- Status
- Jul 24, 2009
- 4,169
- 5
- 22,965
Lets find out!
40,000,000 - 20,169,022 = 19,830,978
19,830,978 ÷ 73,348 = 270.368353602 days
(round to down to the whole number)
270 ÷ 365 = 0.739726027
Well, 0.73 is close enough .75 and that = 3/4. 3/4 of year from now would around late September or early October.
So to answer your question. at your current pace, yes you should hit 40 million before you even hit November.
And just to make sure my math is correct...
http://folding.extremeoverclocking.com/user_future.php?u=322067&f=271
So 1 day difference between my math and what this says but hey, it a very narrow range meaning the September/October time frame about right.
beanoslim
Splendid
- Oct 13, 2010
- 3,612
- 1
- 23,365
Hi HudTech, you should indeed, smp or bigadv Folding with the 950 depending on how much electricity you have spare.
The bigadv is pretty much a 24/7 setup, heavy workload for heavy points.
Otherwise with that cpu you would get some good smp Folding points and still free up your machine.
Happy Folding. 😀
The bigadv is pretty much a 24/7 setup, heavy workload for heavy points.
Otherwise with that cpu you would get some good smp Folding points and still free up your machine.
Happy Folding. 😀
beanoslim
Splendid
- Oct 13, 2010
- 3,612
- 1
- 23,365
Just hit 4.2ghz with my 920, every 200mhz increase added about 2k ppd. 😀 (bigadv)
You need to set up smp2 to start with anyways, then after 10 units [strike]if you decide you fancy a crack at [/strike]you can switch to bigadv and instantly get the bonus. 😀
Loving these new 912pt GPU wu's, not bad for a card that cost £115.
You need to set up smp2 to start with anyways, then after 10 units [strike]if you decide you fancy a crack at [/strike]you can switch to bigadv and instantly get the bonus. 😀
Loving these new 912pt GPU wu's, not bad for a card that cost £115.
- Jul 24, 2009
- 4,169
- 5
- 22,965
Hello Hud-tech and Welcome to the team.
If you need any help on installing the SMP2 client that the Official Install Guide ( http://folding.stanford.edu/English/WinSMPGuide ) is not clear about, I would recommend looking at this video i made a while back of how to install F@H on the computer.
http://www.youtube.com/watch?v=Vr9WLNOGta8
If you need any more help, just let us know! [:graywolf:8]
- Jul 24, 2009
- 4,169
- 5
- 22,965
elkein :
Well obviously moores law isn't real... But would someone better than me apply that to the slight exaggeration above, then see if that 20,000,000,000,000,000 points would be achievable in a lifetime. Just playing devils advocate here. 
:lol:
Looks like the numbers mess with our minds [:rocket_sauce]
yakri
Distinguished
- Oct 10, 2009
- 115
- 0
- 18,690
Ok, I did some cool calculations, unfortunately I couldn't find a calculator powerful enough to calculate the PPD of the entire F@H project in 50 years.
Basically, I went and got the FLOPS/PPD of a sandy bridge processor, added it up into one petaflop x 5 for the total PPD/PFLOPS of the entire project, and got the total PPD.
Of course if I accidentally missed a zero somewhere this is going to be horribly wrong, and it should have a wide margin of error anyway.
Still, according to my math the time it would take for the F@H project to hit 20,000,000,000,000,000 points is 13,150.8164 years.
The total PPD is: 4,166,625,000
If anyone can find a big enough calculator, do 4,166,625,000 to the power of 33 to get the theoretical PPD in 50 years according to moore's law.
In actuality, the PPD in 50 years would probably be MORE than what you'll get from that calculation, but it's good enough for us I suppose. 😛
Basically, I went and got the FLOPS/PPD of a sandy bridge processor, added it up into one petaflop x 5 for the total PPD/PFLOPS of the entire project, and got the total PPD.
Of course if I accidentally missed a zero somewhere this is going to be horribly wrong, and it should have a wide margin of error anyway.
Still, according to my math the time it would take for the F@H project to hit 20,000,000,000,000,000 points is 13,150.8164 years.
The total PPD is: 4,166,625,000
If anyone can find a big enough calculator, do 4,166,625,000 to the power of 33 to get the theoretical PPD in 50 years according to moore's law.
In actuality, the PPD in 50 years would probably be MORE than what you'll get from that calculation, but it's good enough for us I suppose. 😛
yakri
Distinguished
- Oct 10, 2009
- 115
- 0
- 18,690
I went and double checked, and it turns out the farthest google calculator could go was to the power of 32, which is only 18 months short, so here's the result.
