Testing Report: Gigabyte X48T-DQ6 Rev. 1.1

Hello guys,
I'm starting "experimenting" with the Gigabyte X48T-DQ6 Rev. 1.1 to get to know it well before I start reviewing it, I thought that it'd be nice to open a thread here to show my findings and of course share them with you ( and you with me of course :p: ).

"Owners" of the board ( PCB Rev1.1 plz ;) ) feel free to chime in and report your experience, findings, hints, etc ;)

I'll start with a mini photo presentation and then I'll start reporting things as they "occur" to me :D

http://i27.tinypic.com/2gxo3es.jpg
http://i29.tinypic.com/jl1phy.jpg

http://i28.tinypic.com/x1n9g2.jpg
http://i31.tinypic.com/30augk0.jpg
http://i31.tinypic.com/11hdp1s.jpg

the PLL...needed for SetFSB ;)
http://i32.tinypic.com/2dweycy.jpg

PLL is already supported mate ;)

I've just finished testing the board for the max stable FSB with my best Q6600.
At first I tried to hit it from inside BIOS...but anything over 455MHz FSB resulted in no POST.

Tried various FSB/DRAM ratios - Straps ( strap's controlled by the memory divider selection ), nothing.

Then I decided to go with SetFSB in Windows...
And yes! Max stable FSB with my best Q6600 was 485MHz. Not bad ( this CPU's max stable frequency on my best Asus P5K3 Deluxe is 506MHz, and 498MHz on the Abit IP35 Pro )

CPU-z Validation Link for Max Stable FSB with the Q6600 ( Click Me )

The only FSB/DRAM ratio ( involving FSB Strap as well ) that allowed me to reach that FSB stably was the 4.00D

Max stable FSB with this Q6600 and various dividers/straps:

485MHz - 4.00D
460MHz - 4.00B
472MHz - 3.20B
No Go - 4.0A
460MHz - 3.33D
*Mobo Went Crazy* - 2.66D
459MHz - 4.0C
No Go - 3.33C

Edit: I know it's working Dinos thx ;)

Hi BenchZowner,

Thanks for starting this thread. I am interested in buying this board.


The CPU I intend to get is the E8400
The RAM is G Skill PC3-12800 DDR3 1600 (F3-12800CL7D-2GBHZ)

I am keen to hear the thoughts of existing users, especially concerning Over clocking.

Thanks in advance,

David

Hi David,

I'm waiting for a E8400 to arrive today, and I'll start testing it for max FSB, etc.
Hope it does well ( and it looks like it has the potential to do so :D )

More to come later on

Quote:

---

Originally Posted by BenchZowner

Hi David,

I'm waiting for a E8400 to arrive today, and I'll start testing it for max FSB, etc.
Hope it does well ( and it looks like it has the potential to do so :D )

More to come later on

---

The absolute deadline for my purchase is late April 2008. I am holding off as long as possible to get best bang for buck.

Looking forward to your results.

Best wishes,

David

why is the crazy cool backplate present again?

Quote:

---

Originally Posted by LuckyNV

why is the crazy cool backplate present again?

---

Does anyone know if Crazy cool actually works or is it just a Sales Gimick?

Thanks,

David

Quote:

---

Originally Posted by LuckyNV

why is the crazy cool backplate present again?

---

Because it's... crazy marketing cool :D

Kid 2 kid mode "Hey dude! Check this out, this motherboard is full of copper heatsinks, even in the back! It's definitely better than the rest!" :D

Quote:

---

Originally Posted by delewin

Does anyone know if Crazy cool actually works or is it just a Sales Gimick?

---

I will test with Crazy Cool and without to see if there's any difference at all.
However if you're going to use a heatsink/waterblock that make use of a retention plate at the back of the motherboard, you'll have to remove it ( the "Crazy Cool" ) anyways :D

that's my point, its supposed to be a high end enthusiast board, so the space at the back needs to be clear for heatsink/water block backplates :/

Let's take this thing to another level :D

Core 2 Duo E8400 "Wolfdale"
CellShock DDR3-1800 8-7-6-21
Quick FSB Tests ( just some quick quick test )

CPU-z Validation: 520MHz FSB, 400 Strap, DDR3-2081 8-8-8-28 SuperPi 1M Benchable

Damn that seriously blows :( My X38-DS4 pushes 540FSB with the 1600 strap but anything else is crippled. I'll be reviewing the X38-DS4 over the weekend, interested in seeing how these two boards match up. Cheap X38 vs high-end X48T.

