Monero Announces War on ASIC Manufacturers, Hacker News

Monero Declares War on ASIC Manufacturers, Hacker News

My beloved part about it is response to criticism that the switches may pauze the underlying cryptography they are using-

>, Wij do not have cryptographers familiar with this kleuter of thing, sadly. [Two]

Basically, I think it actually weakens the FPGA-resistance of Cryptonight.

Fortunately, the main idea of the switches would discourage ASICs being made for Monero. But it doesn’t seem to discourage FPGAs, which are just one ",resynthesize", away from fixing any code switches. And spil long spil the community becomes habitual to regular switches, then any issues ter the encryption can be motionless straks.

If FPGAs can implement thesis switches efficiently—more efficiently than cpu miners—-and be re-synthesized to implement further switches, doesn’t that advantage them?

FPGAs don’t suffer this problem

Something like an FPGA + Interposer to HMC would be a thick R&,D effort, and just spil centralizing.

BTW: None of the designs I’ve talked about are rigorously speaking commodity. They’d require at a ondergrens, custom-made PCBs. Maybe more advanced technologies for the best technology (again: Custom-made Interposer to HMC + FPGA interface + all the Verilog / VHDL code to make it toebijten).

Spil long spil an FPGA-based shop kept their FPGA PCB secret, spil well spil their code secret, and their Device Drivers secret (You’d very likely run Linux / Windows to talk to the FPGA overheen PCIe) then they’re basically going to be ahead of the surplus of the competition. Eventually, the competition would catch up, but a onveranderlijk R&,D effort into newer designs (ie: testing HBM2 vs HMC vs RLDRAM3 vs QDRIV, building relationships with suppliers, etc. etc.) could lead into a sustainable business edge.

Heck, early on te Monero’s life, some dude got to like 40%+ of the entire network’s Hash Rate by simply writing better CPU code and keeping it for himself, and then spending hundreds-of-thousands of dollars on AWS:

>, By the 14th of May, wij were 45% of the total hashing power on the coin. Things commenced to get a little arousing

FPGAs would be that on steroids. There are way fewer Verilog / FPGA engineers. It also would require custom-built PCBs and hardware engineers to make it all toebijten. I wouldn’t even know who to ask to vormgeving an Hybrid Memory Cube interposer and to gezond it on an FPGA for example.

Not cheap, and most likely not available te bulk, but cheaper than custom-built PCB R&,D.

16.Five MB of internal RAM. 18MB of QDR-IV SRAM. 125MB of RLDRAM Trio. 2GB of HMC.

That’s slew of RAM to pretty much crush Ethash, Cryptonight and more.

Heh, free money incoming if you’ve got a crypto background. Cryptocurrency is a nested fractal of incompetence.

It works just fine if you substitute ",Software development", for the 1st word.

By far, the majority of GPU farms mine Ethereum. This mean any GPU-mineable coin other than Ethereum is trivially vulnerable to majority attacks. The only thing preventing thesis attacks from happening is the financial incentives against performing such a purely disruptive attack.

ASIC-mineable coins are safe from such attacks from GPU miners.

People point out that the high barrier of entry into ASIC manufacturing can lead to monopolies, such spil Bitmain which is a quasi- (not fairly) monopoly with its 70-80% market share te Bitcoin with the S9. But very first and foremost, manufacturing monopolies don’t matter to the function of Bitcoin. A manufacturing monopoly isn’t a hashrate monopoly: Bitmain cannot attack Bitcoin. They themselves only directly own <,10% of the hashrate.¹

¹ Not to be confused with Bitmain’s mining pools which represent more than a third of the hashrate. People usually misunderstand the function and nature of mining pools. End-users choose to mine at whatever pool they want. If Bitmain’s pools began sending malicious mining jobs to stratum clients, the community would react and the end-users would promptly abandon them. So ter practice Bitmain cannot do whatever they want for however long they want with their pools’ hashrate. They love the privilege of signifying this hashrate, and this privilege could evaporate overnight.

(Since you talk about decentralization, being GPU-mineable doesn’t necessarily increase decentralization either. Bitmain could waterput 100 MW of GPUs into mining some random coin.)

digiconomist [0] estimates that current etherum mining cost is 1.7 billion a year, or Four.65 million a day, or 194,000 an hour, or 3234 a minute.

Multiply by Five for cloud on request premiums and you could predominate the etherum network for an entire day for 23.25 million. You could also do it for free if you can manage to do it with stolen credit cards. I’m amazed it hasn’t already happened to a cryptocurrency.

What can kleintje of cool stuff can you actually do when you predominate the etherum network?