4,166,625,000^32 = 6.80928153 × 10(307)
The 307 I believe means 307 zeros in scientific notation.
So. . . . Yeah, kind of a big number.
If you estimate the PPD of the theoretical most powerful single super computer in 2061 it comes out to.
1,666,650,000^33 = 2.09346363 × 10(304)
That estimate is actually on the low side, btw. you'd have to increase the first number by 0.5 to be really accurate, but I'm lowballing it to compensate for the fact that my PPD estimate might be a little high.
tl;dr The number beano threw out there should be doable in what, minutes, seconds? (I've already made my brain hurt too much getting this far, someone else can figure that out.) On a single super computer 50 years from now.
Interesting note: Some super computer building companies have hinted that exaflops, "might not be as far off as you think," which could mean all this could potentially happen a lot sooner than 50 years from now if the right technological leaps are made.
4,166,625,000^32 = 6.80928153 × 10(307)
The 307 I believe means 307 zeros in scientific notation.
So. . . . Yeah, kind of a big number.
If you estimate the PPD of the theoretical most powerful single super computer in 2061 it comes out to.
1,666,650,000^33 = 2.09346363 × 10(304)
That estimate is actually on the low side, btw. you'd have to increase the first number by 0.5 to be really accurate, but I'm lowballing it to compensate for the fact that my PPD estimate might be a little high.
tl;dr The number beano threw out there should be doable in what, minutes, seconds? (I've already made my brain hurt too much getting this far, someone else can figure that out.) On a single super computer 50 years from now.
Interesting note: Some super computer building companies have hinted that exaflops, "might not be as far off as you think," which could mean all this could potentially happen a lot sooner than 50 years from now if the right technological leaps are made.
- Jul 24, 2009
- 4,169
- 5
- 22,965
As for a calculator that can show big numbers....
The calculator in ubuntu 10.10 (in scientific mode) has a handy slider to show the entier length of the number. 😗
http://lh4.ggpht.com/_xiSd9MyLJuA/TUTsclS5o3I/AAAAAAAAAFA/DN4IQjTyd9s/s800/Screenshot.png
http://lh5.ggpht.com/_xiSd9MyLJuA/TUTscnCifyI/AAAAAAAAAFE/h2oDKuuDr4E/s800/Screenshot-1.png
http://lh4.ggpht.com/_xiSd9MyLJuA/TUTuIFfHYQI/AAAAAAAAAFM/QC9w7M8I5-A/s800/Screenshot-2.png
http://lh3.ggpht.com/_xiSd9MyLJuA/TUTuIMD-N7I/AAAAAAAAAFQ/i9acS7QZIOk/s800/Screenshot-3.png
:lol:
yakri
Distinguished
- Oct 10, 2009
- 115
- 0
- 18,690
Hmmm~
I really need to get unbuntu studio installed on this PC, I had it on my old one, but I haven't gotten it dual booted on this one yet because I was getting some new weird error during install, I should really take another crack at it.
The calculator I used just displayed, "ERROR," instead of a number.
Google calc just didn't do anything.
lol
I think you guys went about calculating that number wrong. 
(even though Moore's law is only the doubling of transistors, not performance)
The performance should supposedly double every 18 months if the doubling of transistors meant doubling the performance, so
One Sandy Bridge processor*2^(years/1.5)-I'm only going to go 30 years.
50,000*2^20=52,488,000-This is the PPD one processor would get 30 years from now.
Now divide that number bean threw out by the PPD
20,000,000,000,000,000/52,488,000=~381029475-This is the days it would take to complete that number.
Now make that into years
381029475/365=113,236 years
For 1 processor to make that number in 30 years, that would take it 113,236 more years to get there, but there are going to be more than 200,000 processors folding most likely, so the folding project could get there lightning fast.
113,236/200,000= ~1/2 a year
You forgot that when you have the number to a power, you're more than doubling the amount of PPD per 18 months that F@H would supposedly get. 😉
I don't want to bother trying to calculate the PPD of f@h at that time, because that would mean you have to know what older computers would supposedly be folding still. Just leave it as a simple calculation like this.
(even though Moore's law is only the doubling of transistors, not performance)The performance should supposedly double every 18 months if the doubling of transistors meant doubling the performance, so
One Sandy Bridge processor*2^(years/1.5)-I'm only going to go 30 years.
50,000*2^20=52,488,000-This is the PPD one processor would get 30 years from now.
Now divide that number bean threw out by the PPD
20,000,000,000,000,000/52,488,000=~381029475-This is the days it would take to complete that number.
