depends on (and only on) the Free Desire of Every Single Man to buy COMMON equipment in ANY local store and point it to mine Monero. Historically, before ASICs came to the scene, Monero mining was one of the most DEcentralized among all cryptocurrencies. There are many surveys one can google for. So, "the form of centralized GPU hubs, created by wealthy miners", as you wrote, has much less fundamental kind of centralization than ASIC-based approach. |
|
|
|
Hmm. Can we deny reorgs bigger than 10 blocks unconditionally? Mining is a random process, so it has its distribution in terms of math statistics with mean value for how many blocks need the reorg, and a standard deviation. Standard deviation can be accumulated and counted by each node, so they can simply rejects each single reorg beyond the 3-sigma rule. I think, manual API to check blocks beyond 3-sigma threshold proposed is a technical complication and philosophical violation. |
|
|
|
Hueristic, dEBRUYNE could you deliver this problem to Fluffypony and Monero Core Devs?
Since this attack started in earlier January, NO ONE MONERO CORE DEVS EVER COMMENT THIS. WTF?!
LOL, what do you want, a pre-announcement of centralized checkpoints like Dash's ChainLocks (which won't even go live for another 9-12 months)?  XMR devs have been working on a long-term solution called RandomX, a new PoW that is polymorphic and uses the blockchain for entropy. Please don't expect Fluffy to set his hair on fire and run around screaming, TYPING IN ALL CAPS LIKE A DRAMA QUEEN, just because the "XTend Hyperminer AMAP is a radically new, programmable digital mining fabric that can mine any coin, or even multiple combinations of coins simultaneously." Even if Xtend can chew through RandomX like a boss, I'm working something that will break its teeth.  This is just how the dynamic equilibrium of Nakamoto Consensus evolves. Please don't keep screaming and crying just because we're in one of the periodic recurring 'creative destruction' cycles. The more non-trivial adversaries attack Monero, the more antifragility accrues and the better our long-term chances of resisting extinction pressure. I've read about RandomX a lot. https://github.com/tevador/RandomXThis mining algorithm looks very promising in order to fight against ASICs and FPGA. But it effectively denies mining on GPU too. So all our honest GPU miners will be abandoned by the XMR community if we accept RandomX algo "as-is" on github. But indeed, RandomX looks very intriguing. It adds some random info into mining scratch pad to be collected from the blockchain. If I've understand correctly, it is exactly the idea I wrote few days before here... In order to adopt RandomX to Monero mining, Can we shrink down a standard RandomX data scratch pad from 4Gb to ~256Mb? So, in the client profile, when RandomX data scratch is generated on the fly, part-by-part, only 4Mb parts instead of 64Mb should be generated? I suppose GPU miners could adopt JIT-compiler version of RandomX to GPU mining. Furthermore newbies can play games and mine simultaneously, so this step decreases the threshold for GPU newbies to enter the Monero world, but still VERY resistant to ASICs and FPGA... |
|
|
|
I'm just a small GPU XMR miner since several month, and the last time I saw such a strong rise in hashrate was when bitmain mined in secret with theirs asics only known by them. It seems that they are re-offending. In less than 48 hours 60% of the hashrate is unknown ! Houston, we have a problem    The hashrate has more than doubled in less than two weeks! Something is happening. Either a 51% attack occurs, or specific asics are minning on the network recently. Is it that the most seasoned supervisors of the monero blockchain can confirm or deny what appears to be an attack ? Sorry for my English translated with google translate. Yes this is an attack. It is not so important whether this attack has 51% nature or not. This is a new era of hybrid ASIC/FPGA attack: https://www.prlog.org/12735136-monero-confirms-xtend-onlines-hyperminers-are-berylium-bulletproof.htmlor http://squirrelsresearch.com/cvp-13/TODAY: https://miningpoolstats.stream/monero804 MH/s Total, 340 Known, 464 UNKNOWN. Hueristic, dEBRUYNE could you deliver this problem to Fluffypony and Monero Core Devs? Since this attack started in earlier January, NO ONE MONERO CORE DEVS EVER COMMENT THIS. WTF?! |
|
|
|
Apart of being this one engine scam or not, there are some new mining equipment on the XMR network obviously: https://miningpoolstats.stream/moneroThey removed auto-calculator to avoid panic, but now Total Network Hashrate = 712 MH/s, Known Pools' Hashrate = 328 MH/s, so UNKNOWN Hashrate = 384 MH/s = 54% of Total Network Hashrate. Remember, miner equipment manufacturer NEVER starts to sell miners to public UNTIL they'll become UNPROFITABLE to mine by themselves. So when they start to sell them, you do always lose if buy. I think their equipment is indeed FPGA-like. --- I propose new idea to XMR mining algo: to emulate floating point computations of higher precision that is not common in any hardware. We should imagine how to involve a true random NOISE into computations. The way we involve NOISE should meet some requirements, for example, if one doesn't add NOISE, their results should be worse in metrics of hash mined. So FPGA manufacturer must add NOISE in EACH node of its FPGA logic. If so, FPGA will be as effective as GPU, CPU, etc... I may mistake... above are just free imagines... |
