Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mining Ethereum on M1 Mac GPU

blog.yifangu.com

Post reply on HN
Print view View original

121–129 of 129 posts

Re: Mining Ethereum on M1 Mac GPU

#121

Post by panarky »

xyzzyz wrote: post #45
panarky wrote: post #42

Earlier quoted context omitted.

Are you implying that Germany's electric grid is unreliable? If so, a citation would be good because my sources say Germany's electric supply has been stable.

See for example https://amp.dw.com/en/winds-of-change-push-german-power-grid... or https://www.greentechmedia.com/amp/article/germanys-stressed... .

Neither of these articles says Germany's electricity supply is unreliable. There's no mention of power outages, or equipment damage, or anything like the power crisis in Texas power.

Instead, your citations say renewables are producing more energy than can be used.

That's a good problem. It can be solved by turning some of the supply off, or adding long-distance transmission to move the excess power to where it's needed.

But there's no hint of renewables causing instability or unreliability or power outages.

Re: Mining Ethereum on M1 Mac GPU

#122

Post by stanmancan »

nimbleal wrote: post #113
carlob wrote: post #73

Earlier quoted context omitted.

hydro is really not that green: salmons are losing their spawning runs for an example in your area. Also not renewable, because dams get filled with sediment and lose capacity over time.

Depending on the impounded area they can (in some cases) be worse than fossil fuels for greenhouse gases. https://www.scientificamerican.com/podcast/episode/not-all-h...

I wish that went into more detail! I have no clue how it could be worse than fossil fuels. I get that trees are a store of carbon and if they get submerged they will break down and release it all, but surely over the lifespan of the dam that’s significantly lower than burning coal non-stop? Wouldn’t it be equivalent to a forest fire clearing the area?

I always assumed they would clear the area of trees first anyways. Why let the wood rot when you can cut it down and use it?

Re: Mining Ethereum on M1 Mac GPU

#123

Post by stanmancan »

carlob wrote: post #73
stanmancan wrote: post #44

Earlier quoted context omitted.

We pay about $0.10 in British Columbia, Canada. There is a $0.20 daily base rate and over a certain amount it steps up to $0.14. It’s 95% hydro electric so it’s green too. https://app.bchydro.com/accounts-billing/rates-energy-use/el...

hydro is really not that green: salmons are losing their spawning runs for an example in your area. Also not renewable, because dams get filled with sediment and lose capacity over time.

While it’s not perfect it’s significantly better than burning coal.

Geographically we can’t rely on solar; I’m not sure why we don’t have more wind turbines!

Re: Mining Ethereum on M1 Mac GPU

#124

Post by jmkd »

belval wrote: post #109
jmkd wrote: post #91

Earlier quoted context omitted.

Dams all over the world capture sediment. This not only reduces their capacity, it deprives downstream areas of sediment needed for natural habitat, to fertilise wetlands and to buffer against oceanic incursion. The Sanmenxia Dam in China lost 17% of its capacity in the first 18 months of its operation due to sediment accumulation [1] Countless other dams have already been dismantled due to sediment build up, and man…

Honestly this sounds like mostly poor engineering planification and analysis. If you look at the dams with a lot of sediment it's mostly in South-East Asia and Africa while Europe/North America have little of it.

Asia's rivers have by far the highest level of sediment of any continent [1] meanwhile the US has quietly destroyed over 1000 dysfunctional dams in the post war period [2]

[1] https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1029/WR0...

[2] https://www.pbs.org/wgbh/nova/article/dam-removals/

Re: Mining Ethereum on M1 Mac GPU

#126

Post by OldHand2018 »

ianai wrote: post #110
OldHand2018 wrote: post #28

Earlier quoted context omitted.

I mined Ethereum about 3 years ago using the Nvidia 1060 and they never drew max power. Conversely, you could use nvidia_smi to limit their power draw without making much of an impact on hash rate. I’d be surprised if it was any different with the current lineup of cards. It would be really interesting to see what the M1 draws doing this. Probably not much. But then it’s not doing a whole lot of hashing either.

Was there any critical reason you stopped mining?

I didn’t like the noise and the price had collapsed. I sold the cards for a very good price. I would have made 5 or 6 figures if I had just left them on. Oh well, it’s not the worst financial decision I’ve ever made. Hell, I bought TSLA at $15 and sold at $30. F me.

Re: Mining Ethereum on M1 Mac GPU

#127

Post by nimbleal »

stanmancan wrote: post #122
nimbleal wrote: post #113

Earlier quoted context omitted.

Depending on the impounded area they can (in some cases) be worse than fossil fuels for greenhouse gases. https://www.scientificamerican.com/podcast/episode/not-all-h...

I wish that went into more detail! I have no clue how it could be worse than fossil fuels. I get that trees are a store of carbon and if they get submerged they will break down and release it all, but surely over the lifespan of the dam that’s significantly lower than burning coal non-stop? Wouldn’t it be equivalent to a forest fire clearing the area? I always assumed they would clear the area of trees first anyways.…

> I always assumed they would clear the area of trees first anyways. Why let the wood rot when you can cut it down and use it?

The spaces are often so vast that chopping down the wood could delay a project by impractical lengths of time.

> I get that trees are a store of carbon and if they get submerged they will break down and release it all, but surely over the lifespan of the dam that’s significantly lower than burning coal non-stop? Wouldn’t it be equivalent to a forest fire clearing the area?

My understanding is that microbial breakdown means a lot more carbon is released as methane vs. CO2 from fires — methane being a much more potent greenhouse gas.

Here’s one related paper I found

https://academic.oup.com/bioscience/article/66/11/949/275427...

(I work for a hydro developer but I’m not an expert).

Re: Mining Ethereum on M1 Mac GPU

#128

Post by chillacy »

Scene_Cast2 wrote: post #54

No-one has mentioned that ETH mining is memory bandwidth bound. M1 macs, unfortunately, have slow GPU memory compared to your discrete GPU's GDDR6X. Here's a great post showing that - https://www.vijaypradeep.com/blog/2017-04-28-ethereums-memor... - GPU mining is only 10% slower than the theoretical memory-bound maximum (if compute was infinitely fast).

Shouldn’t it be better to mine RandomX using the CPU?

Re: Mining Ethereum on M1 Mac GPU

#129

Post by krebspsycho »

chillacy wrote: post #128
Scene_Cast2 wrote: post #54

No-one has mentioned that ETH mining is memory bandwidth bound. M1 macs, unfortunately, have slow GPU memory compared to your discrete GPU's GDDR6X. Here's a great post showing that - https://www.vijaypradeep.com/blog/2017-04-28-ethereums-memor... - GPU mining is only 10% slower than the theoretical memory-bound maximum (if compute was infinitely fast).

Shouldn’t it be better to mine RandomX using the CPU?

Generally yes. The RandomX dev seems to have some issues getting to to work well on apple silicon, however the Xmrig dev has it working, but seems it caps out at 2.3kh/s due to not being able to take advantage of JIT or large pages. (At least from what I read of their respective GitHub repos https://github.com/tevador/RandomX/issues/196

And

https://github.com/xmrig/xmrig/issues/1991).

RandomX mining is better at the moment, but that may be because nobody's developed a GPU minter tailored to the uniqueness of apple's GPU, if that's even possible.

Post reply on HN