201–210 of 352 posts
Earlier quoted context omitted.
Do you think Bitmain is winning bids for fab capacity against Nvidia/AMD? I guess their newest miner is slated to be on 5nm...
They can probably pay higher prices when mining rewards are high.
Earlier quoted context omitted.
It still doesn’t help that the Ethereum hashrate limiter is not imposed on the highest-end card (it’s FHR not LHR). The 3090 (aka 30-series Titan cards) are still really hard to find. The the 24GB of double-sided GDDR6x gets super hot though and all implementations of the card did not take that into account. The 3080ti’s are easier to find as they’re LHR and price per hashrate ratio isn’t good. What I mean by easier…
Please don't buy GPU's to mine cryptocurrency, you're contributing to a global chip shortage and I really want to play Halo Infinite in 4k. For real though, cut that out.
Earlier quoted context omitted.
This nuance isn’t helpful for this discussion I think. 1.4T of overall cryptocurrency 2.2T is POW (68%). BTC and ETH alone account for 57% of all cryptocurrency market share [1]. Proof of stake is 9% of the overall market. I’m not sure where the other 23% go but I don’t think any one idea has captured a more significant piece. It’s like arguing that saying the energy sector is responsible for CO2 emissions. Like sure…
You have a point for the purpose of calculating the scale of the impact. But it is also true that a significant part of crypto does not drive demand for hardware or electric power. There are proof of stake coins with multibillion dollar valuations that bring a novel combination of smart contracts and other "platform" capabilities and low cost. This is important because these technologies ensure that crypto can surviv…
I can mint a billion PoS shitcoins and sell myself one for ten dollars and suddenly, I have a valuation of ten billion. That doesn't make it less of a shitcoin. Every PoS coin is one such shitcoin.
Earlier quoted context omitted.
If nvidia straight up just stopped selling any cards, declined all their client/customer base, and just put 100% of their global output to mining cryptocurrencies, I think it's a good chance nvidia can do a 50%+1 attack on both bitcoin and ethereum at the same time. Once they successfully do the attack, they can take over the blockchains and allocate all cryptocurrencies to themselves, cash out, and it'd probably reg…
You haven't been able to GPU mine bitcoin in like a decade, every miner uses dedicated ASICs specifically for mining btc. It's been impossible to compete without ASICs for a very long time.
Since Bitmain asics are currently being made at around 16nm nodes, using 3nm process that Nvidia has access to - think about that for a bit.
PC Hardware can come from an Era with so many manufacturers and true-actual fierce competition, resulting in complete commoditisation of components. However, now in 2022, there are only 2 CPU vendors,2 GPU vendors and even with storage the number of manufacturers is small. Only TSMC can deliver 5nm and Mr Cook has all that capacity on lock (AMD and Intel will get a look only when he is done with it) Mining like with…
I walked into a computer store last night. There were shelves full of CPUs and SSDs and HDDs, and buying several of them at once would have no one batting an eye. Yet there was a policy: only 1 graphics card per customer per month, and that applies solely to graphics cards (and prebuilt computers containing graphics cards). In other words, there is visibly a shortage of graphics cards that is not affecting any other…
Earlier quoted context omitted.
It's already been banned in various countries, including a number of US towns.
Really? I was totally not aware of that, do you have a resource?
Earlier quoted context omitted.
So your theory is that sellers weren’t greedy before?
No, my theory is that this the true core of the disscussion here should be how the economy works in general, rather than a specific case of supply and demand.
Don’t get me wrong, I love generality, but it’s clear that the trees don’t let you see the forest here.
Sellers were as greedy before the price rise, so greed is not the cause.
The difference here is that: 1) Supply difficulties mean there’s less in offer. 2) crypto miners have strong demand, since they can make money on those gpus.
I would argue that the market is working fairly well, since gpus are being allocated for their best use. Gaming is great, but people seem to really love their eth.
The article only attempts to show there's correlation, and ends up by saying "correlation doesn't imply causation, but maybe there's causation". So basically, we don't know. I'm not blaming the author, because determining causation is hard, and checking there's correlation between two variables can still be interesting, but the article falls short of answering the question.
They ran a pretty solid analysis of the subject. Not only do they look at correlation, but they regress out the trend line and show a correlation in the noise. Correlation of noise typically does imply a causal connection. They also look at correlation between GPUs that are used for mining and ones that are not and show that the correlation only holds for GPUs used in mining, so the case seems pretty tight. If you ar…
> We can see a correlation between GPU prices and the value of crypto currencies, but correlation does not mean causation. A causation is plausible in this case.
And then they go on to repeat a very reasonable argument, which I happen to believe to be true, but the kind of argument that everyone was already making before looking at any data.
The fact that the correlations do not hold as much for CPUs is good, but still not enough to stablish causation. As you said, there might be some confounding factors that affect crypto and GPUs but not CPUs.
The article only attempts to show there's correlation, and ends up by saying "correlation doesn't imply causation, but maybe there's causation". So basically, we don't know. I'm not blaming the author, because determining causation is hard, and checking there's correlation between two variables can still be interesting, but the article falls short of answering the question.
Correlation is correlated with causation, so I think this sort of exploration of a question is still valuable.