Back to Recipes Homepage
recipe picture
Recipe: Test your PC power
3.75
Your rating: None Average: 3.8 (4 votes)

Profile

Name: Test your PC power
ID: 46929
Created on: Sat, 09/21/2013 - 04:16
Updated on: Sat, 09/21/2013 - 11:16
Description:

Test your PC on Contest http://fold.it/portal/node/995975



Best For


Comments

Joined: 09/24/2012
Groups: Go Science
User guide

It's not really a game. It only allows you to compare the power of your computer with best computers of other players. Some computers are good for Foldit, other poor. I noticed that Foldit recipe is very CPU resource consuming. I suppose that when a player buys a new computer, he will succeed gaining more points to this test.

1) Click on this contest: http://fold.it/portal/node/995975
2) Go to "Link to join"
Click on "Click here if you would like to join."
3) Open Foldit
4) In the main menu, click on "Contests" on the right pannel
You should see the Contests you joined, including this one. Select it and click on "play".
You'll get a very big protein.
5) Click here above right in order to load this recipe (or, in the recipe menu, type manually the number 46825 to manually load this recipe).
6) Run the recipe once.
7) Your score will rank your pc relative to my best pc I ever tested there
8) Open the right green button on the recipe in order to see the reciupe log. Here you can see some information, for example the CPU time of the first recipe loop. With a better PC, you'll see a lower CPU time there.
9) Save and share the result with yourself or your group for later archive. Whrite there the characteristics of your computer.
10) You might consider adding it here too. This could help people to select best PC adapted to Foldit ... and show that a powerfull PC is good for using recipes (it's less important for hand folding).

Joined: 09/24/2012
Groups: Go Science
Intel Core i3-2350M CPU @ 2.30 GHz 2.30 GHz

CPU time first loop: 13.232 sec
Result report: 15125.471 pts
Total gain: 15.794
49 loops

Computer:

Constructor: ICT Service desk
Processor: Intel Core i3-2350M CPU @ 2.30 GHz 2.30 GHz
RAM: 4.00 Go
System: 64 bits
OS: Windows 7 Enterprise, Service Pack 1

Joined: 09/24/2012
Groups: Go Science
Intel Core 2 Duo CPU @ 2.13 GHz

CPU time first loop: 32.9 sec
Result report: 15110.991 pts
Total gain: 1.141

Computer:

Constructor: Apple
Model: MacBook Air
Processor: Intel Core 2 Duo CPU @ 2.13 GHz
RAM: 2 Go 1067 MHz DDR3
Ver noyau: Darwin 10.8.0
OS: Mac OSX 10.6.8

Joined: 09/24/2012
Groups: Go Science
What should you compare?

the CPU time first loop and the number of loops is the only pertinent results to be compared. Above, the old Mac took 19 loops. It is approximately 2.5 times less powerful than the (new) pc above.

The points are only good for ranking: more point = more powerful computer, but 200 pts is not 2 times more powerful than 100 pts (the points depend on the zone of the protein used for the test). I used a very big protein in order to allow this test last for several years.

I'll add other test results here. Please feel free to add results of other computers and configurations here.

If you save and share your result, when you run the resipe on another computer, previous results are archived in note 1 (and also displayed in the log).

Joined: 09/24/2012
Groups: Go Science
Intel Core i7 CPU @ 2.9 GHz, 2 cores

CPU time first loop: 11.24 sec
Result report: 15125.584 pts
Total gains: 15.907 pts
51 loops

3 clients running at the same time

Computer:

MacBook Pro (Mid 2012)
Processor: Intel Core i7 CPU @ 2.9 GHz, 2 cores
1 processor, 2 cores, niv2 cache 256 Ko/core, niv 3 cache: 4 Mo
RAM: 8 Go 1600 MHz DDR3
Intel HD graphics 4000 512 Mo
OS: OS X 10.8.4 (12E55)

Joined: 09/24/2012
Groups: Go Science
Intel Core i7 CPU @ 2.9 GHz, 2 cores

CPU time first loop: 674 sec
Result report: 13417 pts
Total gains: 3866 pts
1 loops

1 client running at the same time

Computer:

MacBook Pro (Mid 2012)
Processor: Intel Core i7 CPU @ 2.9 GHz, 2 cores
1 processor, 2 cores, niv2 cache 256 Ko/core, niv 3 cache: 4 Mo
RAM: 8 Go 1600 MHz DDR3
Intel HD graphics 4000 512 Mo
OS: OS X 10.11.6 (12E55)

Don't update your system !!! El Capitain dramatically slows down the computer !!

