Intel To Unveil 'Ultra Low-Voltage Bitcoin Mining ASIC' in February (coindesk.com) 165
Intel, one of the world's largest chipmakers, is likely to unveil a specialized crypto-mining chip at the International Solid-State Circuits Conference (ISSCC) in February, according to the conference's agenda. From a report: One of Intel's "Highlighted Chip Releases" at the conference is entitled "Bonanza Mine: An Ultra-Low-Voltage Energy-Efficient Bitcoin Mining ASIC." The session is scheduled for Feb. 23. This brings the company into direct competition with the likes of Bitmain and MicroBT in the market for bitcoin mining ASICs, or application-specific integrated circuits, for the first time.
April first? (Score:3)
My calendar must be wrong.
Comes With a lump of coal (Score:3)
You burn the coal in your house to savor the carbon emission personally
Intels strategy (Score:2)
Intel is likely most afraid of tiawan semi on the fab side and Nvidia on the computer arch side. Taking away Nvidias cash cow n mining would be useful to slow them down . Eventually I think the play is to transition to AI based on ASICS . But that's not a deep pocket yet. Getting there via ASICS again beats nvidias game. However in the meantime Intel is going to work the fab for hire market and any high volume fab operation as the means to fund back next gen fab lines to compete with Taiwan semi
Re:Intels strategy (Score:4, Insightful)
Bitcoin isn't mined with graphic cards.
Thus I fail to see how Intel making a Bitcoin ASIC would impact nVidia in any way.
Re: (Score:2)
Re: (Score:2)
No. This isn't a zero sum game. Bringing a new efficient ASIC to the market for Bitcoin doesn't replace a GPU mining Ethereum or whatever else, it simply leads to two mining machines sitting on the table instead of one.
Re: (Score:2)
Re: They call it Proof of Work, but where's the wo (Score:2)
The short answer is that there isnâ(TM)t enough legitimate work available.
Not even if you had some sort of magical job scheduling system and a magical bus to connect the jobs to CPUs. If you define work as something useful (that in and of itself is difficult :)) then there just isnâ(TM)t enough of it. The computing hardware that would otherwise have been used for bitcoin âoeminingâ would therefore not exist, sadly.
The various crowdsourced computing apps are successful because they trade
Re: (Score:3)
Nvidia isn't used to mine bitcoin. It's utterly uncompetitive in doing that.
It's used to mine ethereum. And ethereum is very ASIC-resistant by design, in that it doesn't require much compute, but it requires a lot of very fast memory.
So intel going into bitcoin mining isn't doing anything to nvidia's position in the crypto market. It's going to compete with bitmain et al ASIC makers.
Re: (Score:2)
Re: (Score:2)
Ethereum ASICs already exist. They're just not very competitive with GPUs because the primary requirement is a lot of fast memory and a very fast memory bus. ETH has intentional software design choices specifically to mitigate against becoming mined mainly by ASICs, as GPUs have a rapid update cycle and take up the most expensive and fastest memory off the market wholesale every generation as memory bandwidth is very much needed in 3d graphics.
Re: (Score:3)
Technically there are ETH ASICs. They aren't that great though, and with PoS coming by June/July it'd be foolish to invest in a purpose-built ETH miner by now anyway.
Re: Intels strategy (Score:2)
PoS isnâ(TM)t coming by ever. Not this June, not June 2027.
Re: (Score:2)
Hey stick to your guns if you want to. Plenty of other chains are doing it. Ethereum is playing catchup. If they don't, they're dead long-term.
Re: (Score:2)
ETH ASICs are not very competitive because the primary points of "competitiveness" in ethereum is memory speed, memory bandwidth and memory availability. GPUs rapidly cycle to newest and greatest in memory because of just how constrained 3d graphics are by those three things. ETH ASICs are just not big enough of a market to keep up.
And no, PoS is not coming to ETH. Ever. Stop with this silly fake news from almost half a decade ago.
Re: (Score:2)
Not very competitive with what? The latest ASICs are slightly better perf/Watt than 3090s and the like. They're just not proportionately better than dGPUs as compared to Bitcoin ASICs.
