[Team.AU]Gigabyte X58-UD5, 3GB Corsair 1866MHz CAS8, Core i7 920 Testing thread

[dinos22]

G'day fellas

Time for some new exciting and fresh silicone torture sessions to begin.

What's getting tortured......

Test Configuration:

• Gigabyte X58-UD5 F2/F4E bios so far
• Intel Core i7 920 retail chip 3837A899
  (onlinecomputer.com.au have lots of stock for Aussies)
• 3GB 1866MHz CAS8 DDR3 Corsair Dominators (TR3X3G1866C8D)
• 1GB Gigabyte GTX280 Graphics Card
• 1000W HX1000 Corsair PSU
• TTBig Typhoon with custom plexi hold own
• (hold down sucks...weak pressure on CPU due to plexi bend)
• 1hp C-BuZz Single Stage (kayl evap) > later
• 160GB IDE Hard Drive
• Windows XP Pro 32bit SP3
• Windows Vista Ultimate SP1 for Vantage
  

WHAT I PLAN TO TEST?

Cooling
With that out of the way i'm planning to do some air testing at first then followed by SS to see how different cooling affects the CPU

RAM/board
got a nice tripple channel kit from Corsair.......planning to test CAS6/7/8/9 OCing and its performance with 32M SuperPi. Once i find the optimum performance efficiency settings i'll do 3D testing as well. It will be interesting to see how IMC responds to voltage and general settings changes. I'm hoping that FELIX will have a full version of Memset out soon for Nehalem which allows subtimings changes in windows

32M SuperPi category testing for 32M SPi Low Clock Challenge thread to set some groundwork for me to relate back to in future testing. What's that? You never been to that thread..... ...better go say hello in the thread and start tweaking that system of yours with SuperPi

Why SuperPi? What is this obsession with a single threaded application you may ask? There are a lot of angry ppl on forums saying that this application is useless and they wonder why ppl bother testing it at all. Even some well known industry ppl (probably politics related there hehe). The bottom line with superpi is that this application is a must for any serious overclocker. If you dont test a lot with superpi you do not know your system efficiency properly imo whether you only bench that app or 3D or whatever. This is a base application that helps tweak the system. if you want superpi stability on multiple cores you run hyperPi basically which will run multiple instances but i am not talking about stability but system tuning. Please keep that in mind

RAM cooling >> AIR vs TEC RAM Cooling with tripple channel DDR3




Forum Member Suggestions
if there is something interesting you want to see i'll do my best to incorporate it

Let's inject some colour here....almost fell asleep myself writing this

[dinos22]

STOCK TESTING ON AIR COOLING

with XMP Profile loaded to run RAM at stock
during load like superpi or hyperpi or prime the CPU was operating at 2833MHz

i will have to retest this 1M as the PP efficiency below 39K seems awefully good...it may be that 1M was actually still running at 22x multi

1M SuperPi


32M SuperPi


8xinstances of 1M SuperPi (via hyperpi)


8xinstances of 32M SuperPi (via hyperpi)


PiFast


Stock AM3 (single GTX280 no OC)





1 Hour Blend Prime before stopping
temps were measured with Speedfan but seeing how low they are compared to others here i will have to use coretemp next time


After stopping prime
idle temperature drops a lot (ambients were below 20C)



Overclocked to 4.2GHz.....let's see where we're at
btw RAM was running at 1000MHz 8-8-8-20-1T @1.68vdimm

1M SuperPi 9.765s


32M SuperPi 8m 38.250s


Pifast 20.84s


I tested the new F4E bios and has some issues overclocking only to finally realise it was uncore related and bios fix is on F4F bios

i did test a lot of different runs as i did this 32M to improve on my last time

32M SuperPi 8m 29.750s on air is really something else
amazingly fast CPUs i have to say...this is equivalent to around 5.7GHz E8600 with similar kinds of tweaks

[dinos22]

Did some more testing last night
it keeps getting better and better

First of all some bios screenshots if you havent had a chance to look at them. I only took shots of interesting things for overclockers















That last screenshot is interesting. Since i was moving to new bios from F4E to F4F i decided to move my saves files to USB stick and reload them in new bios >>> yup it worked but you do get a warning that you are loading a bios setting from different bios etc so do be careful there ..... only issue i have is that the description for file is probably too short...i would like to see a much larger sizes description so i can put some detail in there to remind me what i did with the bios settings at that time


__________________________________________________ ____________________________


Some setup screenshots

Top temp > Ambient ~22C
Bottom Temp > Water temp at idle after a bit of work


You like my new waterblock mount


Water setup pics
OMG Nehalem is HOT...i ran that realtemp software and during prime it was in 80C at 4GHz 1.41v
water temp kept rising and rising....i have to check the rad for dust


Vdroop mod




Phase Change setup




__________________________________________________ _________________



Some results from last nights testing

Max Water Screenshot 4485MHz
http://valid.x86-secret.com/show_oc.php?id=453923



Some Phase Change results

I started off with looser settings first to get a feel for frequency









Surprisingly fast 3DMARK05.....looks like Nehalem FTW here




I started tightening timings here and also later bumped frequency a touch on CPU

[dinos22]

Fastest 32M SuperPi on Core i7 920 in the world

look at the NB Speeds
this board is on fire










Bus Frequency testing (QPI/BCLK)


Here is the full update

I did some QPI/BCLK/bus frequency testing to see if i could break the 222Mhz "wall" I had in my previous tests. It was a bit of a tussle but in the end i managed to get some extra bus speed through few different settings outlined below. It seems as thought PCI Express frequency, uncore "sweetspot" volts and settings you boot at are the main driving factor from my initial testing in achieving higher/stable bus speeds or QPI or BCLK or whatever people call them. I was maxing out at 102.5MHz PCIe frequency in windows and while i also managed to find a sweet spot to let me OC a touch more (103 or so MHz) i was not able to make a run stick even though i did have a CAS6 run running at 228.3MHz bus only to crash on 16th loop...it looked like a 7m30s run ...

Here are my findings so far. Disregard the SuperPi efficiency.......i was just testing frequencies.

I'd love to hear from Gigabyte X58 board owners about their experience and whether these settings help anyone reach higher clocks.

I'm in touch with Gigabyte guys about this so hopefully they can figure out what it is exactly, whether it's the PLL chip or signaling/traces issue or whatever. All i care about is 250+Mhz bus speed stable on any GFX card

Test Configuration:

• Gigabyte X58-UD5 F4n bios (old bios for initial testing)
• Intel Core i7 920 retail chip 3849A910 D0 stepping
• 6GB 2000MHz CAS8 DDR3 Corsair Dominators
• Nvidia 8400 PCIe Graphics Card + Xfire ATI 4890
• 1000W HX1000 Corsair PSU
• 1hp C-BuZz Single Stage (kayl evap) > on CPU
• 160GB IDE Hard Drive
• Windows XP Pro 32bit SP3
• Windows Vista Ultimate SP1
  

BIOS SETTINGS:

• CPU PLL Voltage: Auto or 1.80v - same effect
• CPU VCORE: 1.35-1.36 - same effect
• IOH/ICH 1.1V Voltage: Auto or 1.20v - same effect
• IOH Voltage: 1.42v (i havent tried auto...dont think it matters)
• ICH PCI-E Voltage: Auto or 1.5v - same effect
• QPI/VTT: 1.433v or 1.455v = this chip's sweet spot (any higher or lower and i was unable to push past 101MHz PCIe frequency in windows)
• VDIMM: tried between 1.66v - 1.94v same effect
• BUS SPEED: booting from 200 or 210 same effect, lower or higher i had problems going past 101MHz PCIe frequency in windows
  for the record i could boot at 220 straight from bios to windows...sorry didnt try higher..will do next time
• PCIe FREQ.: Auto or 101MHz - same effect....102 would occasionally boot but every now and then throw an F6 error iirc...anything higher and no boot at all. In windows i did not see 101Mhz PCIe freq. in SetFSB even though i was selecting it in bios so not sure it was actually sticking at all.

222MHz Bus Frequency
101MHz PCI Express Frequency
32M SuperPi stable

recorded in the middle of the run (before i moved on to higher frequency)




223MHz Bus Frequency
101MHz PCI Express Frequency
32M SuperPi stable





225MHz Bus Frequency
101.8MHz PCI Express Frequency
32M SuperPi stable




226MHz Bus Frequency
101.9MHz PCI Express Frequency
32M SuperPi stable






227.5MHz Bus Frequency
102.50MHz PCI Express Frequency
32M SuperPi stable






http://valid.canardpc.com/show_oc.php?id=568106







update!!!!


NB Frequency testing


and just when you think you've seen it all there is some more action to report

i was pushing the memory controller last night to see what this sucker can do on my favourite i7 based board (Gigabyte UD5) and i am mighty impressed with results so far

check out the 4.9GHz NB frequency and 889Mhz 6-6-5-16-1T (tight subtimings on RAM as well)

i was just trying to get some rough idea what sort of performance i could get at 4GHz as there is a nice little challenge at OCX

go you little beauty

time will be more efficienct when i get around to redoing my OSes




i did a 1M SuperPi for good measure but my XP is so screwed up that i ended up on Vista for 1M and the efficiency is mighty nice considering the circumstances

[systemviper]

Sweet, thanks for posting and testing, that is almost my exact config,
X58 Extreme
i7 920 retail chip 3837A728
Kingston hyperX 2k
Evga GTX280sc.
ToughPower 1200w modular

everything should be here monday.
Please post bios settings if possible....
Will you do any 1M super Pi..?

thanks for your time and efftort!

Subscribed...

Another question, My goal after running some benches is to get a 4Ghz 24/7 system. It seems that most 4G overclocks are in the range of 1.4x Vcore. know it's early to tell but do you think that is too high Vcore for 24/7system? I will be running 24/7 on water.

[dinos22]

hi mate
thanks
yes i will take some screens of bios and post up a template soon

[T_Flight]

These Dominators Clock awesome! You have confirmed what I already realized about the i7 platform. A 4-4.2 OC is like running previous gen CPU's at over 5GHz.

I have the 1600 Dominators and want to see if they have just as much in them. If I get to 1800 that would be fabulous, if I got 2000 that would be incredible and I would be stunned! These Dominators really are awesome. They really ARE Dominators. This new low voltage RAM is incredible that it can run that low of a voltage and stay stable.

Thanks for the testing Dino's. I'm really impressed. I'm really anxoius to start tweaking myself. Never in my life would I have believed we'd see the numbers we are seeing.

[Hennels]

Is that windows 98? How much of a benefit does it have over other OS's?

[T_Flight]

Looks like XP in Performance mode with all the eye candy turned off. It takes at least XP SP2 to support PCI-e, so I doubt it's 98. I don't know whether you can set Vista in classic performance mode or not. i haven't messed with it enough. I'm just now learning Vista.

[slim142]

Hey Dinos
What ICs are inside those dominators?

[dinos22]

i would guess and say Samsung based considering that some of the other brands with high clocking ICs at low volts tend to use similar

maybe yellowbeard can shed some light if he pops into thread

confirmed XP with classic theme

[freshy]

thanx dinos, screens of bios and a template sound wild, i have got my ud5 , just waiting on my 920 and 3gb patriot 1600mhz kit and gtz/mount to get here,

thanx again

[slim142]

i would guess and say Samsung based considering that some of the other brands with high clocking ICs at low volts tend to use similar

maybe yellowbeard can shed some light if he pops into thread

confirmed XP with classic theme

Yeah, Yellowbeard would probably have that answer

[TheKarmakazi]

Awesome testing Dinos, great 32M! Is copy wazaa influencing very much with this new architecture?

These Dominators Clock awesome! You have confirmed what I already realized about the i7 platform. A 4-4.2 OC is like running previous gen CPU's at over 5GHz.

In certain benches that can make use of such huge memory bandwidth (superpi) or all the threads (3dmark vantage, wprime).

Other benches you are better off with a much higher clocked wolfdale (aquamark3, 3dmark01,03,05)

Now if Intel can get these chips clocking in the 5.5ghz range (and im not talking about the super cherrys i mean regular chips), then were talking about some serious damage!

[RPGWiZaRD]

i7 looks more tempting day after day, and man do I love this DDR3 kit, can't be very cheap.

What do you think about this board being a "slight" Gigabyte fanboi Dinos?

[dinos22]

Did some more testing last night
it keeps getting better and better

First of all some bios screenshots if you havent had a chance to look at them. I only took shots of interesting things for overclockers















That last screenshot is interesting. Since i was moving to new bios from F4E to F4F i decided to move my saves files to USB stick and reload them in new bios >>> yup it worked but you do get a warning that you are loading a bios setting from different bios etc so do be careful there ..... only issue i have is that the description for file is probably too short...i would like to see a much larger sizes description so i can put some detail in there to remind me what i did with the bios settings at that time

[dinos22]

Some setup screenshots

Top temp > Ambient ~22C
Bottom Temp > Water temp at idle after a bit of work


You like my new waterblock mount


Water setup pics
OMG Nehalem is HOT...i ran that realtemp software and during prime it was in 80C at 4GHz 1.41v
water temp kept rising and rising....i have to check the rad for dust


Vdroop mod




Phase Change setup

[dinos22]

Some results from last nights testing

Max Water Screenshot 4485MHz
http://valid.x86-secret.com/show_oc.php?id=453923



Some Phase Change results

I started off with looser settings first to get a feel for frequency









Surprisingly fast 3DMARK05.....looks like Nehalem FTW here




I started tightening timings here and also later bumped frequency a touch on CPU

[GAR]

HAHAH, dinos your water block is 1 of a kind hahaha, good creativity!!

[dinos22]

hey it even doubles up as my TT Big Typhoon mount lol

[BlindFreddie]

Hi Dinos

Interesting stuff ...
I'm in wait-and-see mode before I buy into i7, but ...
I'm surprised to see that you're using a Nautilus wc setup. I've got one on my Q9650/P5K3 dlx, and I'm not that impressed. The radiator quality isn't that great, and I might even put another rad in the loop.

Meanwhile, I've taken the grille off the top to reduce the restriction of airflow, tho I haven't measured temps before and after. An easy mod that should improve things. I'll put a standard open-mesh wire grille on it at some stage. I might dig out the standard fan and sit an UltraKaze on top of the box first tho.

My other suggestion is to replace your styrofoam waterblock-holddown dooverlacky with 3 bits of wood, available cheap from a rubbish skip out the front of some renovator's dream home near you. 1 piece across the wb, 1 piece across each end of that, with holddown-bolts thru the ends. That should solve your problem of lack of pressure onto the CPU. You can glue and/or screw extra bits on to hold your Big Typhoon if needed.
HTH

[dinos22]

hi mate

i like using Nautilus 500 because of quick release feature in particular
you are absolutely right...........this aluminium radiator is rubbish and you would probably need a new rad to keep the temps down but i do have a tripple rad in the bench room which i've used in my GFX Card benching so that is what i plan to do

nautilus 500 could not hold the temps last night while i was priming at 4GHz with 1.41-1.44vcore as i was tuning the for droop but i am not 100% sure if it is a problem in the setup or too much dust on fan.....lol you would think i could get off my arse and check dont ya....i did have a quick glance from bottom side and while there was some dust there wasnt a heck of a lot

i will inspect it as i did find it weird to see water temps just climbing and climbing with no sign of going down.......i would stop it 10 minutes into prime as the water temps would jump as much as 7C by then eek

i've never had issues with N500 holding temps and would not be measuring water temps if i didnt actually touch the tubing to feel its luke warm and thought WTF is going on here...i've never had such HOT CPUs before thought too

in terms of hold down i will cut a 2mm metal piece so that i have a proper hold down but pressure was better with this wood compared to my plexi hold down i made last week

Dino

[Oliver]

good work Dino and once again congrats on sub 8m 32m with 920

the 1866c9d retail use HCF0 ICs, it will shock me if your kit use other ICs

keep pushing it

[sammurai]

Very god dino.

I have the same mobo and i can`t pass over 190 mhz. I increased the cpu voltaje to 1.4v, QPI to 1.4v and IOH to 1.3v but nothing.

My CPU run 1M@3.9ghz@1.264v

[dinos22]

it took me a while to get here as well man

you have to sit down and spend a lot of time experimenting

F4F bios seems to be easiest to manage so far so try to boot with that one direct to 200 or 210 without setting any volts other than 1.4v for QPI/VTT

sometimes that does the trick

also another thing i found
make sure if you are setting timings manufally that you select NB speed at 4x RAM speed

so 800MHz RAM > 3200MHz NB Frequency