|
|
|
^This is a terrible idea that completely centralizes the network. You can't avoid FPGAs or ASICs forever, they're just another form of computer, albeit a bit less generalized. You can't keep hard forking every six months forever, it introduces centralization and uncertainty in the network/protocol. At some point, hopefully soon, Monero will stop running from this issue and just switch to a proof of work that isn't a pita to verify on low end general purpose consumer hardware and live with FPGAs and/or ASICs.
Official position of Monero Community is to keep entrance threshold level for new miners LOW, furthermore Monero Community does declare, apparently and specially, to resist ASICs, FPGA and any type of NON-COMMON hardware, i.e. hardware that is not available in the local stores of most countries. Recently, Ethereum Community join to Monero Community's fundamental principals to be ASICs, FPGA and NON-COMMON hardware resistant. The situation with ASICs in Bitcoin world is much more centralized, because ASICs are very specific hardware. Currently only one company produces most of Bitcoin ASICs. And of course there is no company that manufactures ASICs or FPGA (specially designed for crypto-mining like http://squirrelsresearch.com/cvp-13/) that will start to sell its NON-COMMON hardware while it is profitable for mining. Any company ALWAYS mines silently by its chips to its own profit and sell them when only they become UNPROFITABLE for mining. (PoS had been recognized by Monero Community as a strong unacceptable centralization too) If you disagree with official Monero Community, feel free to use a lot of Monero forks that are unresistant to ASICs & FPGA. P.S. What about the idea of formal pool registration on the Monero blockchain & sign each mined block by pool's signature. It does not violate the anonymity of a pool ownner. It PREVENTs centralization by rejecting a block candidate mined by a pool close to 51% in hashrate of total network. Any volunteer CAN register new pool by automatic payment a fee to the blockchain and meet formal verification like pool is publically available and has good enough formal rating voted on the blockchain too. P.P.S. the idea of formal pool registration has unique new attribute that is unavailable for any other crypto: Formal Trusted Mining Infrastructure, 51%-Resistant, Entirely Resides On Blockchain. |
|
|
|
Yes. They start to public Unknown Hashrate officially. There is no need to compute it by hand. Now "51.2% unknown". Monero core devs, WAKE UP. Weighted all pro and con I return to an idea of registering all pool signatures on the Monero blockchain. Ones pool mined a block, it should sign it by its signature. Other full nodes should verify in real time that a pool given Does meet some formal requirements to be publically available via http/https and performing regular payments. Each full node thus can calculate whether every single pool violates 51% or not. Unknown pools and pools with invalid signatures should be rejected from block confirmation process... Now let's think more generalized. May be pool verification described above is not necessary. But can we imagine that we store some info on the blockchain that can vary the mining algo by unpredictable way? So only when yet another few blocks are added to totally confirmed (full process resides in a time window ~ 1 day), the new current mining algo is known? Now let's focus out from info on blockchain at all. Can we imagine a formal way that mining algo must use some info from the real world? For example any true random generator or generators reside somewhere on the Internet? UPDATED. This FPGA is our enemy: http://squirrelsresearch.com/cvp-13/It's NOT an ASIC. So onces mining algo changes, new FPGA circuit will be built in few days or weeks depending on mining algo complexity to be understood by human mind... Human mind... May be to add some Deep Learning AI elements into mining algo?! UPDATED 2. AI networks are well known to be able to generate results with rich complicated internal structure. But AI generally can't explain or prove how it has obtained this result. May be AI should generate our miming algo? For example each day or each week or each month? AI should be embedded into blockchain and/or full nodes thus... |
|
|
|
BRAKING NEWS https://miningpoolstats.stream/moneroTotal 662 MH/s Known Pools 320 MH/s, Unknown = 342 vs 320 Known. > 51% UNKNOWN HASHRATE FOR THE FIRST TIME EVER MONERO CORE DEVS WHERE ARE YOU?! YOU DID NOT SAY A WORD ON THIS. (previous year ASIC hashrate was distributed among known pools mostly so we did not observe 51% unknown hashrate. and price was ~$300-400 instead of $43 now) |
|
|
|
Polynomial algo, we were discussing that when we were talking about QC resistance.