Come on Bill, push it a little bit more, I know you can :up: So far, lookin' awesome ;) Do you think the board isn't maxed out yet ? What about the rams, are they clocking good with that board (I know it ain't your P5K3 Deluxe :p:, just out of curiosity wanna see your reaction :D) ...

520MHz is not my max FSB... ;)

I'm waiting for a newer BIOS with better RAM overclocking & dividers 'support'.

I'm starting the default ( stock ) clocks measurements for the review now, gotta remain in schedule.

Whoa!! So far a bit disappointing. It's only 1 board and maybe future bios updates will improve it's performance but wow..... not good.

Quote:

---

Originally Posted by LuckyNV

that's my point, its supposed to be a high end enthusiast board, so the space at the back needs to be clear for heatsink/water block backplates :/

---

Gigabyte has been providing screws as options to remove those crappy backplates for a LONG time

it really isn't an issue as you have everything in the box to allow you to remove it and secure the NB

Quote:

---

Originally Posted by BenchZowner

520MHz is not my max FSB... ;)

I'm waiting for a newer BIOS with better RAM overclocking & dividers 'support'.

I'm starting the default ( stock ) clocks measurements for the review now, gotta remain in schedule.

---

don't hold your breath too much :p

RAM dividers have been problematic on older boards for a long time...i've tested lots of bioses and i can only say that very early bioses for X38 boards are far better FSB OCers as well as everything else for that matter so i prefer to stick with them for great performance :)

just a tip set a loose tRD in bios and boot at 480/500 and setFSB up

also try using high vMCH volts with a monster fan blowing over it to see what the gains are.........this will also allow you to tighen tRD in windows better (to get better tRD you also need to get your timings in bios right.....>> make sure they are tight :))

use Turbo setting for memory performance otherwise performance suffers

Turbo is better than extreme ? [ haven't tested yet :D ]

Mobo won't boot over 455MHz FSB no matter how loose the timings are [ BIOS F4d ]

Yea, the dividers are indeed very very problematic.
Hope for a BIOS fix soon :D

High Vmch is on the way :D [ already done with +0.450V ]

Oh Dinos, clarify some things for me please ( if you can of course ) 'coz my gigabyte contact is away :D

Vmch default voltage is 1.4V [ so +.045V in BIOS means 1.85Vmch ]
Vfsb default voltage is 1.2V [ so +0.2V in BIOS means 1.4Vfsb ]
Vmem default voltage is 1.5V [ so +0.45V in BIOS means 1.95Vmem ]

right ? :p:

looks right to me

I suggested Turbo memory because Extreme can sometimes tighten subtimings too much and prevent RAM/FSB OCs

One more question.... is the D.E.S. software working for you ?
The latest version from Gigabyte's website opens up as a vertical 10pixels gray bar...
( tested in XP & Vista )

i don't have a proper X48 so donno

Today's update:

Max Benchable FSB with a Core 2 Duo E8400 = 536MHz
Max Stable FSB with the same Core 2 Duo E8400 = 528MHz
Max Stable RAM frequency ( 1T CR ) = DDR3-2000 +/- 10MHz

Screenshots & CPU-z Validations:

http://www.benchzone.com/Benchmarkin...dr2000_32m.jpg

CPU-z Validation - DDR3-2000 8-7-6-21 1T
CPU-z Validation - Max Benchable FSB with C2D E8400

One thing about Gigabyte i LOVE, is the build quality, its the best in the biz IMO

fast time Bill

have a look how you stock up against good benchers

http://www.xtremesystems.org/forums/...=32m+challenge

not bad at all mate :up:

I've just finished the measurements for the mobo's review.
I compared it with the Asus P5K3 Deluxe... so it's kinda like P35 vs X48.
Set for the same CPU & RAM clocks on both boards, and then tweaked each board ( sub-timings, strap, performance level, etc ).

CPU: Core 2 Duo E8400 @ 3.6GHz ( 400MHz FSB x 9.0 )
RAM: CellShock DDR3 CL8 Kit @ DDR3-1600 6-5-5-18 ( obviously 1:2 FSB/DRAM Ratio used :D )
VGA: eVGA GeForce 8800GTS 512MB ( Stock )
OS: Microsoft Windows XP Professional SP2
SPU: Creative X-Fi Fatal1ty FPS
[ no LoD, no software tweaks, just clear OS configured for 24/7 usage ]

Gigabyte X48T-DQ6 Rev. 1.1 ( BIOS F4d ) on the left, Asus P5K3 Deluxe ( BIOS v0910 ) on the right:

SuperPi 8M = 156.703s - 155.109s
Hexus PiFast = 25.25s - 25.2s
WinRAR = 2028 - 2130
wPrime = 23.734s - 23.765s
AquaMark 3 = 225929 - 228567
3D Mark2001SE = 66408 - 65862
3D Mark06 = 13376 - 13343
Everest ( Memory Read ) = 10278MB/s - 11129MB/s
Everest ( Memory Write ) = 8528MB/s - 8529MB/s
Everest ( Memory Copy ) = 8928MB/s - 8844MB/s
Everest ( Memory Latency ) = 51ns - 45.9ns

no WAY. The P5K3 is faster? Thats a first attempt at DDR3...the X48 is the third!