Re: (Score:2)
Not competitive with those GPUs, because worth of used ASICs is near zero. Worth of 3090s and the like in the aftermarket is excellent.
Reminder that it's not just the efficiency. It's also the total life cycle.
Re:Intels strategy (Score:4, Funny)
It is "Proof of Steak"...it means that you're bringing some meat to the BBQ.
(poking fun at the unwashed masses who misuse homophones like 'brake' and 'break')
Re: (Score:2)
Pointless old Shitposter, apparently.
Re: (Score:2)
Re: (Score:2)
This logic applies for literally everything, including real miners. Which is also why it's utterly wrong.
This will only help in the short term (Score:5, Insightful)
Even is this chip has much lower power dissipation than current mining ASICs it won't help in the long term. The amount of hash power across the network will just increase to eat up any power savings until the power dissipation is similar to it is today. The thing limiting the hashpower is the price of electricity, not how efficient the mining ASICs are.
It is part of the core design of Bitcoin (and other proof-of-work cryptos) that the inefficiency of the network itself is what secures the network. If you make it cheaper in power to calculate a hash to try to win a Bitcoin reward, the miners will just calculate more hashes to compete. No net benefit (from an energy standpoint).
At best this earns some money for Intel. It won't make Bitcoin any more environmentally friendly.
Re: (Score:2)
Right, but I wonder if units like these would shift the economics away from the use of GPU to mine. If so, then it would at least potentially address the GPU shortage.
Probably not though. We're probably not that lucky enough to escape the grasp of this crypto nonsense ruining everything from worldwide electricity consumption and my terrible gaming rig.
Re: (Score:2)
My understanding is that most Bitcoin mining has already moved to ASICs; it's other cryptocurrency that uses graphics cards. This is more about adding a new competitor to the field than changing the basis for mining.
Re: (Score:3)
Exactly. If it truly was profitable, intel would use it to mine bitcoins rather than selling the chip.
How do you make money during a gold rush? Not by mining gold, but by selling supplies to miners.
Re: (Score:3)
There are significant risks involved with mining. But nothing will stop intel from "test mining" for a month or two with the first batch if they have a superior product where they have vastly superior hash rate and capture a good chunk of market. And then going "want to actually start being competitive in mining bitcoin again? Here's our latest. Buy or leave".
At which point bitcoin per unit of energy collapses as total hash of network jumps, and mining with superior ASIC gets rapidly less economically viabl
Profit from buying AND selling picks and shovels (Score:2)
If it truly was profitable, intel would use it to mine bitcoins rather than selling the chip.
If if was profitable to money with computers, Intel would use them to make money rather than selling the CPUs.
Hmm, that doesn't sound right.
This is how selling any good is. The seller of the thing gets $x now, the buyer expects a benefit of $x+n in the future. The difference is that each party maximizes their profit in accordance with their own acceptance of risk.
Re: (Score:2)
At best this earns some money for Intel. It won't make Bitcoin any more environmentally friendly.
Intel is fully aware of this. Do not buy from Intel.
Re: (Score:2)
*More* raw materials used to construct the efficient cards, *more* landfill from the old ones.
Re: (Score:2)
Besides, I'm hoping that this all becomes a non-issue and that proof-of-stake becomes the dominant strategy. PoW is just so wasteful.
uh (Score:2, Insightful)
Re: (Score:2)
In this specific case, the OP is right because there are specific technological control systems in BTC that will quickly eliminate the savings, and there won't be any corresponding benefits (except to Intel vs other asic makers in zero sum). BTC has been completely resistant to brute force head on attack (by node takeover, which the PoW system suppresses) for many years now. It's now like adding 10,000 more locks to an already secure front door, except these take more energy.
BTC developers (i.e. Blockst
So lead by example (Score:3)
Anti-Crypto folks should probably disconnect from the Internet and shut their computers and TVs and all electronics off.
So lead by example: do the rest of the world a favor and unplug yourself from the internet.
Re: (Score:2)
Nah, you first!
Re: (Score:2)
Bullshit, go get the logs from /. if they'll give em to you. I wasn't the AC. I just hate it when people post like that. You want people to harm themselves? Harm yourself instead.
