some run benches with classic AIR and normal ambient

[Tomasis]

1,3v ..wow pretty low!

[richierich]

Nice, looks like you got a 1090T that's binned for almost 95W. What's the stepping?

[FlanK3r]

1015 DPMW, the same, which MAD222 has (he got 6950 MHz with it)

[Riska]

Nice cpu you have there..
Mine could do 4.2ghz at 1.306v on water prime stable. But it is sold now.

http://i438.photobucket.com/albums/q...ore42GHZ-1.jpg

[richierich]

Nice cpu you have there..
Mine could do 4.2ghz at 1.306v on water prime stable. But it is sold now.

What stepping? I sound like a broken record lol.

[BeepBeep2]

Holy bajeebus, 4.2 stable at 1.30v. Wow.

[FlanK3r]

hm, very nice Riska, maybe mistake sold it :-(. Now im waiting at MAD222 gold chip, he promise me max air stable clock and thinking, he can got 4350-4400 MHz (gold chip, he got 6.96 GHz !)

[Tomasis]

even more impressive

[FlanK3r]

riska? what temps do u had?

[Mad222]

I remember my promise, Flankerp!
Will get back my 2nd 1090T this week from the lcoal media
You know how much I paid for the power bill last month?....USD 300
That's why I dun do much 3Dbenching @ home...I'm broke

hm, very nice Riska, maybe mistake sold it :-(. Now im waiting at MAD222 gold chip, he promise me max air stable clock and thinking, he can got 4350-4400 MHz (gold chip, he got 6.96 GHz !)

[Riska]

I had 40-45c max my water is made of 2x Xspc RX360 radis in a loop and a laing 10w pump with xspc res/top on and a xspc Delta3 coolinghead for cpu and 6x zalmans zmf3 fans at 5v

[FlanK3r]

MAD:dont kill chip under LINX, rules voltage max 1.55V for CPU, ok?:-D
UFff, so much money as my holiday to Italy...

[Mad222]

Flankerp
Just tried for a few mins with only 1g available memory
What do you want me to do ?
Only have 790FXTA-UD5 @ home
Wow...it's hot, 44c @ 100% full load, man!

[FlanK3r]

nice bro, u are lucky bastard !

Set higher problem size (in x86 with 3.5GB its about 14 300) and again 20 loops (your temps will bit higher ) Whats your ambients celsius? I cant broken still more than 4300MHz, maybe some setings, dont know. Show me next LINX, il so happy

[Mad222]

Ok, Flankerp
4.36g error on loop 14/15
reduced 1 fsb & had no problem
19c ambient on air

[crazydiamond]

holy !! xtremely impressive!

[FlanK3r]

very good MAD! More voltage doesnt helped u? This chip looking +40 MHz better than my :-). Im looking, u have good cooler, temps (i think real temps, Coretemp show at some chips 10 C diference) max about 54 C.

Diamont:think, this is maybe the best AIR cooling Thuban the world.

[Mad222]

Hahah...
I was using yasya with GC2 & air condition on
The main pt is the CPU itself that doesn't need high volt.
Anything higher than 1.488v would get me an error @ loop 14/15
With lower NB like 2600-2800, I think 4.4GHz Linx stable can be done !
This is the weird chip I mentioned b4....boot up to Windows @ 4.56g with NB 3180/Ram 1800+ & complete PI8m/16m without any problem, but can't do well on LN2
I thought this was a 7GHz chip, I'll try another LN2 session with this chip @ diff. setting & volt. to see if it can hit 7GHz

PS: My kids have just gone to bed, I can do some more Linx run now !

[FlanK3r]

whau, try max LINX with default NB clock :-). Im curios how much better than tuned chip. We have the same problem, i tried higher voltage and got to errors :-)

PS: and yes, your PS like me very much (maybe one from LINX testing, how much u can get with only 1.3V stable, how much with 1.4, 1.425 and 1.45V.It will be also interesting)

[Mad222]

Can't pass 4.4GHz
Temp. higher than 45c by corertemp reading would give erorrs
Now running 4.38GHz @ 14135 problem size for 10 loops (just did a few times & Linx stopped @ loop 10...damn)

[FlanK3r]

its hard time tweaking LINX stability (maybe harder than Cinebench runs ) Now i have here a bit HOT, so for me is today 4300+ to close

[Mad222]

10 loops is just fine
Need to do better Temp. control in order to break through 4.4GHz @ bigger problem size & 20 loop stable

[FlanK3r]

whau, and freat temps now

[FlanK3r]

Today im posting boring result with default voltage and good stable clock x4 965 BE C3. At weekend was a bit hot, so 1.392V is ideal for CPU life :-). Second, in this weather i have no chance broken benchmarks at highest setings+ think, i can got higher stable clock with ambients about 22C than yesterday... Setings:
Gigabyte MA790X-UD4
2x 2GB Kingston Hyper X 1066MHz cl5
AMD x4 965 BE C3
CoolerMaster Hyper 212
ATI HD3870 512 MB
Windows XP SP3 32-bit

First setings at screen lower wasnt the best, but im tetsing not often in PRIME95 :-). So 3698 MHz + 2630 MHz at NB and 1.392V


LinX: 3794 MHz and 2710 MHz at NB


But benchmarks was runing by me only at stablec clock 3780 MHz + 2700 MHz NB.

Everest memory benchmark


Time for superpi 1M :-)


And for 32M superpi:


Looking at the tests with the archives, first winrar bench


And 7zip benchmark


Now some next multicore calculating and memory benchmarks.
MaxxPi 4M


MaxxPi 16M


MaxMEM


MaxxFloops-like this benchmark now


Science benchmark for scientists


Geekbench


Fritzchessbench


wprime 32M and 1024M


3D and rendering in Cinebenches :-)




At latest i tested x264 benchmark, but results are strange (dont know, what version of x264 benchmark is it)


To end i tested only few configurations, how good is IMC with CPU-NB at 1.425V? I got 2827 MHz at NB, think, it will stable


.. and in LInX



And than quickly benchmarks at 4100 MHz




[IMG]