:eek:

Thanks for the testing - shame about the FSB.

IMO we really need a mixture of the features of P35 and X38/48.. PCI-E2.0, improved power phases and massive FSB.

Very interesting results. Sounds like I have one of these boards coming my way also so I'm hoping I can get some decent quad FSB clocking going with sub-ambient cooling. Depressing those comparisons with the P5K, I'd have pegged X38 as being a snappier/tighter chipset.

I haven't compiled the review charts yet, but it looks like the performance battle between the Gigabyte & the Asus is 50-50%.
For example if you check my previous post, you'll see that the Gigabyte is faster in 3D Mark2001SE, whereas the Asus is "significantly" faster in AquaMark 3.

After finishing some reviews that I have to clear fast and are in top priority, I'll do some "benching" comparison between the two ( or even more ) in SuperPi 1m/32m , and the 3Ds to see which one's best for each "scenario".

Thanks for the thread Bill :)

I was interested in this board because my P35 doesn't even boot Wolfy after 3 new BIOSes which are supposed to support it but the perf. there 'aint looking like I want it now. :(

Let's hope it's a BIOS issue, they can help much.

Quote:

---

Originally Posted by delewin

Hi BenchZowner,

Thanks for starting this thread. I am interested in buying this board.


The CPU I intend to get is the E8400
The RAM is G Skill PC3-12800 DDR3 1600 (F3-12800CL7D-2GBHZ)

I am keen to hear the thoughts of existing users, especially concerning Over clocking.

Thanks in advance,

David

---

ah ur in luck cuz i use that the stuff ur talking about minus the 8400 :)

rams do good and i usually use the 4.00B strap aka 333mhz.

that'll be about 1800+mhz and 7-7-7-18 1T. 2.15V or so.

i think the thing about this board is bios. i went from f2i to f4h and the difference was stunning! i was getting auto reboots and all at 4.8ghz on dice. once i hit the f4h bios. performance went up and so did stability. no reboots nothing even above 5ghz.

i didnt really do indepth tests but i can say it could be something to do with chipsets too. my friend Harshal from India used a p35 and i the x48t dq6 rev 1.1. we got similiar cpu scores for 3dmark06 at 8332 and 8336. however, my 9650 was running 400mhz slower than his 9770.

perhaps he needs a clean install or something or perhaps it could just be the board? :) just sharing a possibility

Quote:

---

Originally Posted by BenchZowner

I haven't compiled the review charts yet, but it looks like the performance battle between the Gigabyte & the Asus is 50-50%.
For example if you check my previous post, you'll see that the Gigabyte is faster in 3D Mark2001SE, whereas the Asus is "significantly" faster in AquaMark 3.

After finishing some reviews that I have to clear fast and are in top priority, I'll do some "benching" comparison between the two ( or even more ) in SuperPi 1m/32m , and the 3Ds to see which one's best for each "scenario".

---

asus has always been flying but i must say Gigabyte has been doing a good job so far! taking our feedback and all into consideration.. i dont get mosfets burning at 1.95V anymore! :D

Hey NightRaven,
Thanks for the contribution man.

Apart from the dividers problems ( which will more than likely get better in future BIOS releases ) it's a pretty solid board.

One more thing to take review measurements for and then I'm ready to mod the P5K3 Deluxe & take these 2 boards for a good benching test ( gotta know which one's faster and where :D ).

Quote:

---

Originally Posted by BenchZowner

Hey NightRaven,
Thanks for the contribution man.

Apart from the dividers problems ( which will more than likely get better in future BIOS releases ) it's a pretty solid board.

One more thing to take review measurements for and then I'm ready to mod the P5K3 Deluxe & take these 2 boards for a good benching test ( gotta know which one's faster and where :D ).

