According to Billy Finn, the noise is actually getting louder. He’s been tracking the decibel levels on Louise Avenue since 2022, when Hyperscale Data began operating. Back then, the sound level was around 52 decibels. Today, they’re typically around 61 decibels, and sometimes go as high as 78 decibels, he told the paper.
Inside his house, it goes down to 39 decibels, which is about the level of a quiet office or library, according to the American Academy of Audiology. But it jumps to 62 decibels when he opens the door, sounding a bit like a passenger jet taxiing on the runway in the distance.
And that’s 24/7/365.
Remember that the next time you “ask ChatGPT” instead of making a web search.
Decibel scale is logarithmic, so down at around the 60 level, the difference between 30 and 60 is basically the difference between a normal conversation and someone whispering.
…sounding a bit like a passenger jet taxiing on the runway in the distance.
That could be because they use turbine generators for power when/if the local grid can’t support what they want. The major difference between the turbines used in aircraft and the ones for datacenters is the load. Giant fans for planes, generators for datacenters.
Please ignore the obvious onesidedness of the link below. Its there to show I’m not making this ip. It’s shitty that this is allowed and they aim to siphon off as much money as they can.
There is a difference between using X amount of resource to provide an actual output that justify spending X amount of resource, and using X*1000 resources to provide zilch.
Citation on it being zilch because they do a fair bit of actually useful work these days while the search engines mostly seem to be going downhill (Kagi being an exception)
It does not use AI data centers. Those are not remotely on the same scale as what we called a data center before this LLM hype started. This infrastructure will not have another use once the bubble pops.
It doesn’t need nearly as many. AI inference is orders of magnitude more expensive than a single search query (ignoring the fact that Google does it’s own inference with search queries now). And that doesn’t even include training, which is stupidly expensive to do.
The sound: https://www.tiktok.com/@ayathetigress/video/7650601803972627726
And that’s 24/7/365.
Remember that the next time you “ask ChatGPT” instead of making a web search.
He must have a well-insulated house if opening and closing the door makes a 23 decibel difference.
Too bad he can’t open his windows, though. Or enjoy his yard.
Can’t wait for the backlash to tear these datacenters down, or the economic cliff to shut down their operations.
Not at the range of things under 70 decibels.
Decibel scale is logarithmic, so down at around the 60 level, the difference between 30 and 60 is basically the difference between a normal conversation and someone whispering.
Maybe I just lived in a poorly insulated house, because I swear I could hear outside as plainly with the windows closed as if they were open.
It was an old house, mind. Also drafty…
That could be because they use turbine generators for power when/if the local grid can’t support what they want. The major difference between the turbines used in aircraft and the ones for datacenters is the load. Giant fans for planes, generators for datacenters.
Please ignore the obvious onesidedness of the link below. Its there to show I’m not making this ip. It’s shitty that this is allowed and they aim to siphon off as much money as they can.
https://www.greengasturbines.com/blog/gas-turbines-for-data-centers-hyperscaler-power
“green” gas turbines, what the fuck?
Mmm, nice clean diesel 🥴
Which is illegal according to EPA regulations, last I knew.
Yeah, here we are.
Are you under the impression that a web search does not use data centers?
Google in particular has a lot of infra to support their search, which actually used to be good
There is a difference between using X amount of resource to provide an actual output that justify spending X amount of resource, and using X*1000 resources to provide zilch.
Citation on it being zilch because they do a fair bit of actually useful work these days while the search engines mostly seem to be going downhill (Kagi being an exception)
It does not use AI data centers. Those are not remotely on the same scale as what we called a data center before this LLM hype started. This infrastructure will not have another use once the bubble pops.
It was said that one was built in 2022. That’s long before they started doing gigawatt scale bullshit.
It doesn’t need nearly as many. AI inference is orders of magnitude more expensive than a single search query (ignoring the fact that Google does it’s own inference with search queries now). And that doesn’t even include training, which is stupidly expensive to do.
Sure. But let’s not pretend Web search is innocent here. Wanna be eco friendly, walk to your local library.
Using this a lot lately, huh
derail