1. Pi
- Posted by Kat <kat at kogeijin.com> Dec 10, 2002
- 455 views
This is what string math is for: A team of Japanese researchers say they have calculated Pi to 1.24 trillion digits. The most accurate previous calculation only stretched to 206 billion places. They calculated the value for pi with a Hitachi supercomputer over 400 hours in September. Kanada’s team spent five years designing the program used in the September experiment, Kudo said. The Hitachi supercomputer is capable of 2 trillion calculations per second. Kat
2. Re: Pi
- Posted by Derek Parnell <ddparnell at bigpond.com> Dec 10, 2002
- 416 views
This brings to mind at least two questions=2E=2E=2E 1) How do you prove that they are right=3F 2) Why bother anyway=3F With that sort of accuracy=2C we can measure the = known universe with an error of less than a sub-atomic particle=2E That = is definite overkill=2E I guess they had to spend that grant money = somehow =3B-) -------- cheers=2C Derek ----- Original Message ----- From=3A Kat =3Ckat=40kogeijin=2Ecom=3E Date=3A Tuesday=2C December 10=2C 2002 11=3A33 am Subject=3A Pi =3E = =3E This is what string math is for=3A =3E A team of Japanese researchers say they have calculated Pi to 1=2E24 = =3E trillion digits=2E = =3E The most accurate previous calculation only stretched to 206 = =3E billion places=2E They = =3E calculated the value for pi with a Hitachi supercomputer over 400 = =3E hours in = =3E September=2E Kanada=92s team spent five years designing the program = =3E used in the = =3E September experiment=2C Kudo said=2E The Hitachi supercomputer is = =3E capable of 2 = =3E trillion calculations per second=2E =3E = =3E Kat =3E = =3E =3D=3D=5E=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3E This email was sent to=3A ddparnell=40bigpond=2Ecom =3E = =3E http=3A//topica=2Ecom/u/=3Fb1dd66=2Eb2mL5y=2EZGRwYXJuOr send an email= to=3A = =3E EUforum-unsubscribe=40topica=2Ecom =3E = =3E = =3E =
3. Re: Pi
- Posted by Kat <kat at kogeijin.com> Dec 10, 2002
- 412 views
On 10 Dec 2002, at 11:41, Derek Parnell wrote: > > This brings to mind at least two questions... > > 1) How do you prove that they are right? > > 2) Why bother anyway? With that sort of accuracy, we can measure the > known universe with an error of less than a sub-atomic particle. That > is definite overkill. I guess they had to spend that grant money > somehow I don't know the cost of superputer time, but i think i recall a figure of $10,000/hour being bandied about once. So 400 hours = $4Million. That's not counting the 5 years of the superputer to develop the code, and the salaries (at top university pay scales) of all those involved. Sounds like it would have been easier and quicker, if less expensive, to use Matt's bignum lib! > -------- > cheers, > Derek > > ----- Original Message ----- > From: Kat <kat at kogeijin.com> > Date: Tuesday, December 10, 2002 11:33 am > Subject: Pi > > > > This is what string math is for: > > A team of Japanese researchers say they have calculated Pi to 1.24 > > trillion digits. > > The most accurate previous calculation only stretched to 206 > > billion places. They > > calculated the value for pi with a Hitachi supercomputer over 400 > > hours in > > September. Kanada’s team spent five years designing the program > > used in the > > September experiment, Kudo said. The Hitachi supercomputer is > > capable of 2 > > trillion calculations per second. > > > > Kat > > > > > > http://topica.com/u/?b1dd66.b2mL5y.ZGRwYXJuOr send an email to: > > EUforum-unsubscribe at topica.com > > > > > ==^^=============================================================== > This email was sent to: kat at kogeijin.com > > >
4. Re: Pi
- Posted by jordah at btopenworld.com Dec 10, 2002
- 418 views
Why would some one need a number that long? a billion maybe. A trillion... Was it to show off the supercomputer or the actual program? jordah ----- Original Message ----- From: "Kat" <kat at kogeijin.com> To: "EUforum" <EUforum at topica.com> Subject: Pi > > This is what string math is for: > A team of Japanese researchers say they have calculated Pi to 1.24 trillion digits. > The most accurate previous calculation only stretched to 206 billion places. They > calculated the value for pi with a Hitachi supercomputer over 400 hours in > September. Kanada's team spent five years designing the program used in the > September experiment, Kudo said. The Hitachi supercomputer is capable of 2 > trillion calculations per second. > > Kat > > > >
5. Re: Pi
- Posted by xanax at bellsouth.net Dec 10, 2002
- 440 views