Re:This will only help in the short term (Score:4)
In the cryptocurrency case the problem is just made harder if solving it becomes more energy efficient; with incentives to consume as much compute as the current price of electricity vs. the current value of whatever is being mined will support, no matter what hashrate that is.
Re: (Score:2)
Re: (Score:2)
It won't. GPUs are being eaten up to mine Ethereum, which is very memory-bound. ASICs are trash at mining Ethereum, but GPUs have loads of memory bandwidth so they are the tool of choice.
This announcement will do the square root of jack shit to return the GPU market to the dynamics of yesteryear. If anything, it will drive down the costs of ASICs due to having another entry into the market.
You have to wonder if (Score:2)
Re: (Score:2)
Gotta wonder which node they're using for these ASICs? Are these pipecleaners for 7nm/Intel 4 like Loihi 2?
I wonder if you could use it for low energy (Score:2)
graphics rendering. Just in theory.
Re: (Score:2)
graphics rendering. Just in theory.
Use Intel for graphics rendering? What are you smoking?
Just in time for the market crash (Score:3)
Intel, Always on time
These show up in landfills (Score:5, Insightful)
It doesn't even lead to freedom. Big players can easily manipulate the price and the exchanges have shown they can control who exchanges what. There's zero upsides to crypto.
Re:These show up in landfills (Score:4, Insightful)
The importance of the exchanges shows that crypto is already failing at the goal of being decentralized. So it isn't decentralized, it isn't private, it isn't efficient, and it can't handle the scale of transactions needed to run a real economy. So what exactly is cryptocurrency good for besides scamming greater fools?
Re: (Score:2)
It's failed at all its goals, and for the same reason.
1. Enable cheap micro transactions. Nope, super expensive.
2. Decentralized currency. Nope, nobody can afford to use it for routine transactions because of #1.
3. Distributed finance. Nope, most transactions are settled by market makers (exchanges, elaborate schemes like Lightning) because #1.
Re: (Score:2)
WTF are you talking about? The mempool is empty. You can get your transaction in the next block for 5 sat/byte. And that is high these days. Expensive Bitcoin transactions hasn't been a thing for a couple of years now. Your argument is so old and so wrong that I wonder why you even making a comment for anything other than quick circlejerk karma
Re: (Score:2)
Lol. That's how badly it's failed, the bitcoin bros think that's cheap.
And that's the price *right now*. It was many times higher before, and probably will be again when transaction fees are the sole income for miners.
Re: (Score:2)
Re: (Score:2)
The importance of the exchanges shows that crypto is already failing at the goal of being decentralized. So it isn't decentralized, it isn't private, it isn't efficient, and it can't handle the scale of transactions needed to run a real economy. So what exactly is cryptocurrency good for besides scamming greater fools?
Ask Assange, who had to run his legal defence fund over BTC, because people in power decided to unperson him [bitcoin.com] and he got booted off all electronic transation systems.
Oh boy, time to spend some Kharma (Score:2)
Not that he had a lot of options (the charges against him are dodgy as fuck), but what I'm getting at here is that he was just fine with or without BTC. He doesn'
Re: (Score:2)
Assange is just a pawn at this point, He was used by the Russians to help elect their boy Trump (look it up, he held onto docs unfavorable to Trump while putting out anti-Hilary stuff before even cleaning it of innocent names, of which there were tons because of the size of the docs. That's not Journalism, a Journalist doesn't do that. Period).
I don't know if what you are saying is true or not, but that definitely is what journalists do. Journalists have a PoV and report things which support that PoV, on the whole. There are exceptions, but they're rare enough not to affect the mean position of "journalist".
Re: (Score:2)
Congratulations, you tried to tar and feather all of blockchain everywhere based on the shortcomings of Bitcoin.
Re: (Score:3, Insightful)
I'm not tarring all crypto with the problems of Bitcoin. Most of those problems are inherent to the idea of a public blockchain. Bitcoin is especially bad at efficiency and scalability, but there are limits to how much better you can get. If you try to share every financial transaction globally- which is the whole foundation of blockchains- you have an inherent problem with privacy, since anyone who wants can see the whole global transaction history. The best you can hope for is pseudonymity, which can b
Re: (Score:2)
Bullshit.
