Black Hole - Benchmark (OLD Version)

Status
Not open for further replies.

Virssagòn

VIP Member
Some results:



73da0532_bh_51x102.5.png


5c363341_BlackHoleBench-22.01.2013.PNG


3220a422_Capture2.PNG


6c0d8258_vj0D9wu.png


6f893f7f_BlackholeBenchmark.PNG


9002355d_Best1.PNG


bh2600k.jpg


bh990x23460.jpg


bh3930k.jpg


Just wanted to post for some interested people :D
 

Virssagòn

VIP Member
Timings make sense, but not much in this version (previous version was very sensitive in timings, dtna found that out)

9-9-9-24
99924.jpg


7-7-7-22
77722.jpg



@dtna: how much did you lower the timings? And what timings should have more effect? I'm not really into ram things...
 

87dtna

Active Member
Higher ram speed with looser timings made more improvement than lower speed with tighter timings (to a point, 1333 cas 7 is still gonna beat 1600 cas 9 as you showed).

People with 2133 ram will have a fairly large advantage, particularly if it's cas 9 ram as well. 2133 VS 1600 both cas 9 would yield anywhere from 500-1000 points better.
 

Virssagòn

VIP Member
Higher ram speed with looser timings made more improvement than lower speed with tighter timings (to a point, 1333 cas 7 is still gonna beat 1600 cas 9 as you showed).

People with 2133 ram will have a fairly large advantage, particularly if it's cas 9 ram as well. 2133 VS 1600 both cas 9 would yield anywhere from 500-1000 points better.

I thought you said lower timings was better then increasing ram speed? :S
How much voltage can I give my ram to not damage?
my hyperX ram needs more then the ripsjaws... that's not really handy :D
 

87dtna

Active Member
To a point. You can see 1333 cas 7 was better than 1600 cas 9. But 1333 cas 8 would probably be not as good at 1600 cas 9.

Once you go over 2000 ram speed though the results were highly changed. I believe my 2000mhz cas 9 score was ~800 points higher than than 1600 cas 9 and 1000 points higher than 1333 cas 9. This was with the first V4. 4.1 seems seems less affect and I'm not sure about 4.2.
 
Last edited:

87dtna

Active Member
I just downloaded and started testing the beta 4.2, and the 4 threaded test is only using 2 threads and my score is horrendous! lol
 
Last edited:

Virssagòn

VIP Member
I just downloaded and started testing the beta 4.2, and the 4 threaded test is only using 2 threads and my score is horrendous! lol

Try running again, I got someone with the same problem and his was solved with another run for some reason...

On your cpu it normally runs at thread #0 2 4 6. Because that are the physicals and give best times...
 

87dtna

Active Member
lol it's messed up. I ran it again, this time it started with cores 2 and 6 running, and halfway through the 4 thread test core 0 bounced in there.

First time I ran it my 4 thread score was 1600, now its 2600. Should be more like 5k looking at other results so I'll try again.
 
Last edited:

87dtna

Active Member
Finally lol. Though I'm not sure how an I5 760 at 3.7ghz scores 12k and the I7 860 at 4.2ghz only 2k more.

The multithreaded score seems quite low, like I don't even have HT on.

BH42I786042ghz1910mhzram.jpg
 
Last edited:

Virssagòn

VIP Member
Yeh, the bench don't let much difference between i5 and i7. I5 2500k @ 5.2 and i7 2600k @ 5.3 only had a rough ~1100 point difference.
Someone with i7 950 @4.2ghz scores 14356. Mine @4.3ghz 15488, so there is a pretty nice difference between the old i7 and the new. :p
 
Status
Not open for further replies.
Top