---

haha u do just that bro. i dont have my p5k3 anymore but imho... the x48 is faster and faster with a better bios.. could be just phsychological but let us know man! :)

im awaiting some more super pi runs if you could do them. I have the X38-DQ6 and I would like to see how they compare to each other. My 8400 Max FSB on here was 531Mhz. Once you find the sweet memory strap and timings these boards are a joy to play with.

Good luck with the rest of your benches! Are you going to do any subambient runs?

I'm planning a LN2 session for this saturday/sunday...hopefully nothing unexpected will happen, so it's going to be 3 boards ( P5K3 Deluxe, Gigabyte X48T-DQ6, Asus Commando ) & 2 CPUs ( E8400 & Q6600 ).

I guess it's been quite a long time ( 'bout a month or so ) without any form of bad luck for me... so it was about time...to happen again.

All out of sudden, the board isn't working :(
Looks like there's a short-circuit happening somewhere on the board or it's just a dead component :(

I'll take a better look at the board tomorrow and see what I can do about it.
Damn!

Sorry to hear that Bill, I'm really getting pissed, each board has it's quirks, formula here doesn't go 400FSB /600 ram, P5K Dlx does, Mars no boot (bummer last two are P35 pfff)... why don't we get something that works flat out out of the box... releasing products is nice but I feel that the need to release them earlier then the competition is really more important than the quality of the product... but that has been going on for years now lol... maybe I need another Hobby :)

btw nice benches lately, I need to catchup again :)

Quote:

---

Originally Posted by BenchZowner

I guess it's been quite a long time ( 'bout a month or so ) without any form of bad luck for me... so it was about time...to happen again.

All out of sudden, the board isn't working :(
Looks like there's a short-circuit happening somewhere on the board or it's just a dead component :(

I'll take a better look at the board tomorrow and see what I can do about it.
Damn!

---

ouch sorry to hear man :( which cpu were you testing when this happened ?

check if any component blew up or burnt up as well

oh bummer!

yea check mosfets?

its weird cuz i haven heard of a single rev 1.1 die yet!

haha even Mr Kill A Gigabyte T_M hasnt killed this one! :D

The board died while...sitting inside Gigabyte's retail package...
It wasn't running...I had it packed to carry it with me today...
I just unpacked it to check a new CPU for max FSB...but it never posted :(

That's the second motherboard to die on me while doing nothing, and more precisely sitting inside its package.
First I had my old & trusty Asus P5W DH Deluxe die like that...and now this one too.

I know I'm all covered up with the words "Bad Luck Guy" but that much ? :(
What's next ? New QX9650 D.O.A. ? ( knock on wood! )

damn thats some weird stuff u got there bro!

tried another cpu? er running cmos batt other way round and stuff?

3 CPUs tested.
Power up with nothing on it ( still wouldn't )
Test with no CMOS battery
Clear CMOS & no Battery for 2 hours ( "asus striker trick" :D :p: )

Something's short-circuiting the board or something's dead ( no physical damage can be spotted )...and I think it short-circuits with the 12V line.

Sorry to hear that, poke and prod the board for a loose component?

Could someone please do me a big favor ?
I need screenshots from the BIOS ( unfortunately every section ) for the review :(
Oh and if you do that ( I owe you big time :) ) could you please have the timings set to the lowest values, and the voltage settings to the highest values.

Thanks in advance ;)

I can supply pictures of bios for you but it will be a x38 dS6 if thats any good for you and the bios is F8

p.m if thats any good.

i have had a number of asus boards die like that.

found it it was something to do with the ram and it corrupting the bios... you'd turn it on and the only thing that would happen is the fans would come on.
12v/3.3v/5v measure points on the board all seems good. but you'd hit the switch and nothing would happen.
you could pull the ram out and the cpu and it would not even beep at you.

something to think about if you can hotflash the bios chip... or get another one from gigabyte

Quote:

---

Originally Posted by -Acid-

I can supply pictures of bios for you but it will be a x38 dS6 if thats any good for you and the bios is F8

p.m if thats any good.

---

Thanks for offering m8, my friend Tim sorted me out some time ago ;)

Quote:

---

Originally Posted by Lestat

i have had a number of asus boards die like that.

found it it was something to do with the ram and it corrupting the bios... you'd turn it on and the only thing that would happen is the fans would come on.
12v/3.3v/5v measure points on the board all seems good. but you'd hit the switch and nothing would happen.
you could pull the ram out and the cpu and it would not even beep at you.

something to think about if you can hotflash the bios chip... or get another one from gigabyte

---

Nah, this one's definitely not a BIOS Corruption.
There's something causing a short-circuit somewhere on the 12V rail.