Joined: 09/24/2012
Groups: Go Science
Intel Core Duo T2300 1.66GHz

CPU time first loop: 26.8 sec
Result report: 15112.437 pts
Total gain: 2.76
24 loops

1 client

Computer:

Vaio vgn-sz2M/B
Processor: Intel Core Duo T2300 1.66GHz
1 processor, 2 cores
RAM: 1GB/Go DDR2 SDRAM (533MHz)
Graphics: NVIDIA geforce Go 7400 with turbochache supporting 128MB/Mo
OS: Ms Windows XP home edition

Joined: 09/24/2012
Groups: Go Science
Intel Core i7-2600 CPU@3.40GHz 3.40GHz

CPU time first loop: 5,1 sec
Result report: 15129.582
Total gain: 19.905
127 loops

2 clients

Computer:
Acer Predator G3600
Processor: Intel Core i7-2600 CPU@3.40GHz 3.40GHz
64 bits
RAM 8 Go
Graphics: NVIDIA geforce GT440
OS: Windows 7 familial edition, Service Pack 1

bob1928's picture
User offline. Last seen 1 year 1 week ago. Offline
Joined: 12/16/2012
Groups: Gargleblasters
can not join

1) Click on this contest: http://fold.it/portal/node/995975
2) Go to "Link to join"
Click on "Click here if you would like to join."

closest "Link to join" is "join contest with magic key", which begets "wrong magic key".

Not very helpful

Joined: 09/24/2012
Groups: Go Science
Oops Corrected! Here is the link to join
bob1928's picture
User offline. Last seen 1 year 1 week ago. Offline
Joined: 12/16/2012
Groups: Gargleblasters
Gateway DX 4300

59 loops, 10.275 sec/1st loop
AMD Phenom II X4 810 processor 2.6 GHZ, 8Gb RAM
ATI Radon HD3200 graphic care
(64bit, quad core)
W7, 2nd client running during test
2 more clients open but static
precision wiggle for test
=========================
2nd run, fast wiggle,
2 other clients running
71 loops, avg 8.6 sec/loop
=========================
It appears the choice of fast/slow wiggle is important

Joined: 09/24/2012
Groups: Go Science
Intel Core i3-2350M CPU @ 2.30 GHz 2.30 GHz, Less Precise Wiggle

With less precise wiggle enabled, the result is as follows:
CPU time first loop: 10.6 sec against 13.2sec with default Wiggle (see above)
Result report: 15141.234 against 15125.471 pts
Total gain: 31.557 against 15.794 (more loops in same time 10' gains more per unit of time)
68 loops instead of 49 loops

Computer:

Constructor: ICT Service desk
Processor: Intel Core i3-2350M CPU @ 2.30 GHz 2.30 GHz
RAM: 4.00 Go
System: 64 bits
OS: Windows 7 Enterprise, Service Pack 1

Conclusion:: Less Precise Wiggle is about 30% faster and more rentable per unit of time on big proteins

Joined: 03/03/2013
Groups: Beta Folders
potential

Could this recipe work with other proteins?

Joined: 09/24/2012
Groups: Go Science
yes

For any own tests. (or you can create another contest - but 10 minutes might be to little for small puzzle and a powerfull computer - just edit the recipe then).

Joined: 03/03/2013
Groups: Beta Folders
Mac OS X 10.5.8

1st loop 18 seconds.
Result 15121.298
11.621 total gain
38 loops.

1 client

iMac OS X 10.5.8
2.4 GHz Intel Core 2 Duo
1 GB RAM
Processor speed 2.4 GHz

Joined: 03/03/2013
Groups: Beta Folders
Retest

10.87 seconds first loop
15125.735
16.058 total gain
54 loops.

The previous one was while working on the internet.

Joined: 09/24/2012
Groups: Go Science
Intel Core Duo T2300 1.66GHz -NewChapter Low (disable cut bands)

START SCORE IS LOWER With NC=> just compare loops and gain

Intel Core Duo T2300 1.66GHz
CPU time first loop: 26.8 sec
Result report: 11318.390 pts
Total gain: 5.062
25 loops

1 client

Computer:

Vaio vgn-sz2M/B
Processor: Intel Core Duo T2300 1.66GHz
1 processor, 2 cores
RAM: 1GB/Go DDR2 SDRAM (533MHz)
Graphics: NVIDIA geforce Go 7400 with turbochache supporting 128MB/Mo
OS: Ms Windows XP home edition

Conclusion: gain is higher in NC with Low & no cut bands

Joined: 09/24/2012
Groups: Go Science
Intel Core Duo T2300 1.66GHz -NewChapter Low (Enable cut bands)

ntel Core Duo T2300 1.66GHz -NewChapter Low (Enable cut bands)
START SCORE IS LOWER With NC=> just compare loops and gain

Intel Core Duo T2300 1.66GHz
CPU time first loop: 33.9s
Result report: 11314.846
Total gain: 1.518 Much less than with disabled cut bands !
16 loops faster than disabled cut bands

1 client

Computer:

Vaio vgn-sz2M/B
Processor: Intel Core Duo T2300 1.66GHz
1 processor, 2 cores
RAM: 1GB/Go DDR2 SDRAM (533MHz)
Graphics: NVIDIA geforce Go 7400 with turbochache supporting 128MB/Mo
OS: Ms Windows XP home edition

Conclusion: gain is higher in NC with Low & no cut bands

Joined: 09/24/2012
Groups: Go Science
Intel Core Duo T2300 1.66GHz -NewChapter MEDIUM

