I.E., mega-, giga-, tera-, etc. I believe we’re currently up to peta (10[sup]15[/sup]) to describe the number of operations per second of a planned supercomputer. Current the largest prefix is yotta = 10[sup]24[/sup], comparable to Avogadro’s number or the number of individual atoms in a modest sample of matter. Will we ever need a bigger prefix to describe computing power?
Yes. But I can’t talk about it without risking irreparable damage to your future timeline.
Well, if we accept Moore’s Law (or some variant of it, since there is actually some question just how fast computer power increases), then the number of operations per second increases by a factor of 2 every two years, which is equivalent to an increase of about 1000 every twenty years. That means if we are talking about 10^15 operations a second today we’ll be talking about 10^18 operations a second in 2031, 10^21 operations a second in 2051, 10^24 operations a second in 2071, 10^27 operations a second in 2091, etc. Now you’re going to say, “But does anyone really know that computer power is going to increase that fast?” Of course no one really knows. We’re not psychics. All we can do is look at the graph of increasing computer power over the history of computers and make a guess about what well happen in the future.
Cosmic Relief writes:
> But I can’t talk about it without risking irreparable damage to your future
> timeline.
Especially since it turns out that Lumpy is actually the person who will invent . . . um, maybe we better not talk about this here.
‘Yotta’ isn’t the highest named prefix even now. 1000 yottabytes = 1 brontobyte.
1000 brontobytes = 1 geopbyte.
I don’t know how high we will get for things like individual processors or hard drive capacity but petabyte sized hard drives are a certainty within a reasonable and some supercomputer will have storage space of over an exabyte within a few decades. An exabyte is about one million of the terbayte sized hard drives found in a current desktop computer. It is a lot to pack into one computer but can probably be done. Right now, the exabyte is used in real life to estimate things like total internet storage.
It appears those aren’t official(yet).
In any case, if we’re counting bytes, we should really be using the binary versions of the prefixes, since we count them in base 2, not base 10.
We’re OK up to a google, I mean googol.
Since that exceeds the number of subatomic particles in the observable universe, that wouldn’t be possible unless hmm… I just had an idea.
I still favor the Marxian versions.
On the other end, electrical engineers use pico (10** -12) regularly and femto (10** -15) occasionally. (picofarads capacitance and femptoamps leakage currents respectively would be the most common)
ETA, pico was needed for decades before it came in to use. WW2 EEs would say “micromicro”
Yeah, I’m really hoping someday I can get a computer with a 1 harpobyte drive running a 4 grouchohertz processor.
Yobibyte? Zebibit? Pebinibble?
Data units or Gungan appetizers? You decide.
Perhaps we should just tell him, it’ll make the next 10 years a lot easier for him.
Life attempts to imitate South Park:
I’m sure that kibibyte is some kind of dogfood.
Sounds lik a fast food joint from The Flintstones, although those of a nitpicky pursuasion would call it an Apatobyte.
You can get those prefixes in a few lab tests. Mean Corpuscular Hemoglobin (MCH) is measured in picograms, and Mean Corpuscular Volume (MCV) is measured in femtolitres (I’m a computer geek not a lab tech, but I help [del]play with[/del] analyze data coming back from drug trials).
Wouldn’t that be a number rather than a prefix? In any case, if you accept that then don’t forget a googolplex. Watch your spelling on that - a Googleplexseems to be something else entirely :D.
Thirty-five years ago a colleague and I “fixed” an IBM mainframe by placing a 100 pF cap between ground and a key clock ! :dubious: :smack: All it “fixed” was a diagnostic that did an unnecessary test demonstrably inapplicable to system operation, but customer preferred it that way; that cap may have stayed on the clock-pin in that box for years for all I know. Either way, we thought it was fun to write up a report about the 100 picoFarad capacitor fix.
Other than picoFarad caps, I regarded most of the very big and very little numbers as jokes then. I once knew (intimately) the program manager for the world’s first terabit computer memory; he may have been first to use the term “terabit” for all I know. Tomorrow I’m buying an external hard drive; I may splurge and get a 1.2 terabyte drive which now costs less than an evening’s entertainment.