Live data from Hacker News

The Ethereum merge is done

coindesk.com

151–160 of 1001 posts

Re: The Ethereum merge is done

#151
post #90
post #22

The bottom just fell out of the used GPU market. NVidia Tesla 80 24GB on sale for US$79 in quantity. [1] [1] https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2380057.m...

Is a card like this useful to run Stable Diffusion, etc?

From what I gather you would want the M40 if anything as it is a single GPU with access to all of the 24GB for SD inference. But if you are going to do smaller you might as well get a 3060 12GB as it will be about 3x faster.

Re: The Ethereum merge is done

#152
post #91

Earlier quoted context omitted.

What's Substack? Is that new Medium?

It's like Medium but wants you to pay for newsletters.

Which is fairer imo than paying for the whole site. I prefer creator control on which content is paid and which content is not.

Re: The Ethereum merge is done

#153
post #18

Good. The merge worked and they are not burning up the planet. Now when are those Deep Learning systems in these data centers going to stop incinerating the planet with their useless deep learning models that not only are broken but require constant retraining and wasteful CO2 energy usage for years without any efficient alternatives?

All datacenters and network infrastructure in the world account for 2,5% at most (I don't remember the source anymore, I read this months ago) so I can't imagine machine learning is more than maybe 0,5% total.

So instead of going after actually useful technology that consumes a tiny fraction, how about you look at some of the truly unnecessary users that consume much more than that. What if we added up all the electricity used by the entire supply chains of all the smartphones people throw away prematurely due to planned obsolescence? Or all the other unrepairable disposable devices? How about all the clothing that's made of increasingly shittier materials and often needs to be replaced every two years? Some estimates for the clothing industry are as high as 10% of global GHG emission, primarily from fossil fuel power plants. And don't get me started on all the cars that wouldn't need to be produced if we fixed public transport and americans stopped building cities like idiots.

The only reason people go after the IT sector is that it's easy to stick a current probe on the wire and get a relatively large number. Turning it off is simple and makes the number, however small, go down. All the actual big users are complex systems that are difficult to analyze and "turn off".

Re: The Ethereum merge is done

#154

Earlier quoted context omitted.

I wouldnt worry, the whole system up to this point has used "technobabble" as a means to confuse and impress outsiders. When reading up on it, there is no meaning to find besides "yep, its a linked list allright".

I could say the same thing about reading fields I don't generally understand, and it can seem like "technobabble" because I don't understand the meaning of words they are using, since some things are written with a certain audience in mind that possesses the knowledge to understand the content, like many academic papers. However, I don't regularly dismiss fields like that, but rather I understand that not everything…

Genuine research states claims for the methods and discovery, making it often quite easy to work backwards from the conclusions to the theory. No such logic seems to exist in the crypto culture.

Re: The Ethereum merge is done

#155

There are two interesting things I want to watch from this. The first is I'm interested to see what kind of bull run ETH goes on. The merge has been incredibly long coming, it has huge risks and I think that puts downward pressure on price, you really don't want to be doing stuff in ETH at the moment because there's a fairly good chance something goes wrong, someone stealds $XBn and runs off and the Ethereum guys go…

>So as that risk dissipates I would expect a decent price run. The risk you're describing is a social-political one. As long as the users do not fork with the etherium developers, that will not happen. Change to PoS has nothing to do with it. I look forward to the increased GPU supply.

> The risk you're describing is a social-political one. As long as the users do not fork with the etherium developers, that will not happen. Change to PoS has nothing to do with it.

Are you saying that the market did not price in any technical risk? I don’t follow crypto markets closely (and don’t own any), but given the number of crypto bug exploits and the unprecedented nature of this merge, that seems unlikely to me.

Even a social-political risk averted is a risk averted.

Re: The Ethereum merge is done

#156
post #146

For those interested in understanding the tech rather than the typical bashing things as beneath them, I wrote up a detailed technical explainer of how Ethereum PoS works: https://0xfoobar.substack.com/p/ethereum-proof-of-stake

Isn't it a little unfortunate that the acronym for Proof of Stake would be PoS? Couldn't they think of something else?

You mean people might confuse it with Point of Sale? /s

Re: The Ethereum merge is done

#159
post #35

The validator of the first proof of stake block earned just over 45 ETH as everyone clamored to get their transaction in this historic block: https://etherscan.io/block/15537394

The validator is randomly chosen, no? So we essentially have a lottery as a banking system now?

There's a new block about every 13 seconds. Each validator will get its share of "wins" over time.
Post reply on HN