Can we imagine an algo effective on both CPU and GPU, not effective on both FPGA/ASICs, such as it relies on some real-world data that can't be predicted? if no, can we just fork 4 times a year, not 2? |
|
|
|
Off hand and on first thought a surprise fork does not sound like a bad idea as it will once again de-incentify asic development but forced registering of pools/miners is against all crypto stands for.
I don't persist on the idea of forced registering of pools/miners. I accept any better idea with great enthusiasm. What our practical experience show today: 1. Hard fork twice a year to alter the mining algo don't kill ASICs - they evolve faster than we think. Just 3 of 6 months passed since previous fork - and ASICs are on the scene again. Note cn/2 was designed ESPECIALLY to be ASIC-resistant compared to cn/1. 2. It is obviously a difficult task to distinguish CPU/GPUs and ASICs in terms of mining efficacy: new algo should preserve the CPU/GPU mining efficacy while be violent against ASIC mining efficacy. I've just read about CryptonightR in more depth and I like it. It meets the 2. requirement above. Let's CryptonightR will be our first step to fight against ASICs: https://github.com/SChernykh/CryptonightRAny thoughts? May be somebody sees hidden disadvantages in CryptonightR? |
|
|
|
Thanks. Man, that castle went through a lot of drama, I recall an earlier robbery / water damage situation as well. Surprised he didn't have a boat that sank as well ;-) But I guess it is best to let sleeping dogs lie. Sad news if Rpietila is still insane... I remember his great contribution in the community life. There was a discussion when Rpietila want to make a Monero real-world meeting in his castle... or a meeting had been done indeed?.. my memories about that discussion are dim now... |
|
|
|
Right now : 614 - 311 unknown vs 313 known = 303 UNKNOWN vs 311 known. ----------------------------------------------------------------------------------------- 608..614..617 it is increasing every single minute... 51% ATTACK is ongoing. Summarizing all the discussion above: Unknown hashrate ARE ASICs for almost SURE, because The Community fights against ASICs thus forces them to use hidden, unknown pools. Honest miners are becoming bankrupts right now: The Ratio Hashrate / Price = 608 MH/s / $46 = 13.2 Mh/s/$, BUT A YEAR AGO: https://bitinfocharts.com/comparison/monero-hashrate.html#1yhttps://bitinfocharts.com/comparison/monero-price.html#1yThe Ratio Hashrate / Price was ~650 MH/s / ~$350 =~ 1.8 Mh/s/$, and that time ASICs were NOT on the scene yet - they came in February 2018 increasing the hashrate by factor ~2 to 1GH/s (and now we observe the same factor ~2 from ~300 to ~600). I VOTE FOR IMMEDIATE HARDFORK WITH ALTERED MINING ALGO LIKE THIS ONE: https://github.com/SChernykh/CryptonightRCryptonightR is normalized in hashrate as our current cn/2 and is efficient with GPUs (but much not as well with ASICs/FPGAs). ------------------ Alternative/additional way I proposed: each pool should register in XMR blockchain its signature. Each block mined should be signed by pool's signature. Unsigned or mal-signed blocks should be rejected by the XMR consensus. This way can also count hashrate distribution among pools exactly, so the consensus may reject a block mined by a pool close to 51% hashrate. This proposal doesn't violate an anonymity of a pool owner: just pool as web-site must be public, https certificates may be done by "Let's Encrypt" for example. |
|
|
|
So when they has to manually do something for their FPGAs, the FPGAs has to be close to them, since they cannot every day travel all around? This makes impossible to rent the FPGAs all around the planet? This sounds a good idea.