Now make that into years
381029475/365=113,236 years
For 1 processor to make that number in 30 years, that would take it 113,236 more years to get there, but there are going to be more than 200,000 processors folding most likely, so the folding project could get there lightning fast.
113,236/200,000= ~1/2 a year
You forgot that when you have the number to a power, you're more than doubling the amount of PPD per 18 months that F@H would supposedly get. 😉
I don't want to bother trying to calculate the PPD of f@h at that time, because that would mean you have to know what older computers would supposedly be folding still. Just leave it as a simple calculation like this.
- Jul 24, 2009
- 4,169
- 5
- 22,965
Maybe absorbed into another unknown project. This has happen in F@H past. Although it was another project going into F@H.
Genome@home
http://en.wikipedia.org/wiki/Genome@home
yakri
Distinguished
- Oct 10, 2009
- 115
- 0
- 18,690
Haserath :
I think you guys went about calculating that number wrong. (even though Moore's law is only the doubling of transistors, not performance)
The performance should supposedly double every 18 months if the doubling of transistors meant doubling the performance, so
One Sandy Bridge processor*2^(years/1.5)-I'm only going to go 30 years.
50,000*2^20=52,488,000-This is the PPD one processor would get 30 years from now.
Now divide that number bean threw out by the PPD
20,000,000,000,000,000/52,488,000=~381029475-This is the days it would take to complete that number.
Now make that into years
381029475/365=113,236 years
For 1 processor to make that number in 30 years, that would take it 113,236 more years to get there, but there are going to be more than 200,000 processors folding most likely, so the folding project could get there lightning fast.
113,236/200,000= ~1/2 a year
You forgot that when you have the number to a power, you're more than doubling the amount of PPD per 18 months that F@H would supposedly get. 😉
I don't want to bother trying to calculate the PPD of f@h at that time, because that would mean you have to know what older computers would supposedly be folding still. Just leave it as a simple calculation like this.
(even though Moore's law is only the doubling of transistors, not performance)The performance should supposedly double every 18 months if the doubling of transistors meant doubling the performance, so
One Sandy Bridge processor*2^(years/1.5)-I'm only going to go 30 years.
50,000*2^20=52,488,000-This is the PPD one processor would get 30 years from now.
Now divide that number bean threw out by the PPD
20,000,000,000,000,000/52,488,000=~381029475-This is the days it would take to complete that number.
Now make that into years
381029475/365=113,236 years
For 1 processor to make that number in 30 years, that would take it 113,236 more years to get there, but there are going to be more than 200,000 processors folding most likely, so the folding project could get there lightning fast.
113,236/200,000= ~1/2 a year
You forgot that when you have the number to a power, you're more than doubling the amount of PPD per 18 months that F@H would supposedly get. 😉
I don't want to bother trying to calculate the PPD of f@h at that time, because that would mean you have to know what older computers would supposedly be folding still. Just leave it as a simple calculation like this.
Yeah, you have a point in that I forgot moore's law is about the number of transistors, not computing power.
But don't forget that I was calculating for what one super computer could do in 50 years, based on PPD/FLOPS. 50 years has around 33 18 month sets, so I was calculating to the power of 33. ---- Fuuuuuuu, I'm going to just interject this in here so you can see how my thoughts went instead of editing it. I just realized that I should have squared the number 33 times instead of going to the 33rd power. ***.
But super computers have been nearly doubling performance every 18 months, not quite I don't think, but they aren't too far off. So would a formula like,
where (a) is the power(number of FLOPS) of a present day super computer, and (b) is the ratio by which the super computer's FLOPS increases every 18 months, then squared and repeated 33 times work? Then you could use the PPD per PFLOPS to find the theoretically possible PPD, right?
shovenose
Distinguished
- Dec 6, 2009
- 3,377
- 0
- 20,960
why wont the mac os x powerpc client do anything?i want to run folding@home on my old ibook g4. im running os x 10.4 tiger, and its connected to the internet with the internal wireless card.
the client opens fine, etc. it just shows a black screen and doesnt work on any wu
the client opens fine, etc. it just shows a black screen and doesnt work on any wu
- Jul 24, 2009
- 4,169
- 5
- 22,965
I'll look into this issue.
Although i cant say i'll be speedy returning an answer. Currently been experiencing internet issues with my ISP (Hargray).
They say there updating there email system but at the moment when they said they were upgrading is when my internet been going all out of wack. There times i can load a page like there nothing going on and other times i cant load any pages.
Be right back with an answer.
- Status
TRENDING THREADS
-
-
I swapped my mini-LED display for a $1,300 OLED monitor. Here’s what happened- Started by Admin
- Replies: 27
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
-
Facebook
Twitter
Instagram