3d Mark 2003-Wednesday

[Dr. Data]

Hey guys been busy lately sorry.

Anyway, I have been watching the counter since Saturday and by my estimates we should have this available wednesday morning.

http://www.madonion.com

[Swedishlf]

I disagree, I did some calculations just now, and it looks to me to be a bit sooner.

The countdown runs through 5000 points every 15 seconds, or thereabouts.

So the timer read exactly 17500000 at 11:35AM central time, meaning the timer would run out 14 hours and 30 minutes from then or at 3:05AM Tuesday morning...

17500000/5000 * 15 /60/60 =14.5hrs


Anyone else wanna try?

[Dr. Data]

LOL, I agree its sooner.

As crazy as this may sound....it appears to be accelerating, so any calculation you do could be off because its gradually getting faster? Saturday morning around 9am it was at 61xxxxx, and Sunday around lunch time it was at 57xxxxx, so around 10xxxxx per 24 hours right? Wrong

One of the girls in my office just put it at Tuesday afternoon, but I have a sneaking suspicion it might be available late tonight somtime....

[phenix]

By my calculations 15 hrs

17000000

The underlines digit is 5 seconds

[wrathchild_67]

So who's gonna start the obligatory "Post your 3DMark03" scores here!" thread?

[phenix]

I did.(since i'll be the first to own the program)

[Swedishlf]

Okay, huge geek points here hopefully...

As best I can tell it's accelrating at a rate of .000603 number/second^2

At number 17500000, the rate was almost exactly 333.33 numbers/second.

Simple physics formulas from here:

x=VoT + 1/2 aT^2

Solve for T

T= (V (+_) (Vo^2+2*a*x)^(1/2))/a

Two solutions present themselves, 50217.3 seconds and -1154930 seconds... since there's no negative time...

50217.3 seconds ~13 hours, 56 minutes, 57.3 seconds

so at 11:35, that's my best guess at how long it had to run placing the release time at:

01:01:57.3 AM Central time...

So if anyone knows what time zone www.madonion.com is run from, that might make sense. If they are on mountain time, that's almost exactly midnight tonight But I could be wrong I just did quick sloppy calculations and not so good measurements.

Yes, I was that bored.

[Dr. Data]

Unless the rate of acceleration is also variable.....

What would be the term for an accelerating rate of acceleration?

Parabolic curve?

[Swedishlf]

I found an entire website about that... I think the unofficial but commonly accepted term for the time derivative of acceleration is "jolt" or "jerk" or something to that effect, i'll see if I can dig it up.

Edit:

The page in question is here

[wrathchild_67]

Actually, I think the servers are located in Finland.

[phenix]

I don't think the time is accelerating it's more likely an illusion.

Good work Swedishlf. Sad...but good.

[Dr. Data]

Hmmm.

I guess Ill get up at 2Am EST and start the dl. With the Internet traffic being less(hopefully) that time of night maybe the mirrors wont be overloaded, but my bet is that it will be days before Ill be able to get it through on a dial up....

[phenix]

http://discuss.futuremark.com/documents/counter2.html

[RDX]

My guess is 1:00:00 eastern standard clock which would be Noon central time. My calculations as follows.

At 8:45:00 (+/- .2sec) EST the meter was at 11,700,000 (Measured with the boulder atomic clock time)

At 8:53:20 EST the meter was at 11,600,000

At 9:01:40 EST the meter was at 11,500,000

Following this, the pattern is clearly linear with the countdown decreasing at a rate of 200 points per second.

11,500,000*500sec/100,000 = 57500seconds left after 9:01:40 or 15hrs 58min 20 sec.

That works out exactly to 1:00 EST.

[Swedishlf]

Wait everyone, turns out I'm a moron...

I was using my rubik's cube timer to time the intervals, I forgot that it counts down for 10 seconds before it starts counting back up... so all of my intervals should have 10 added to them...

new estimate is (using the same formula as before)

23 hours 24 minutes and 20.7 seconds after 11:35 AM CT Feb 10

or almost exactly 11:00AM CT Feb 11 Which is 12:00PM ET... I did take into account acceleration which may or may not be a product of errors in the measurement. But this is much much closer to what others have calculated.

Please accept my appologies for the mix up!