Printable View
I posted a screenshot above of the same version donmarkoni is using, v1.02.25. This is not the version we are looking for as shown above by polroger. That version he is running is for the Formula Board but doesn't work on the RE3.Quote:
Originally Posted by R37ribution
wanna see some action with air cooling? roomtemp ~20°C
cooled with Prolimatech Megalems and 1 120mm Noctua fan
http://www.abload.de/img/i5-4.8ghz-lukuecopyin5p.png
Edit: Power measurement
- i5-2500K 4.8GHz @ 1.376v (idle 1.6GHz @ 1.40v)
- 8GB DDR3-1600 CL7 @ 1.60v
- P67 Mainboard
- 2 * MSI GTX460 1GB Cyclone (2D)
- 500GB Samsung SATA HDD
- 3 * 120mm Fans
- Zalman Fanmate
- Seasonic X-750W NT
Idle @ Desktop (81W)
http://www.abload.de/img/p10407059ikf.jpg
100% load with LinX (193W)
http://www.abload.de/img/p1040702mdv1.jpg
My new CPU
http://10pix.ru/img1/1366/2961652.th.jpg
Cold water
Soon I will do as a result of the rules and lay out, it's still a trial;)
Oh yea, I had that working with my R3G. Its nothing too special. Just a different front end for the same app you have with all the other apps integrated into it like ProbeII etc. I'll admit it does look nice tho. I just didn't like all the other stuff installed so I removed the suite and just installed TurboV EvoQuote:
Originally Posted by deachus
@chrisch is that ur cpu ?
yea, otherwise i wouldnt show screens :D
Higher memory usage actually seems beneficial to higher GFlops. It is supposed to be a stress test primarily and therefore using HT with it's lower result in GFlops is the right way IMO simply because it uses more power / generates more heat. At least in my tests it has. Typically my 860 with Linpack will run ~15% slower but use ~10% more power while using 8 threads over using 4 threads on separate physical cores with HT still enabled. IMO though it's also nice to know what sort of performance can be expected which is difficult if the resulting GFlops are skewed by bugs.Quote:
Originally Posted by Gautam
One thing i did find is that if OC'ing from Windows there is a possibility of wrong readings depending on the timer used which can result in lower than real readings when BCLK is increased. Rather than have an example showing lower clocks it is more obvious reducing BCLK as this will result in higher than real GFlops.
Example of incorrect reading using dual core with HT. BCLK dropped from 100MHz to 50MHz. In case your wondering, no I don't normally run the memory clock or BCLK that low :p:.
http://img600.imageshack.us/img600/4782/45gflops.png
Normally at 4.5GHz this dual core would be a little under 30GFlops.
Not sure what CPU this is... ;)
http://farm6.static.flickr.com/5005/...c3c15b92_o.png
Damn you guys with your teasing! I am trying to save up on some GPUs here! :rofl:
Seriously, though, looks awesome, and very promising. And just think about golden retail samples... :slobber:
SB can't come soon enough! :D
golden retails = 5GHz aircooled below 1.4v vcore :D
100mhz base clock hmmmmmmmmm:rofl:Quote:
Originally Posted by initialised
ha ha ha ha you got that wad saved for the new ati, s . . . lol me toQuote:
Originally Posted by zalbard
wrong!, 5.5 below 1.3 and can handle insane voltages ha :DQuote:
Originally Posted by Chri$ch
Another cpu submission for my W3540...
Batch #3845B010
Ambient temps: 6/7 C.
Cooling: TRUE Black with push/pull.
LLC @ 100%
Load:
http://i695.photobucket.com/albums/v...nX4C4Tload.png
Voltages measured with my Fluke MM during load...
Vcore: ~ 1.423v
Vtt: ~ 1.346v
Idle:
http://i695.photobucket.com/albums/v...nX4C4Tidle.png
This chip normally overclocks nicely with pretty decent (low) volts but it seems that at 4.5GHz+ the IMC gets rather weak and quite fussy. I found it difficult to get a successful 20 run LinX. I did try to pass with HT but had no luck... perhaps too much stress for the IMC?
Now with UD4 and also air...
voltage reading is wrong in CPU-Z but ET6 shows it correct :)
http://www.abload.de/img/4.8ghz-airhfax.png
http://www.abload.de/img/p1040719rd11.jpg
Sandy Bridge is looking good!Quote:
Originally Posted by Chri$ch
yes it is, good enough to sell off a bunch of stuff:ROTF:Quote:
Originally Posted by PolRoger
@chrisch can u run some vantage benches and some super PI scores for comparison hehe
Nice Chri$ch, 4.81 @ 1.38v aircooled. Insane.
Working on it.Quote:
Originally Posted by Chri$ch
what a retail can do will i see in a few minutes :D
http://www.abload.de/img/p1040735lbdv.jpg
http://www.awardfabrik.de/forum/show...&postcount=124
Last time I listened to that kinda talk in this thread I ended up toasting a 980 ES!Quote:
Originally Posted by Johnny87au
Surely it didn't happen at 1.3V! :DQuote:
Originally Posted by initialised
1.3V is hardly insane and not quite enough for 4.5GHz on these.
Hi!
What is considered max safe voltage for SB on air - for benching and for 24/7?
My retail i7 2600K (batch: L038A646) is doing 4.6GHz 1.332V load, 1.344 idle in Win, 1.35V in bios, temp in low 70s, HT on.