Intel Core Duo T2300 1.66GHz -NewChapter Low (Disable cut bands)

START SCORE IS LOWER With NC=> just compare loops and gain

Intel Core Duo T2300 1.66GHz
CPU time first loop: 33.9s
Result report: 11316.551
Total gain: 3.223 less than Low
23 loops slower than Low

1 client

Computer:

Vaio vgn-sz2M/B
Processor: Intel Core Duo T2300 1.66GHz
1 processor, 2 cores
RAM: 1GB/Go DDR2 SDRAM (533MHz)
Graphics: NVIDIA geforce Go 7400 with turbochache supporting 128MB/Mo
OS: Ms Windows XP home edition

Conclusion: gain is higher and faster in NC with Low & no cut bands

Joined: 09/24/2012
Groups: Go Science
Intel Core Duo T2300 1.66GHz -NewChapter HIGH

Intel Core Duo T2300 1.66GHz -NewChapter HIGH (Disable cut bands)

START SCORE IS LOWER With NC=> just compare loops and gain

Intel Core Duo T2300 1.66GHz
CPU time first loop: 271 sec. (oops, correction above: was 28, 34, 32 resp. for former tests above)
Result report: 11314.640
Total gain: 0.235 much less than Low and even Medium
16 loops slower than Low and Medium

1 client

Computer:

Vaio vgn-sz2M/B
Processor: Intel Core Duo T2300 1.66GHz
1 processor, 2 cores
RAM: 1GB/Go DDR2 SDRAM (533MHz)
Graphics: NVIDIA geforce Go 7400 with turbochache supporting 128MB/Mo
OS: Ms Windows XP home edition

Conclusion: Yield is higher with Lower & no cut bands
Preference is:
Cut bands unchecked > Enable cut bands
Low > Medium >> High

Joined: 09/24/2012
Groups: Go Science
Intel Core i3-3120M 2.50 GHz , Intel HD Graphics 4000, Windows 7

Cores: 4

"Windows performance" = 5.0
CPU: 6.9
RAM: 5.9
Graphic: 5.0
Game graphics: 6.3
HD: 5.9

1) Heavy conditions
3 clients (CPU usage >=100%) windows maximized.
All heavy Foldit settings possible (view everything, music, High Wiggle Power etc)

Result:

Score: 11329.335
Loops: 41
Total gain: 16.007
1 loop in 21.1 sec

2) Light conditions High Wiggle Power
1 client (CPU usage = 20-40%), window minimized
Light Foldit settings (no sound, view almost nothing) unless stil High Wiggle Power

Result:

Score: 11335.537
Loops: 110
Total gain: 22.209
1 loop in 6.4 sec

3) Light conditions Low Wiggle Power
1 client (CPU usage = 20-40%), window minimized
Light Foldit settings (no sound, view almost nothing) unless stil High Wiggle Power

Result:

Score: 11333.223
Loops: 73
Total gain: 19.905
1 loop in 6.2 sec

actiasluna's picture
User offline. Last seen 11 hours 11 min ago. Offline
Joined: 03/05/2015
Groups: Gargleblasters
iMac 3.4 GHz Intel Core i7

iMac 3.4 GHz Intel Core i7 (quad core)
Mac OSX Yosemite
NVIDIA GeForce GTX 680MX 2048 MB video card
8 GB RAM

Heavy conditions
6 clients open
Auto wiggle
Score: 9623.493 in 49 loops
Result: 15.63 sec/1st loop
Total gain 18.767

Heavy conditions
6 clients open
High wiggle
Score: 9623.537 in 51 loops
Result: 13.39 sec/1st loop
Total gain: 18.811

Will run again with just the test client open.

actiasluna's picture
User offline. Last seen 11 hours 11 min ago. Offline
Joined: 03/05/2015
Groups: Gargleblasters
Same test after update to El

Same test after update to El Capitan

Running 4 clients
Auto wiggle
Score: 9623.934 in 60 loops.
Result: 11.04 sec/1st loop
Total gain: 19.208

Since conditions aren't identical to my original test I can't statistically presume El Capitan is faster but it "feels" faster.

actiasluna's picture
User offline. Last seen 11 hours 11 min ago. Offline
Joined: 03/05/2015
Groups: Gargleblasters
... and...LWP same conditions

... and...LWP same conditions result Score 9623.934 in 60 loops same result as AWP

HWP same conditions result Score 9623.914 in 59 loops. Result: 11:32 sec/1st loop. Total gain 19.188

Joined: 06/06/2013
Groups: Gargleblasters
result of test old dell window 7s i5 machine

ran "normal for me" condition, as one of two clients. ran auto wiggle power default, minimized the protein display to hydro, line and no stubs or sidechains, reduced size and pushed protein off screen
9626.577 in 101 loops with result 7.31 seconds for first loop. so not as slow as Actiasluna with heavy load of 4 clients. my machine speed varies noticabley with protein size. Other client running the 70 AA monocyte

Want to try?
Add to Cookbook!
To download recipes to your cookbook, you need to have the game client running.
Parent
Children

none

Authors
Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons