I heard a discussion like this a while back only lightbulbs were the whippingboy. That ended abruptly when I asked how many of them left their computers running 24/7, unplugged their Xbox or anything with a vampire light, etc.
Yeah, it's called "fetish"; people get provided with fetishes they are supposed worry about, right here right now. And this is entirely disconnected from rational thinking; it is just ideology. In practice, things still work like the witchburns worked.
I'm wondering why indeed nobody seems to consider energy consumption of a desktop machine - all discussion is about laptop, and is only about battery consumption.
I love to run my machines 24/7 (or a least 16/7 - and anyway, when the thing is reachable from outside for smtp/http, it should actually be kept running), but then I try to care to have somehow a little more efficient hardware and config.
For instance, CPU: even a T-model (i5-3570T in this case) will continuously eat some 18W for the uncore:
Code:
Intel(R) Core(TM) i5-3570T CPU @ 2.30GHz
(Arch: Ivy Bridge, Limit: 36W)
18.92W [===================================> ]
Package: Uncore: x86 Cores: GPU:
Current: 18.92W Current: 18.20W Current: 0.72W Current: 0.00W
I was told this would be fixed in Haswell/Broadwell series, so now Im looking for one of those.
Other issue: disk spindown.
Many things (like databases, etc) are used only part-time. With a proper disk layout the disks could be stopped for most of the time.
But there it gets weird: disks have features like idle/standby/sleep and APM (these are two different things!), but they are highly individual per each drive model (and sparsely, or not at all, documented) - and some decision is necessary, because spindown goes against the disk-lifetime also (something between 10'000 and 300'000 is usually supported). So, I conclude: almost nobody is actually using these, otherwise they were better documented.
But the highlight, in that regard, is Mr. Tom Lane, chief developer of the postgres database (and certainly a very good engineer, otherwise). Postgres does continuously access the filesystem every five or ten minutes, for no reason at all. Or, more precisely, the reason given by the developers is: people might accidentially swap the disk with the database files, while keeping the database program running, and that might lead to data corruption, so the program checks every few minutes that the files are still there.
So, consequentially, for a database that is only used during business hours, disks will never spin down even if configured to do so. When I asked the developers if they really intend to achive this result, Mr. Tom Lane explained that given the decision to either protect the people from data corruption or otherwose support some moron who wants to spindown their disk, it is clear he must protect the people.
So, I don't know exactly how many sqare miles of rainforest Tom Lane kills per day, but given the popularity of postgres, and the fact that some disk manufacturers (e.g. Seagate) nowadays configure disks for auto-spindown already, I assume it figures to quite an amount...
So, thats what I mean with "fetish". "Environment" is a fetish, "CO2" is a fetish", "energy saving" is a fetish. People carry these fetishes around and think them important. But they don't spend a thought on what that actually would mean if one would apply some consequential, rational thinking. (Indeed, rational thinking is nowadays considered very nazi.) Instead, these fetishes are used for propaganda: to promote things where somebody can make real big money from: LED lights for houses, lithium batteries for cars, etc.
Carbon based lifeforms should have a carbon footprint. Where is the flaw in that logic?
Indeed, they do. And Maturana+Varela (famous systems-theory scientists) have shown that self organizing systems (which nature/creation certainly is) can only exist in a context of energy abundance.
But what is also true, is that this planet can only use the energy that it receives from that nice atomic reactor up in the skies - and that is a certain, very specific amount per day. (Otherwise you would need to do fusion or fission, and how that could be done cleanly has yet to be shown.)
So, I think it was already 50 years ago, I said that it is not a wise idea to burn all the oil that is fetched out of the ground - because that is something like a bank-account, an inheritance conserved over millions of years, and burning it up just because it's there, within only a few decads, is certainly very stupid.
But nowadays people come up with the climate-lie, and that is again a fetish. And yes, I say, 'climate-lie', because it is bullshit - it's just a consequential damage. If people would have listened to what I was saying 50 years ago, that problem would not have appeared.
But then, again, people do not consider to change behaviour and fix the problem. They just decide: if it does not work out to burn all the oil for sports, then lets instead kill all the lithium for sports! And that will turn out a lot worse, because that stuff does not grow in woods at all. We will have to fetch it from the asteroid belt later on.
optimuspc.com
