Simple WCG PPD Estimator

[brandonwh64]

I can update it if I new what values to add.

C# is like pointerless C/C++.

Yea you have to go easy on me, this is only my third week of C++ class LOL, Im learning but picking up pretty quick. It would have been easier to have his code, Is there a way we could decompile his last update?

 

[KieX]

Values are based off the formula:

Multi = Points/MHz
(Where points are averaged over a few days).

Then you simply multiply the Multi x MHz. Detract 10% for 32-bit. Divide by 7 for the BOINC points

Give me a few minutes and I can work out the values for:
-i7 Sandybridge
-i7 Sandybridge-E

For the bulldozers, would need to ask mjkmike as I think he's the only one crunching that in our team?

EDIT: Got them!

            _ProcessorDatabase.Add(new Processor("Intel", "Core i7", "SB Quad-core", 9));
            _ProcessorDatabase.Add(new Processor("Intel", "Core i7", "SB-E Hexa-core", 13.5));

 

[FordGT90Concept]

Looking through the source, the numbers I posted appear to be for 32-bit. It times by 1.1 for 64-bit.

It does...

BaseScore = ProcessorValue * ProcessorSpeed * ProcessorCount
WCG = BaseScore * (CpuUsage / 100)
BOINC = WCG / 7

 

[brandonwh64]

Ok from what I see here are the values from one of the previous releases

C2D (Conroe): 2
C2D (Wolfdale): 2.2
C2Q (Kentsfield): 4
C2Q (Yorkfield): 4.45
Ci3/Ci5 (Clarkdale): 4.1
Ci5 (Lynnfield): 5.9
Ci7 (quad-core): 8.1
Ci7 (hex-core): 12.1
Athlon 64X2 / Athlon X2: 1.7
PhI X3: 2.3
PhI X4: 3.2
AII X2: 2.0
AII X3: 3.0
AII X4: 3.9
PhII X2: 2.1
PhII X3: 3.1
PhII X4: 4.2
PhII X6: 6.7

 

[FordGT90Concept]

EDIT: Got them!

            _ProcessorDatabase.Add(new Processor("Intel", "Core i7", "SB Quad-core", 9));
            _ProcessorDatabase.Add(new Processor("Intel", "Core i7", "SB-E Hexa-core", 13.5))

Added to my source.

 

[brandonwh64]

I could probably base it off last weeks assignments in class just add a few more details

#include <iostream>
#include <iomanip>

using namespace std;

int main(void)
{
char first [20],
last [20];
double monpay;


cout << "Employee First Name: " << endl;
cin >> first;
cout << "Employee Last Name: " << endl;
cin >> last;
cout << "Monthly Pay: " << endl;
cin >> monpay;


double fed, state, ssmed;

fed = monpay * 0.15;
state = monpay * 0.035;
ssmed = monpay * 0.085;


cout << "Employee: " << first << last << endl
<< "Gross Amount: " << fixed << setprecision(2) << setw(15) << setfill('.') << monpay << "$" << endl
<< "Federal Tax: " << fixed << setprecision(2) << setw(15) << setfill('.') << fed << "$" << endl
<< "State Tax: " << fixed << setprecision(2) << setw(15) << setfill('.') << state << "$" << endl
<< "Social Security and Medicare Tax: " << fixed << setprecision(2) << setw(15) << setfill('.') << ssmed << "$" << endl
<< "Health Insurance: " << fixed << setprecision(2) << setw(15) << setfill('.') << "75$" << endl
<< "Net Pay: " << fixed << setprecision(2) << setw(15) << setfill('.') << monpay - fed - state - ssmed - 75 << "$" << endl;


cout << "Press any key to exit." << endl;
cin.ignore(2);

return 0;
}

Added to my source.



Those are different than what I have, should I change them to that?

From what I tested in his latest one, they are accurate.

 

[KieX]

Looking through the source, the numbers I posted appear to be for 32-bit. It times by 1.1 for 64-bit.

It does...

BaseScore = ProcessorValue * ProcessorSpeed * ProcessorCount
WCG = BaseScore * (CpuUsage / 100)
BOINC = WCG / 7

Ah, works it out the other way round. So:
SB Quad: ProcessorValue = 9
SB-E Hex: ProcessorValue = 13.5

Using 4.5GHz 2600K as an example on 32bit 100% CPU use:

BaseScore = ProcessorValue * ProcessorSpeed * ProcessorCount
40500 = 9 * 4500 * 1
WCG = BaseScore * (CpuUsage / 100)
40500 = 40500 *1
BOINC = WCG / 7
5786 = 40500 / 7

Then the 1.1x to get 64bit results.

Added to my source.

I missed a ";" on the last one just noticed

 

[FordGT90Concept]

From what I tested in his latest one, they are accurate.

I double checked and they all match mine...

 

[brandonwh64]

Cin << Mhz * CPU / 7;
^ will give you the 32bit points and without the / 7 you get WCG points right?

 

[FordGT90Concept]

It calculates for WCG (the numbers you gave are for WCG). BOINC is 7 times less than WCG.

 

[[XC] Oj101]

If anyone would like an update, I am the author of the web based version.

 

[Chicken Patty]

If anyone would like an update, I am the author of the web based version.

What's the link to yours?

 

[[XC] Oj101]

I wrote it specifically for Ion, I never hosted it myself. If you want I'll have a copy up within the next few hours, I must just dig up the source code.

 

[[XC] Oj101]

Found it in my sent items I've uploaded it to my site.

http://flyingsuicide.net/wcg/index.php

 

[Chicken Patty]

This is the same as the one in the OP correct? Obviously this being a web based version.

 

[[XC] Oj101]

It's similar, it's the one listed here, I wrote it for the OP.

 

[Delta6326]

Hey awesome program I use it all the time. Was wondering if any one knew how much one of these would give out a day PPD? AMD Opteron 6274 Interlagos 2.2GHz 16MB L3 Cache S... ??

 

[Chicken Patty]

F150 might be able to have an idea as he has an AMD Quad CPU setup crunching. I would say though at least 8000 ppd.

 

[KieX]

Hey awesome program I use it all the time. Was wondering if any one knew how much one of these would give out a day PPD? AMD Opteron 6274 Interlagos 2.2GHz 16MB L3 Cache S... ??

A 4x 6272 cruncher will average 15-17K PPD. That would mean each CPU can do roughly 3.8K-4.3K PPD. 2.1GHz vs 2.2GHz of the 6274 would be a small increase perhaps 4.4K?

Been inactive the last few days, but you can see from the other days what they can do:
http://stats.free-dc.org/stats.php?page=host&proj=bwcg&hostid=1820078

 

[Delta6326]

Ok thanks. I have been looking at a server rack-mount systems.

 

[KieX]

Ok thanks. I have been looking at a server rack-mount systems.

That sounds like a whole lot of fun to be had

 

[FordGT90Concept]

And headaches when it breaks. XD

 

[[Ion]]

Hey guys, I'm back! I recognize that my program is...more than a bit out of date. If I can get people to PM/post values for the new Sandy & Ivy Bridge CPUs & Bulldozer, I'll get things up-to-date!

 

[mstenholm]

Welcome back

I noticed that a kpresler (from Sweden ) re-joined.

 

[Mindweaver]

Hey guys, I'm back! I recognize that my program is...more than a bit out of date. If I can get people to PM/post values for the new Sandy & Ivy Bridge CPUs & Bulldozer, I'll get things up-to-date!

Welcome back [Ion]!