This sounds like the prime-number venture. actually paying people to find primes above what has already been found but with todays computers, it'd take years just to match what they already have. Xanax ----- Original Message ----- From: <jordah at btopenworld.com> To: "EUforum" <EUforum at topica.com> Subject: Re: Pi : : Why would some one need a number that long? a billion maybe. A trillion... : : Was it to show off the supercomputer or the actual program? : : jordah : ----- Original Message ----- : From: "Kat" <kat at kogeijin.com> : To: "EUforum" <EUforum at topica.com> : Sent: Tuesday, December 10, 2002 12:33 AM : Subject: Pi : : : > : > This is what string math is for: : > A team of Japanese researchers say they have calculated Pi to 1.24 : trillion digits. : > The most accurate previous calculation only stretched to 206 billion : places. They : > calculated the value for pi with a Hitachi supercomputer over 400 hours in : > September. Kanada's team spent five years designing the program used in : the : > September experiment, Kudo said. The Hitachi supercomputer is capable of 2 : > trillion calculations per second. : > : > Kat : > : > : > : > : : : : :
6. Re: Pi
- Posted by xanax at bellsouth.net Dec 10, 2002
- 425 views
This sounds like the prime-number venture. actually paying people to find primes above what has already been found but with todays computers, it'd take years just to match what they already have. Xanax ----- Original Message ----- From: <jordah at btopenworld.com> To: "EUforum" <EUforum at topica.com> Subject: Re: Pi : : Why would some one need a number that long? a billion maybe. A trillion... : : Was it to show off the supercomputer or the actual program? : : jordah : ----- Original Message ----- : From: "Kat" <kat at kogeijin.com> : To: "EUforum" <EUforum at topica.com> : Sent: Tuesday, December 10, 2002 12:33 AM : Subject: Pi : : : > : > This is what string math is for: : > A team of Japanese researchers say they have calculated Pi to 1.24 : trillion digits. : > The most accurate previous calculation only stretched to 206 billion : places. They : > calculated the value for pi with a Hitachi supercomputer over 400 hours in : > September. Kanada's team spent five years designing the program used in : the : > September experiment, Kudo said. The Hitachi supercomputer is capable of 2 : > trillion calculations per second. : > : > Kat : > : > : > : > : : : : :
7. Re: Pi
- Posted by Kat <kat at kogeijin.com> Dec 10, 2002
- 436 views
On 7 Dec 2002, at 13:02, jordah at btopenworld.com wrote: > > Why would some one need a number that long? a billion maybe. A trillion... > > Was it to show off the supercomputer or the actual program? I don't know. Possibly to test the computer, possibly to test the program. It's my understanding that in astronomy, in the million of calculations involved in travel away from earth, significant errors can creep in at rediculously small decimal point places, and if eliminated, can save fuel. But i don't know at which point it stops being significant versus a waste of cpu horsepower/time. Kat > jordah > ----- Original Message ----- > From: "Kat" <kat at kogeijin.com> > To: "EUforum" <EUforum at topica.com> > Sent: Tuesday, December 10, 2002 12:33 AM > Subject: Pi > > > > This is what string math is for: > > A team of Japanese researchers say they have calculated Pi to 1.24 > trillion digits. > > The most accurate previous calculation only stretched to 206 billion > places. They > > calculated the value for pi with a Hitachi supercomputer over 400 hours in > > September. Kanada's team spent five years designing the program used in > the > > September experiment, Kudo said. The Hitachi supercomputer is capable of 2 > > trillion calculations per second. > > > > Kat > > > > > > >
8. Re: Pi
- Posted by jbrown1050 at hotpop.com Dec 10, 2002
- 439 views
On Mon, Dec 09, 2002 at 10:12:43PM -0600, Kat wrote: > It's > my understanding that in astronomy, in the million of calculations > involved in > travel away from earth, significant errors can creep in at rediculously > small > decimal point places, and if eliminated, can save fuel. But i don't know > at > which point it stops being significant versus a waste of cpu > horsepower/time. > > Kat I'm not sure if any one does. Hence the quest for on-going (perhaps one day infinite??) accuracy. jbrown
9. Re: Pi
- Posted by Dan Moyer <DANIELMOYER at prodigy.net> Dec 11, 2002
- 439 views
It's not possible to have "infinite" accuracy, 'cause there could ALWAYS be more digits, that's what infinite means. Dan Moyer ----- Original Message ----- From: <jbrown1050 at hotpop.com> To: "EUforum" <EUforum at topica.com> Subject: Re: Pi > > On Mon, Dec 09, 2002 at 10:12:43PM -0600, Kat wrote: > > It's > > my understanding that in astronomy, in the million of calculations > > involved in > > travel away from earth, significant errors can creep in at rediculously > > small > > decimal point places, and if eliminated, can save fuel. But i don't know > > at > > which point it stops being significant versus a waste of cpu > > horsepower/time. > > > > Kat > > I'm not sure if any one does. Hence the quest for on-going (perhaps one day > infinite??) accuracy. > > jbrown > > > >