Live data from Hacker News

Microsoft has sunk a data centre in the sea to investigate energy efficiency

bbc.com

261–270 of 397 posts

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#261
post #248

I'm sure the environmental impact of this one unit will be negligible, but if the worlds largest tech companies start deploying thousands of these - imaginable, since it's a pretty good idea from an engineering perspective - I could see that causing some real problems.

> I could see that causing some real problems Can you be more specific? What type of problems and why? Temperature-related, waste related, hazardous substances? And why would this become a problem at-scale?

  Prof Ian Bitterlin, a data centre consultant for nearly 30
  years, is sceptical about the environmental impact of going
  underwater.
  
  "You just end up with a warmer sea and bigger fish," he says.
The reason I think it could become a problem at-scale is, again, because it's a great idea from an engineering perspective. Cheap land, cheap cooling, space-efficient (no need for human amenities in the building, like walking space and bathrooms and security checkpoints). I could imagine a scenario a decade or two down the road when most of the world's data centers are dumped off the coasts, at which point there would likely be impact. Setting aside the heat, there's the widespread disruption of the sea floor. Also, while some companies might take environmental precautions, I'm sure that others would not. Why would you bother spending the money to salvage a pod in the case of a tank breach or a fire?

It's far from the biggest environmental concern we have, and it could be done with minimal impact if those doing it cared enough, but in my experience large companies won't go to any expense that doesn't make or save them money.

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#262
post #55
post #36

At what point does this become a problem? They're using the naturally cool sea water to regulate the temperature in the container, but that heat needs to go somewhere... it heats the water. If this becomes a popular alternative to classic server farms, at what point does it start to increase the temperature of the sea? I know, the oceans are big, but so is our data.

All the big power plants use rivers for cooling. It doesn't seem to be a terribly big problem unless there is a drought.

Thermal pollution can cause significant ecological impact [1] when left unchecked; however, thermal discharge is regulated under the Clean Water Act [2], hence the usage of cooling ponds [3] and towers [4].

[1] https://en.m.wikipedia.org/wiki/Thermal_pollution

[2] https://www3.epa.gov/region1/npdes/merrimackstation/pdfs/ar/...

[3] https://en.m.wikipedia.org/wiki/Cooling_pond

[4] https://en.m.wikipedia.org/wiki/Cooling_tower

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#263

Earlier quoted context omitted.

I think part of it is that even problems that are small/unnoticeable at the individual level become big when multiplied by a huge N. For example, I don't think I've ever had a hard drive fail in ~30 years in computing, well without being dropped that is, yet look at the Backblaze reports: https://www.backblaze.com/blog/hard-drive-benchmark-stats-20...

My guess is that you change disks often. I've had multiple fails over the years, either from overheating or more often from overuse for extended periods of time (aka more than five years). Although I've never had a failure before the guarantee expired.

Yesterday I lost 3TB Toshiba TD01* series drive 23,75 months in (home) use. Only 4800 hours powered on. And it started to seek fail, even SMART showed it. It failed one week before warranty ended. Ha!

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#264
post #242

Well everyone is talking about the energy usage, which is the point of the experiment. But what happens after 5 years (the expected life of the datacenter). Most of the computers in there will be worthless then. Will they bring up the datacenter and reload it with new equipment? Will the cost-benefit be in favor of re-equipment or just sink a new one in? If its going to be cheaper to just drop a new one in, we will h…

Even if the equipment is outdated and broken, the scrap might be worth something to someone.

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#265
post #156
post #144

Earlier quoted context omitted.

It works because of how hard drives work - the heads are extended by slightly heating the bar the heads are connected to. If you're head crashing, you need a little less length, hence freezer.

This isn't quite true, the heads are kept from crashing by air pressure and a crash does irreversible damage. The "clicking" sound is either the heads flicking back and forth as the drive tries to seek, or the central motor trying and failing to start platters spinning which are jammed on the bearing. What freezing the drive may achieve is either un-sticking it from the bearings (see also the "bang drive on edge" tec…

While they may make legitimate points in the article, it's hard to consider someone that makes money on HDD recovery like the authors of the article having an unbiased viewpoint here.

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#266
post #198

Earlier quoted context omitted.

Naturally scarce stores of value like Bitcoin, gold, diamonds, etc. have always "wasted" resources, whether they be power, human resource, or polluting the environment. On one hand you could regard them as immoral because of this, on the other hand they enable the freedom to store value without trust.

At this point, we know that Bitcoin doesn't scale to handle the volume of transactions needed for a global digital currency. Ultimately, if you own Bitcoin in hopes of the value going up, it will only go up if someone is willing to pay more for it. There's no other utilitarian use of it. This means that Bitcoin is a 0-sum game*, and a tremendous waste of resources. At least with gold mining we end up with a material…

I'm not a huge fan of bitcoin and have never owned more than 200 dollars worth, but that's a super weird argument to make.

Of course something only goes up in value if someone is willing to pay more for it, that's true of art, cars, houses, or whatever. You could argue that a house has more utilitarian value, as you could always live there, but that completely ignores the other persons goals. A house isn't going to be much value if jobs are scarce in that area, but bitcoin would still have value in that scenario.

All that energy is going to make sure you can send a large amount of value between two parties from almost anywhere in the globe with a minimum of transaction costs and a decent amount of anonymity.

This also ignores the fact of risk pricing. I live in Backwatermenistan and the political stability and the currency is looking unstable. I could buy a bitcoin from someone else who has one with local currency, I lower my risk of ruin from currency inflation by paying a price that I deem reasonable, the other guy gets cash which he could use to start a business or anything else that's inside his risk tolerance. No one lost in that scenario.

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#267
post #36

At what point does this become a problem? They're using the naturally cool sea water to regulate the temperature in the container, but that heat needs to go somewhere... it heats the water. If this becomes a popular alternative to classic server farms, at what point does it start to increase the temperature of the sea? I know, the oceans are big, but so is our data.

Wouldn't it be much smarter to recycle that heat energy? "Our goal is to recover and reuse all the heat produced." - https://www.telia.fi/yrityksille/english/telia-helsinki-data...

Re: Microsoft has sunk a data centre in the sea to investigate energy efficiency

#268
I wonder why they are submerging these instead of just making them a floating barge.

I can't imagine submersion is significantly better for cooling, the energy density very likely already requires a circulating water system, and the added issues dealing with the pressure at depth seem risky.

Possibly there is a benefit not related to cooling. Perhaps weather/waves? Things are probably alot calmer 30ft below the surface. The tossing and turning on the surface may place additional streses on things like HDD spidles.

Post reply on HN