While these numbers are sort of interesting, they are misleading, because the 1.18 cents/search number that the author provides is not actually the cost of a single search.
A more useful analysis would be to estimate the marginal cost of each additional search (e.g. if I go make a search right now, how much extra does that cost Google?).
Likewise, there is a similar number that represents the marginal cost of each additional webpage that Google indexes.
Power consumption will vary based on usage on modern machines and Google could do automatic boot/shutdown of machines to match demand. But the marginal cost in electricity alone should still be pretty small.
According to this link someone else found [1], they claim to use 0.0003 kWh electricity per amortized search, which at e.g. 0.06 USD/kWh (typical [2]) is 1.8*10^-5 USD, or 0.26% of search costs (according to the estimates in the article).
(In a related estimate, 0.0003 kWh/search means 54 CPU-core-seconds of indexing per search, assuming (without any justification and probably wrongly) that (i) most electricity consumption is by CPUs during indexing and (ii) they are using x86 machines with about 20W/core power consumption. I'm sure someone else here can fix/improve this estimate.)
There has been a form of innovation - in some server farms that consist of thousands of interchangeable computers, they simply don't try to cool them, and instead just exchange air with the outdoors (possibly with a bit of filtering). If your software doesn't mind having nodes frequently drop out due to overheating, then it apparently is cheaper to replace melted CPUs than to keep them all cool.
Which is also why more data centers, including Google data centers, are being built in cool climates. If you're exchanging for cold outside air, you get to keep more of your servers from melting down without paying more for cooling.
I dont think it works that way - in centralised cooling spaces, they have to keep their cooling machines running. So, it costs the same whether you are generating that amount of heat or not.
To cut down on cooling, you need to architect your farm to have smaller, independent cooling systems. However, your system must also be architected such that weird things like "only machine is active in my cooling cluster, so I cant shut down the cooling for that area" cannot happen.
I'm not sure if things like that have been solved.
Correct, HVAC systems are designed to handle variable capacity and to stand down as demand lessens in the same way as a data center is. It's such long standing practice that most recent homes even have variable speed compressors on the condensing unit and multi-speed fans on the air handler.
I think pedrocr means that the baseline would be moved down, so that less cooling in total would be required, not that they reduce the amount of cooling used during usage dips.
The calculations are quite tricky given the granularity of data the author is working with. I agree that a marginal cost calulation is a bit more useful as the cost of development staff is not really relevant here as this is an R&D cost. The exception being those employees that work on technologies to sustain the growth of pages and searches and the IT staff to maintain the infrastructure. Without management accounting data the figures will be rough estimates at best.
Well, you would need to calculate the marginal cost if you were really interested in what a particular search costs, but I think this number is actually says much more.
It is the amount that Google is spending on average on a search and thus it is the amount that Google has to earn back on average per search in order to be profitable.
Also the 0.68 cents is cost of search-related revenues divided by number of searches. However, presumably there are only revenues when someone clicks on an ad on the search results page. So a more accurate number would also factor in the ratio of ad clicks to searches. If there's one ad click per 10 pages on average, these numbers are off by an order of magnitude.
"A more useful analysis would be to estimate the marginal cost of each additional search (e.g. if I go make a search right now, how much extra does that cost Google?)."
Interesting, but tricky. Is the price of the new servers they need to buy for the peak processing marginal or fixed cost?
I agree in principle, but this fraction would be easy to get if they just had to buy a server after a fixed number of clicks. But they only have to if the peak usage grows. Peak usage grows, they buy some servers, then go on for some time processing huge amounts of clicks. Things might get even more complicated if they decide to spend the (now) excess processor power on something else.
Having read this I searched around for other figures on how many searches Google serves each month, and while I didn't find any better sources, I did come across a Google blog entry [1] about the environmental cost of searches.
It interested me to know that, based on the average MPG of cars in the US [2] and a rough cost of a gallon of petrol [3], it would cost 1.6% of a GBP penny, or 1/40th of a USD cent, to emit the same amount of CO2 in a car as a single Google search (0.2 grams of CO2 [1]).
I'm not sure why I felt the above was interesting enough to write in a comment, but for some reason it does interest me that, if Bill's calculations are correct, it costs Google 27 times as much per gram of CO2 emitted as it does drivers (excluding car costs etc.) Even assuming I'm showing signs of insanity chosing to work that out, the amount of CO2 emitted is good to know.
That's a very significant piece of the emissions footprint of a car. I've seen numbers that say as much as 30% of the emissions resulting from a car come from the car's construction and distribution.
I have seen numbers that suggest it costs about as much to keep our road system maintained as the fuel costs to drive on it. Basically add up the cost of bridge / road matience, deicing, DMV, highway police, truck weigh stations, drivers ed, and new roads and you find our highway system is heavily subsidized.
While it's true that all that costs money and creates CO2, as do the cars themselves (as lukeschlather commented), I suspect that Google aren't taking into account the cost and the CO2 of the infrastructure of the internet, that actually allows people to search through them.
Should we call this estimate: 'cost of Google as measured by searches'? Similarly, you can divide the cost by number of Google employees and call it 'cost of Google search per employee'?
The total cost (including revenue from advertising) is negative, otherwise google wouldn't be profitable.
Anyway, I found his transparency and links to the sources way more interesting than the actual result (which I somehow still doubt on an intuitive level)
I'm sure that's right. There are more eyeballs doing searches on Google, and there is an auction between advertisers pay to put their ads in front of users. More advertisers chasing users probably gives them more revenue per search.
A user is a user is a user. Unless google users are that much more likely to buy something after clicking my ad, why would you pay more per individual view? Sure the absolute numbers should go up with more views, but it doesn't seem like money well spent to pay more for similar adspace.
That would assume that people using Adwords compare costs with the other ad networks before deciding to advertise. I would guess this is most often not the case.
i think the issue here is that a search is one service payed by another (adwords) and so on.
for the cost of a search you have to devide all employees into their group and take out only the cost for searches. but than comes adwords and pay for it...
tough the adwords code probably runs on the same servers that deliver searches (unified system)... i guess its really hard to cslculate but i would guess a search is only a fraction of a cent even before balancing it out win profits.
A more useful analysis would be to estimate the marginal cost of each additional search (e.g. if I go make a search right now, how much extra does that cost Google?).
Likewise, there is a similar number that represents the marginal cost of each additional webpage that Google indexes.