Sure 23 millions seems joy, but treating that many GPU is expensive and if they doesn’t have differents markets to support that usage, it’s a big loss to them (and if they actually have the market, then, that single 23 millions isn’t worth much either).

I see your point about ASICs raising the caf but wouldn’t that waterput contributions out of reach of all but high-wealth or dedicated miners? Which is basically what has happened to Bitcoin. It is downright uneconomical for anyone but people with free to very cheap electric current and lots of ASICs.

I believe the point of Monero is that power shouldn’t be consolidated to a single few.

Not at all. ASICs don’t increase the gap inbetween the low-hashrate and high-hashrate miners any more so than GPUs. It’s all proportional. If you have x% of the global hashrate, you get x% of the revenues. This is one of the fattest misconception te the mining world. You see journalists claiming that ASICS make mining harsh for smaller miners. But this has absolutely nothing to do with ASICs. It all comes down to electrical costs: the large miners can afford to set up shop ter area with cheap electric current while puny miners have to make do with whatever rates they pay where they live/work. The difference inbetween $0.03/kWh ter Douglas County, WA (where many industrial miners are located) and $0.20-0.30/kWh ter SoCal is enormous. That’s why large miners profit more. This would be exactly the same if the coin wasgoed GPU-mineable.

This once again reduces the playing field to a puny, powerful group of players and essentially no one else can participate. Doesn’t this defeat the purpose of a decentralised currency based on not having ",trusted knots",?

Comparatively, I do have a nice graphics card (which I can use for gaming and stuff), and I can rent from cloud providers at price points I can afford if I desired more access to cards.

(1) Their higher capital cost, limited availability, and limited ",appeal", promotes effective centralization of mining.

(Two) Due to #1 mining prizes go to fewer an fewer arms, promoting inequality and centralization of wealth.

(Trio) ASICs increase the overall wastefulness of PoW currencies since they’re e-waste. Mining ASICs are worthless for any other purpose, while CPU-mineable coins at least promote the production of things that are otherwise useful. If a miner liquidates a bunch of CPUs and GPUs they will see use on the market, while ASICs are junk.

Similarly, locations with the cheapest power force centralization of location, which coerces centralization of state regimes presiding overheen miners.

Monero doesn’t suffer from thesis centralizing compels spil applying generalized computing power can be done with spare cycles and is still profitable at the margin. Te other words, you might not buy a rekentuig to mine Monero, but if you have one, it might make sense to use your spare cycles for mining. This gets you a widely distributed polyculture, which is the best kleintje of decentralization.

I don’t think so. Competitors will always attempt to catch up given how profitable it is to sell mining equipments. Wij certainly have bot going through a phase overheen the last Two years where Bitmain has bot vooraanstaande, but today wij have Trio manufacturers who have miners just spil efficient spil the Bitmain S9 (0.Ten J/GH):

• Canaan who just released the Avalon 821 (0.11 J/GH)

• Halong is promising a miner just spil efficient te ",Two months",

Here ter the US, it’s very common to have a vuurlijn doorheen which is never locked.

You care about having a safe vooraanzicht wegens to the extent that you anticipate future attacks. If you don’t, then you don’t.

More specifically, the fact that a trivial attack didn’t toebijten yet is not always an argument against the need to anticipate that attack.

Just like mining pool wasgoed not an expected event, humans will always figure out a hack to make more money where there’s money. Especially when there is no legal implication for doing so.

Thesis monero developers think they’re doing good for their network, but all this is doing is weaken their network based on the limited skill they have about what’s possible presently. I would even go further to say this is arrogance.

There are many ways to compromise a network, and the best way to protect against this is to strengthening it, instead of making foolish decisions based on some conspiracy theory that nobody has everzwijn actually seen played out.

For example if someone figures out a way to bring together the coinhive JS mining library and a clever web exploit scheme to create a global ultimate monero mining worm, it could compromise the entire network and they will have no choice but to hard fork. The only way to protect against this sort of attack is to strengthen the network. And only then you deserve to worry about centralization.

I know this ",miner centralization", is a controversial topic, but what I know for sure is that people are making decisions to permanently switch protocols that are already working, ter order to ",fix", some hypothetical situation that may or may not toebijten.

If humans were so good at prediction, wij would not have good depressions. The best way to overeenkomst with thesis issues is to solve them when they actually toebijten. I think this is the best for all cryptocurrencies because IF some sort of centralization actually does end up happening, people can fork and take a phat chunk of network with them always.

The people who think wij can’t recover from thesis hypothetical centralization are making some assumptions that are more likely to be false than true.



Leave a Reply

Your email address will not be published. Required fields are marked *