You must log in or register to comment.
The more I think about it, the less sense this graphic has
- If not per sq km, it should at least be per capita
just checking on wikipedia, divided by area GB should have bar around twice high as Germany. 209k m^2 vs 357k m^2 - and what does it mean 1 datacenter in the first place?
big as a city sprawling datacenter complex and a bunch of racks in the cellar both count as 1?
If you follow the trail you get Cloudscene as a source: https://cloudscene.com/region/datacenters-in-north-america
They seem to be some cloud services marketplaces, where they link up buyers and sellers. I suspect it only lists the data centers that they have listed that are included in the graphic. That would make a lot of sense, since Chinese data centers used to service people in China are unlikely to be listed, which is why it says in all of China there are only 300 data centers.
Yeah, this dataset seems very incomplete/limited. I’d also argue that the US probably doesn’t have over 5000, as many of these vendors have their “own DC” that’s just hoteled inside the same giant multi-building complex.
- If not per sq km, it should at least be per capita
Personally I think it’d be interesting to see this per capita, so here’s my back of a napkin math for data centers per 1 million pop (c. 2022):
- NL - 16.78
- US - 16.15
- AU - 11.72
- CA - 8.63
- GB - 7.68
- DE - 6.22
- FR - 4.63
- JP - 1.75
- RU - 1.74
- CN - 0.32
Worth noting of course that this only lists the quantity of discrete data centers and says nothing about the capacity of those data centers. I think it’d be really interesting to break down total compute power and total storage by country and by population.
I’d also be interested to know what qualifies as a “data center”? For example, are ASIC based crypto mining operations counted, even though their machinery cannot be repurposed to any other function? That would certainly account for a chunk of the the US (almost all of it in Texas).
Yeah, that’s the proper way to think about it. And honestly, it should be servers or racks per capita (i.e. some standard unit), not just “datacenters,” since those can be of varying size.
I would really want a measure of actual compute power, like teraflops per capita or something. Still imperfect, but better than just counting the number of buildings.
Yeah, almost any metric is more useful than this one, and I’m an American who “benefits” from this stat.
I think it is was supposed to come across as anti American not beneficial. People have been knocking data centers for their overuse of energy, thereby trying to say the U.S. is using mass amounts of energy. They just wanted to make a chart that showed the U.S. having way more of these “problem facilities” without any more thorough information.
I read it as the opposite. There’s a certain amount of processing and storage the world needs, and if the majority of that is happening in the US, that’s better for the US due to more jobs and more oversight by law enforcement (FBI, NSA, etc). It’s also indicative that the tech industry is stronger in the US than elsewhere, which can be good for investors.
That said, it’s a completely worthless if it’s just an indicator of population density among developed nations. It should be something based on per-capita use of/access to datacenters.
OK, interesting. I’m a little unclear on how they’re calculating rMax and rPeak though?
🤷
The first thing i found while looking for processing power / population. Then i got too lazy to calculate.
Rpeak values are calculated using the advertised clock rate of the CPU. For the efficiency of the systems you should take into account the Turbo CPU clock rate where it applies
I’m gonna speculate, but I don’t think these systems can run all cpu cores on turbo due to power and thermal limitations and because that wouldn’t be good cost/processing power wise (since you need excessive cooling to do that). Rather, they fire turbo on groups of CPUs, allowing the CPUs to cool down till the sequence wraps around
So, I think rPeak is the processing power achieved when the computer is turboing a large chunk of CPUs
Sorry for my messy writing, I’m a little tired
A graph of power used for data centres per capita would be interesting, and I’d think Iceland would be pretty high up that one.
Surprised that Ireland isn’t on this list. Surely they need loads of data centres for all the US companies they shield from taxes 😉.
Germany only 521? Seems a bit low.
What counts as a “data center”? How many rooms and how many racks does it need to have?
I believe it needs to be a building that is only dedicated to servers.
Not an office building or factory with a server room attached.
In 2020 there have been around 3000 data centers in Germany. Sounds more plausible to me.
Why? Because data protection and stuff is so good here? Wouldn’t that apply to any other EU country too?
Surprised china has few
If you follow the source trail it lists Cloudscene as the source, who seem to be some marketplace for buying and selling cloud services. I highly suspect it’s a count of the data centers they have listed by their sellers, which would bias the US and explain why there are so few for China.
Could be they have way more but don’t tell us about them
Or the have fewer but larger.
Or as another user pointed out: The source might be questionable to use as valid data globally.
Could be fewer but larger. China likes megastructures. Could just be that their compute power is much more heavily consolidated.
Yeah, same, I found it a bit surprising they have fewer than Germany
They also may be much larger data centers than in other countries. This source is just a total count of individual data centers.
I’m shocked the UK is as high as it is. Land costs a huge amount here, as does energy (highest in Europe).
Then again, the UK has an unusually large services sector after Thatcher basically decided to kill the manufacturing sector, and the UK is probably the IT leader in Europe, so I guess it has that going for it.
China being so ridiculously low has me questioning the data though…
The UK’s average energy price is high, but it’s also very variable as when it’s cloudy and calm 20% of demand needs to be imported from France/Norway so wholesale energy is very expensive, but when it’s sunny and windy wholesale energy is free or even at negative cost and 20% of generation gets exported to France/Norway, where their energy is more expensive
If you have the option to run datacentres at minimal or even negative energy cost maybe 20% of the time, then shift load elsewhere the rest of the time, then that may be a reasonable proposition
Surprised Ireland didn’t even make it to this ranking.