But how UTC time is dealing with this? Has the manual handling to be done at very same moment for all the FPGAs in the one entity?
Actually I don't know HOW. Just my fantasies to start a discussion... Now I think my opponents that say there are NO ASICs today despite dramatic hashrate grow - they are right. Even if they aren't - next hardfork and mining algo adjustment WILL BE this April. My main pain is that new huge hashrate (currently 232 of 547 MH/s) doesn't belong to known pools: imagine they are Ethereum miners that switching to Monero - why they don't use existing well-known pools with good reputation?.. |
|
|
|
There was a proposal from hyc for a revolutionary new algorithm, named RandomJS, that would work best on commodity CPUs. I haven't looked at it too deeply but the little that I saw I liked. Doing ASICs or FPGAs for it would be extremely hard and the results would most probably not be efficient.
I think the downside of RandomJS is that GPUs would be at a significant disadvantage, and there are lots of miners that have invested large sums into GPU rigs.
I don't know what the current status of RandomJS is, but you can contact hyc on reddit and ask him.
RandomJS is obsolete. Its author proposes RandomX. RandomX requires main computer RAM ~ 4Gb, too huge. Furthermore RandomX denies GPU mining de-facto (not only FPGA & ASIC). But almost all our honest CryptoNight V8 miners are GPU miners. We must save the possibility of GPU mining. |
|
|
|
wow, the hashrate is going up! Nee equipment?
no wonder why my rigs are doing less recently.
I was warning about this few weeks ago. But community was too lazy. They told me like "man, relax". Today I read discussions on the known pools, for example, https://monero.hashvault.pro/, and all the miners are in deep anxiety. Today, according to: https://miningpoolstats.stream/moneroKnown Pools Hashrate = 322 MH/s Unknown Pools Hashrate = 555 - 322 = 233 MH/s And this 233 MH/s of Unknown Hashrate was 10 Mh/s just TWO WEEKS ago. IT IS GROWING TOO FAST. HONEST MINERS WILL BECOME BANKRUPTS SOON. Whether Unknown Hashrate comes from FPGA, ASICs, Botnets, or Ethereum GPU miners - does not matter so much. Main danger is that 233 of 555 MH/s are Unknown Hashrate. It is suitable only for attempt of 51% attack that was made a week ago against Ethereum Classic. WE MUST propose mining algo more resistant to both ASICs and FPGAs. Could we embed for example some kind a dependence on UTC time into mining algo? The goal is to enforce FPGAs owners to justify their hardware manually every day. While honest GPU users will use UTC time from central system clock of their main computers? Any better ideas welcome! |
|
|
|
Man, I don't say there is a reason for panic, there is a point to think on our steps to PREVENT future panic. Your mistake is that you consider high network hashrate only APART of other signals of danger, you just see one step near you, not far a bit. 1. There is no big deal with hashrate itself, THE PROBLEM is how is this hashrate distributed among Known Pools. The danger is that 100% of new creepy hashrate DOES NOT BELONG TO KNOWN POOLS AT ALL. Unknown hashrate ~= 200MH/s while Known hashrate (belongs to Known Pools) ~=300 MH/s, so Unknown hashrate ~= 60% of Known hashrate. Note, I even don't speak that ASIC or Botnet owners MAY use known pools : they MAY but they DON'T. THIS IS THE MAIN PROBLEM THAT SEEMS LIKE 51% TRY-ATTACK. 