Not all blockchains suffer from such problems: e.g. there are blockchains that are efficient, can handle transactions at scale, and/or are private. Nobody's done all of them at once.
Re: (Score:2)
Re: (Score:2)
Intel owns Altera. They do FPGAs. It's not difficult for them to do ASICs.
Re: (Score:2)
Re: (Score:2)
Intel knows how to do ultra-low voltage FPGAs so.
Re: (Score:2)
Re: (Score:2)
Look on the bright side: Intel still hasn't announced any major wins for their new Foundry 2.0 initiative. Maybe they have some spare wafers to push out the door? $10 says its Intel 4 they'll be using for the ASICs. Cuz they can't fab anything else on it yet. Except maybe Loihi 2.
Re: (Score:2)
Cryptos need to move to Proof of Stake. Problem solved.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2, Funny)
It doesn't even lead to freedom. Big players can easily manipulate the price and the exchanges have shown they can control who exchanges what. There's zero upsides to crypto.
Of course, all "crypto", not just proof-of-work ones. And that, my friends, is how you can tell that CO2/ewaste/chip shortage is not the reason, just a convenient pretext for the crypto hating.
To be fair, he could just be completely ignorant. Lack of knowledge seldom stops people from acting like they have justifiable opinions.
I'm aware of proof of stake (Score:2)
It's more like a shuffle (Score:3)
Libertarianism just doesn't work. People with money and power use that money and power to get more and then pass that money and power on to their prodigy who repeat the cycle. They do this without regard to your life or mine.
I like competition just fi
Re: (Score:2)
Complete waste (Score:2)
Wasted engineering effort to produce something that doesn't actually produce anything or do anything useful, other than transfer money from one pocket to another. This indicates that Intel is going down the tubes.
Today on CryptoTurd (Score:2)
Intel Corp confirmed for eco-terrorists (Score:2, Insightful)
Re: (Score:2)
It's the lack of carbon taxes that are the big problem. There's no inherent negative externalities in bitcoin, only in certain forms of producing energy.
With carbon taxes, bitcoin tends to free market and benefits everyone.
Re: (Score:2)
This trick never works. (Score:2)
Saving energy? Never. Consider:
Candle wax expensive. Not many candles used; people go to bed early.
Along comes whale oil. Much cheaper; gets used more; people stay up later in brighter rooms.
Ditto for piped-in natural gas.
Ditto for electrifying America.
Ditto for transitioning from incandescent to LED lighting.
Reducing unit cost always leads to increased usage, not reduced spending.
Re: (Score:2)
I replaced all the standard incandescent light bulbs in the house with LEDs. With the new, higher-efficiency LED bulbs, even if I keep all the lights on, all the time, they still use less energy during 24h than a typical day with incandescents.
If the new Intel ASICs are 10x more efficient than existing ones, I don't think there will be 10x more people mining bitcoin.
Re: This trick never works. (Score:2)
If the new cards are 10 times more efficient then mining bitcoin will be 10 times more profitable. Of course more people will decide to mine if itâ(TM)s 10 times more profitable. Either that or the same people will mine 10 times more until the profitability comes back down.
Re: (Score:2)
The way I know Intel they will take all the money that they can get. This will actually be good here, it will reduce the money spent on electricity.
Re: (Score:2)
There will, however, be miners increasing their equipment 10x in such a hypothetical.
BTC perversely incentives burning energy as much as possible, because the difficulty adjusts to fill as much energy as is being contributed.
Re: (Score:2)
I replaced all the standard incandescent light bulbs in the house with LEDs. With the new, higher-efficiency LED bulbs, even if I keep all the lights on, all the time, they still use less energy during 24h than a typical day with incandescents. If the new Intel ASICs are 10x more efficient than existing ones, I don't think there will be 10x more people mining bitcoin.
