*STABLE* Black Hole Benchmark
- Thread starter Virssagòn
- Start date
Virssagòn
VIP Member
Had to install Windows 7 on a Fujitsu MX130 S2 server today, thought I'd run BH
Seems very low, but then again, it's an underclocked 8350
I've to admit that this is a very very low score. Any chance of getting the system run BH a second time?
turbobooster
Member
difference quad channel and dual channel.
oke did a blackhole test with dual channel 16gb and 16gb quad channel.
exactly the same settings for the rest.
the lowest score is the dual so the 1 on the right.
oke did a blackhole test with dual channel 16gb and 16gb quad channel.
exactly the same settings for the rest.
the lowest score is the dual so the 1 on the right.
Virssagòn
VIP Member
Kaveri does pretty good at higher clocks. The only problem is that up from 4.6GHz, you'll need to increase the voltage by very much to make it stable again... So Kaveri is maybe an improvement in IPC, but certainly not an improvement in overclocking abilities, so overclocking headroom stays the same. Richland has a higher base clock, but can reach 5GHz easily. Anyway, could have been the heat as well, but my temp sensors seem to be faulty.
After exams I'll go for at least 5GHz!
After exams I'll go for at least 5GHz!
Okedokey
Well-Known Member
These are the systems I'll be testing after exams:
Got them here already, wish me luck with those reviews
Nice man, just make sure you give it epic cooling and volts!!! I wanna see a fire!
Virssagòn
VIP Member
Well, these scores are including 2 laptops and 1 locked i7 4790, so you won't see much fire there... Or do you mean that Kaveri setup? I'll be putting an open loop on the kaveri, like I did with richland when I ran blackhole on 5.2GHz.
turbobooster
Member
24/7 overclock
turbobooster
Member
it will, this is just my 24/7 overclock, problem for a higher overclock 24/7 is the corsair h75
Okedokey
Well-Known Member
You have it in push pull fan installation?
Make sure Qfan is disabled in the bios or connect the CPU fans to a direct 12V source (e.g. molex to fan converter) to ensure they're at max and see what you get for temps that way. Also putting good quality TIM on every so often helps.
I have a 420 mm (10 cm thick) and a 240mm rad on my CPU so i can run it at 5GHz less than 15oC delta. But boy, does it warm up my study!
turbobooster
Member
Yes its push pull, and when I run benchmarks or stresstest the fans are running at 100%
At the moment the TIM what came with the cooler its on it, normally a use cooler master IC Essential E2
Okedokey
Well-Known Member
Yeah well do yourself a favour, get some Arctic Silver 5 and replace that TIM. I have seen up to 10oC drop doing so...
Virssagòn
VIP Member
Doesn't help that much though. In my tests the Chill Factor III from thermalright came out as best TIM. I tried these: Arctic Silver 5, be quiet! DC, default scythe paste, default CM TIM (came out as worst with a 6c difference compared to chill factor III) and noctua's default.
Be aware that every paste needs up to a week regular use to run on its full potential. Especially the Arctic Silver 5
turbobooster
Member
tried the artic silver, after that I bought the cooler master gold, temp dropped for me, but run out of it, so that's why I have the thim on what came on the cooler