2. Don't compare hashrates alone. Compare them IN PAIR with XMR price. For simplicity consider exactly 1 year period. Today hashrate ~=500Mhz, price ~=$45, but A YEAR AGO hashrate ~=500Mhz AS WELL (nothing changed) but price was ~= $400. You can see easily that mining profit drooped in ONE YEAR for TEN TIMES, not 2, not 5, but 10 times. Our Honest Miners that pay for electricity are becoming bankrupts right now! All we remember Bitcoin past at 2010-2011. When did Bitcoin mining begin to be completely UNprofitable to mine with GPU?! WHEN FPGA and ASICS came to the scene!!! But official politics of our community is to be FPGA- and ASIC-resistant, or generally speaking TO BE RESISTANT AGAINST ANY SPECIFIC MINING HARDWARE. When honest miners with GPUs that pay for electricity can't mine and become bankrupts - THIS IS DIRECT VIOLATION OF THE COMMUNITY POLITICS. Hmm maybe there's some truth to what you're saying. The Monero Classic chain (the one with ASIC miners) dropped from 60.2B difficulty yesterday to 21B difficulty today. Maybe somebody is switching off their CryptoNightV0 ASICs and switching on their CryptoNightV2 ASICs?  edit and it looks like the XMC hashrate will continue to drop, their block times are super slow at the moment. May be you are right... Can you place here links to XMC hashrate/difficulty chart? |
|
|
|
Thanks for the links  Holly Shit! Today, according to https://miningpoolstats.stream/moneroXMR Network Hashrate = 528 MH/s Sum Hashrate Belongs To Known Pools = 330 MH/s UNKNOWN POOLS Hashrate = 528-330 = 198 MH/s ~= 200 MH/s UNKNOWN POOLS Hashrate / Sum Hashrate Belongs To Known Pools = 0.6 i.e. 60%. XMR Price = $42 XMR Network Hashrate / XMR Price = 528 / 42 = 12.5. Exactly year ago, January 2018, XMR Network Hashrate ~= 500 MH/s, XMR price ~= $400, XMR Network Hashrate / XMR Price ~= 1.25 https://bitinfocharts.com/comparison/monero-hashrate.html#1yhttps://bitinfocharts.com/comparison/monero-price.html#1yYou see, 1 year passed, XMR Network Hashrate ~= 500 MH/s as it was, but price dropped catastrophically from $400 to $40, ------------------- Dear Community Members, do you still believe this is NOT either 51% Attack or Hidden ASICs?! |
|
|
|
People that support a privacy coin like to solo mine or mine on private pools, what's wrong with that?
And do you think a 12 fold growth (from a low percentage to high) in a few days is normal organic growth? I confirm, it is obviously abnormal even... a bit mysterious and CREEPY: last XMR hardfork made on October 8 2018 had involved Cryptonote v8 (cn/2) mining algorithm that has significant anti-ASIC improvements (I remember it because I did read a discussion on cn/2). Today, just 3 (THREE) months passed. How can one make a huge amount of new ASICs against cn/2 algo?! In 3 months?! 200 EXTRA UNKNOWN MH/s are too huge for even ASICs made in 3 months! We remember, ASICs are HARDWARE not software. Right now: https://miningpoolstats.stream/monero309 MH/s Known hashrate, 200 Mh/s UNKNOWN, total 508 Mh/s P.S. May it be a bug in a block candidate confirmation unit? Or generally speaking, a bug in XMR source code somewhere... may be in cn/2 algo itself? |
|
|
|
Thank you for your warning.
I checked but don't see the pool that has dominant hashrates as you said.
Can you tell me which pool has 62% of Monero total hashrates, please.
This is UNKNOWN. All "legal" pools listed in https://miningpoolstats.stream/monerohave total hashrate just 312 MH/s absolute now. This is the PROBLEM: Unknown part of hashrate is 62% (507 MH/s absolute) as big compared to known part belongs to known pools. But just A MONTH AGO, Unknown part of hashrate was 5-10%! |
|
|
|
|
Show Posts