The truth is somewhere in the middle. More efficient "mining" of bitcoin will result in an increase to mining operations until the price of bitcoin drops such that additional mining isn't worth the risk. Whether the new equilibrium is better or worse for the environment is hard to predict. Lightbulbs are interesting because for households we've pretty much reached peak lighting, however more efficient lighting has and will continue leading to more street, park and industrial lights being installed as runnin
Re: This trick never works. (Score:3)
Even worse, bitcoin is designed to be like this. The only constraint on bitcoin energy usage in the cost of electricity compared to the cost of a bitcoin. So the only way to reduce energy usage for bitcoin is to either raise the cost of electricity/hardware or reduce the price of a single bitcoin so it is less profitable to mine. The creator of bitcoin likely never predicted the use of dedicated rigs to mine bitcoin or calculated how much electricity would be consumed if the price of a single bitcoin re
Re: (Score:2)
No, it works when there's a bound on demand. For example, even I wanted to do so, lighting up to everything full brightness throughout my home would be less energy consumed than using some incandescent some of the time. I will never demand more lighting than my whole house all the time, there's just no point.
However, in the case of BTC mining, it's correct. The miners will consume as much energy as they can, and increasing hash rate simply makes the hashes worth less rather than lowering energy demand bec
So nice to hear (Score:3)
I'm so glad Intel is more worried about people wasting energy on mining crypto than they are about getting their chips out the door so companies can ship their products.
If they have that much excess capacity to squander on something like this, they can get their regular chips to manufacturers.
Re: (Score:2)
Re: (Score:2)
Intel actually from what I know is pretty much keeping up with demand. GPUs, AMD CPUs, and various random components are the ones coming up short.
I'm looking forward to Etherium moving to proof of stake, as that would alleviate the GPU demand significantly, which will in turn alleviate GPU consumption of the fab capacity and free up capacity for normal stuff.
Looks like Intel is desperate (Score:2)
A few years ago they would not have touched something this shady with a 10 foot pole.
Re: (Score:2)
Intel has been looking desperate for awhile now.
Re: (Score:2)
Indeed, they have. They have been trying to get back where they were and failing for a few years now. Got fat on ripping off dumb people and forgot that they are an engineering company. Now, if only MS had a real competitor...
This is a waste (Score:2, Insightful)
Great, now we can use our overloaded silicon fabs to produce something is pretty much worthless to society as a whole.
Serious question (Score:2)
Answer - why real work isn't done (Score:3)
Re: (Score:2)
Good post. Just to be completely clear for others, any class of problems that can be easily verified but has a huge search space and a predictable runtime can work. A hash is a textbook example but they could be other similar problems. One might think 'factoring primes' but the runtime is extremely unbounded. We don't get a new block until the problem is solved - it can't take a week occasionally.
Monero and Raven are a couple of projects that use a rotating problem set at each hardfork, for those interested
Re: (Score:2)
The proof of work in bitcoin is to find some random data that produces a hash that's smaller than some target value. What you're describing is the process of updating the hash tree (i.e. writing a block to the blockchain) which is what you get to do *after* you've come up with a lucky solution to the proof of work problem.
Generating the block is utterly trivial;
Re: (Score:2)
They tried with GridCoin and CureCoin. Neither one took off. Had Satoshi thought it up originally it would have worked, but by the time these projects came to light, they had been swept aside by Ethereum et al.
They went the way of most altcoins, which is into the toilet.
Re: (Score:2)
You can. There are some systems based on those. One issue is that you have a central authority issuing the problems.
Efficiency (Score:2)
An Ultra-Low-Voltage Energy-Efficient Bitcoin Mining ASIC
The intrinsic value produced per watt is still zero.
Re: (Score:3)
Do you have any idea how valuable an inherently deflationary, completely unregulated "currency" is for money laundering? Plus, if (God forbid) any country adopts them as main currency, the inherently deflationary bit will have the remarkable effect of supercharging the transfer of all wealth there to the already rich.
Ethereum (Score:2)
Except that the change to proof of stake makes them worthless soon, ASICs for mining where GPUs are used is the game changer we need. Banning proof of work is the other game changer we need, but I don't hold out hope for lawmakers to do that anytime soon.
WHY? (Score:2)
This shit is arbitary. It has no REAL value, except we blindly assign it such.
Surprise Surprise (Score:2)
Re: (Score:2)
So we are going to maintain the same energy draw...
If not greater [wikipedia.org] energy draw.