Multi-Threaded Super-Pi Contest

Jet

Jet

VIP Member
Information about this Benchmark:
http://www.xtremesystems.org/forums/showthread.php?t=221773

Download from here:
http://www.numberworld.org/y-cruncher/

Warning:
This program can be more intensive than prime95.
Version 0.4.3 stress test runs hotter than both prime95 and Linpack on Core i7 with HT enabled.
Click to expand...

Rules
-Post screenshot of your 32 million digit time and your 256 million digit time:

Instructions
Download file, unzip, and then open y-cruncher.exe. Enter the following:
Option 0: (Benchmark Pi)
Option 1: Multi-threaded
Option 0: Choose SuperPi Sizes
Then, choose 32M. Run that, take a screenshot, and then do the 256M and 512M after that! (Warning: you might find your overclock isn't as stable as you think it is!)

32M Standings:
1. 7.526 -- 87dtna (Xeon @ 3785)
2. 8.146 -- Jet (i7 920 @ 4624)
3. 9.159 -- Jet (i7 860 @ 4017)
4. 9.472 -- THERMAL REACTOR (i7 920 @ 4397)
5. 9.638 -- Jet (i7 920 @ 4208)
6. 9.828 -- jasonn20 (Phenom II X6 @ 3800)
7. 10.462 -- Sam_VDC (i5 750 @ 4410)
8. 10.623 -- 87dtna -- (i5 750 @ 4300)
9. 12.507 -- The_Other_One (i7 920 @ 3320)
10. 13.394 -- linkin (Athlon II X4 635 @ 3625)
11. 13.449 -- jevery (Q9650 @ 4230)
12. 15.384 -- bigrich0086 (Q9450 @ 3600)
13. 15.544 -- 87dtna (Phenom II 550 (unlocked to X4) @ 3825)
14. 15.600 -- OMEGA (Q9550 @ 3825)
15. 16.701 -- G25r8cer (Phenom X4 920 @ 3616)
16. 17.283 -- Intel_man (i5 750 @ Stock)
17. 17.981 -- Nevakonaza (Q6600 @ 3717)
18. 23.319 -- 2048Megabytes (Phenom II X4 940 @ 3000)
19. 24.122 -- Glliw (Phenom II X4 940 @ 3000)
20. 24.756 -- Mark50 (E8400 @ 4203)
21. 33.707 -- Michael (E6600 @ 3285)
22. 43.372 -- Jet (P8700 @ 2520)
23. 43.835 -- Michael (P8600 @ 2400)
24. 47.640 -- mx344 (Phenom X3 8750 @ Stock)
25. 51.109 -- 2048Megabytes (Athlon II 245 @ 2907)
26. 59.084 -- 2048Megabytes (Athlon 7750 @ 2700)
27. 92.787 -- 2048Megabytes (Athlon X2 4600+ @ 2400)
28. 385.511 -- The_Other_One (Atom N280)
29. 570.421 -- Drenlin (Celeron @ 2791)

256M Standings:
1. 76.032 -- 87dtna (Xeon @ 3785)
2. 94.356 -- jasonn20 (Phenom II X6 @ 3800)
3. 96.330 -- Jet (i7 860 @ 4017)
4. 98.593 -- THERMAL REACTOR (i7 920 @ 4397)
5. 103.614 -- Jet (i7 920 @ 4203)
6. 116.369 -- Sam_VDC (i5 750 @ 4200)
7. 117.153 -- 87dtna (i5 750 @ 4200)
8. 128.117 -- The_Other_One (i7 @ 3320)
9. 171.357 -- 87dtna (Phenom II 550 (unlocked to X4) @ 3825)
10. 177.891 -- Intel_man (i5 720 @ Stock)
11. 189.166 -- G25r8cer (Phenom X4 920 @ 3214)
12. 197.598 -- Nevakonaza (Q6600 @ 3717)
13. 280.233 -- Mark50 (E8400 @ 4104)
14. 382.398 -- Michael (E6600 @ 3285)
15. 475.326 -- Jet (P8700 @ 2520)
16. 480.453 -- Michael (P8600 @ 2400)

512M Standings:
1. 207.906 -- jasonn20 (Phenom II X6 @ 3799)
2. 222.382 -- Jet (i7 860 @ 4017)
3. 226.908 -- Jet (i7 920@ 4202)
4. 255.760 -- Sam_VDC (i5 750 @ 4200)
5. 257.275 -- 87dtna (i5 750 @ 4200)
6. 281.771 -- The_Other_One (i7 @ 3320)
7. 382.560 -- 87dtna (Phenom II 550 (unlocked to X4) @ 3624)
8. 425.686 -- Nevakonaza (Q6600 @ 3717)
9. 623.840 -- Mark50 (E8400 @ 4104)
10. 1045.970 -- Jet (P8700 @ 2520)
11. 1069.610 -- Michael (P8600 @ 2400)
 
Last edited:
8

87dtna

Active Member
Is this what you want? Not sure if I did it right-

PItest.jpg


Test was done with my Phenom II 550 unlocked to an X4 at 3.8 ghz.
 
Last edited:
linkin

linkin

VIP Member
I think it should be 1m times like the other non-multithreaded superpi. of course if it ends up being like a 256k test (short and not much difference) we can go higher.
 
poke349

poke349

New Member
*Slight hijack with a little insider info...* ;)

I'm in the process of putting together a specially tuned binary for my roommate's newly built Phenom II X3 rig.

Assuming it gives a worthwhile speedup for AMD, v0.4.4 will be out in a couple of days - which should make this benchmark a little bit more vendor-neutral. (It's very Intel favoring right now since there are tuned binaries for both Core 2 and Core i7, but none for AMD.)


Aside from that, everything else will be the same. (Speed-wise.)


As far as benchmark sizes go... I honestly think 1M is too small. It's too short, too inconsistent, and too small to benefit much from multi-threading.
I would suggest one small benchmark (25m or 32M), and a large one (1b or 1G). Which should be analogous to SuperPi 1M and 32M. (as far as run-times are concerned)
 
Last edited:
Shane

Shane

Super Moderator
Staff member
I would contribute to the thread if i could find out how the hell to run the damn tests :D

Why didnt they have make it simple like SuperPi? :eek:
 
poke349

poke349

New Member
I suggest you include version #s in the ranking list. Or you can request that only v0.4.3 or v0.4.4 be used. (v0.4.4 will be coming out in a couple of days)

This benchmark is an evolving program. It'll be updated over time to utilize new hardware features (such as SSE, AVX, etc...).

So speeds may or may not be consistent between versions. (AMD K10 is gonna get a boost with v0.4.4.)


Enjoy ;)
 
Last edited:
8

87dtna

Active Member
Come on man how many times are you going to change it??? I just spent the last hour getting my system stable enough for 512m to test good. Now I come back to the thread to post me results and you've changed it to 256.

Is it gonna freakin stay at 32m and 256m now???
 
Jet

Jet

VIP Member
87dtna said:
Come on man how many times are you going to change it??? I just spent the last hour getting my system stable enough for 512m to test good. Now I come back to the thread to post me results and you've changed it to 256.

Is it gonna freakin stay at 32m and 256m now???
Click to expand...

Hahaha!

I'm sorry--I was just getting the bugs worked out. I found out that the 512M requires 64 bit and over 2GB of ram, so I thought that would limit who could do it--so I went with the 256M. If I hadn't seen that, I would have kept it at 512M! I hope you understand...
 
Shane

Shane

Super Moderator
Staff member
Jet please update teh scoreboard....just wait until the i7 users come then it will get fun :p
 
You must log in or register to reply here.
Top