On the other side, my Asus would power on with the ATX 24pin plug only, but would short-circuit and power off immediately with the 4pin ATX plugged in ( probably a dead mosfet )

Any news on the x48 front? I just ordered the gigabyte x48-dq6, so I'll be following this thread closely.

@BenchZoner are you going to be testing a q6600 in this board? just seen the q6600 post. thx

Quote:

---

Originally Posted by BenchZowner

3 CPUs tested.
Power up with nothing on it ( still wouldn't )
Test with no CMOS battery
Clear CMOS & no Battery for 2 hours ( "asus striker trick" :D :p: )

Something's short-circuiting the board or something's dead ( no physical damage can be spotted )...and I think it short-circuits with the 12V line.

---

hmm i think i just had the same problem as u with one of my boards bro.

but yea i have 2 v1.1s and one died just the way like yours. while my other one is good.

i assume yours has red caps all over?

This is my sample ( Rev 1.1 of course )

http://www.benchzone.com/datas/users...t_page_pic.jpg

Well... I'd say a mix of red & blues :p:

That's weird... two samples die for no reason...
Could be a coincidence... could be bad batch of some components ( e.g. really bad mosfets batch )... who knows...

SOrry for your mobo loss m8 :(

Lol i think us i4memory guys should change our name to iKillMobos.

I blew up a Asus P5B Deluxe/Wifi-AP last night, in my dumbness I swapped a cpu out but left a thumbscrew (didnt see it) sitting under mobo. When I fired it up i smelt lots of burning so quick shut it off. But the southbridge (or something very close to it) totally fried and smelled like burnt peanutbutter. Luckily nothing but the mobo was damaged.

At least I can carve up the heatsinks from it. Is there anything else worth salvaging off a dead mobo?

Capacitors and chokes can be useful for GPU mods. Otherwise that's about it but it's a good start.

Alot of parts can be used to repair other boards, caps, chokes, mosfets, plugs. in fact i'm looking for an old p4 socket 478 board (dead is ok) to get the 4 pin plug for the cpu power to repair a board.

Hi guys, i'm losing my mind on this mobo.
I'm trying to make my system run properly, but in no case I can make my ram modules to run at 1600mhz. Already changed both modules. I cannot reach speeds over 1280 and SPD infos show 1333. System crashes before bios screen if upper speeds are set. Please tell me something.

This is my configuration:
Model Name : GA-X48T-DQ6
--------------------------
M/B Rev : 1.1
BIOS Ver : F4 (F4H already Tested)
Serial No. : SN081100054828

-------------------------------------------------------------------------------
VGA Brand : GALAXY Model : 8800GTS
CPU Brand : Intel Model : Q6600 Speed : 2400MHZ
Operation System : VISTA64BIT SP :
Memory Brand : Corsair Type : DDR3
Memory Size : 2X1GB Speed : 1600
Power Supply : 750 W


ddr3 modules: 2x CM3X1024-1600C7DHX

some pics assembling the system following.

other pics

i know this will be tedious but you need to write out all your memory settings so others can evaluate your problem.

my first thought is to ask you how much vdimm are you using.

Thx for reply Ace, in fact i try many different settings, i try starting with auto settings provided by SPD (1333 9-9-9-24) and after i set manually 1600 7-7-7-20 trying voltage range from 1.80(default ram setting) to 2,35.
But never... all settings putting memory over 1280 block the mobo and reset the overclock bios settings

Quote:

---

Originally Posted by Valdo

Hi guys, i'm losing my mind on this mobo.
I'm trying to make my system run properly, but in no case I can make my ram modules to run at 1600mhz. Already changed both modules. I cannot reach speeds over 1280 and SPD infos show 1333. System crashes before bios screen if upper speeds are set. Please tell me something.

This is my configuration:
Model Name : GA-X48T-DQ6
--------------------------
M/B Rev : 1.1
BIOS Ver : F4 (F4H already Tested)
Serial No. : SN081100054828

-------------------------------------------------------------------------------
VGA Brand : GALAXY Model : 8800GTS
CPU Brand : Intel Model : Q6600 Speed : 2400MHZ
Operation System : VISTA64BIT SP :
Memory Brand : Corsair Type : DDR3
Memory Size : 2X1GB Speed : 1600
Power Supply : 750 W


ddr3 modules: 2x CM3X1024-1600C7DHX

some pics assembling the system following.

---

Use the PINK DIMM Slots &
Try these settings for a start:

CPU Clock Ratio = 7x
CPU Host Control = Enabled
CPU Host Frequency = 400MHz
PCI Express Frequency = 105MHz
Performance Enhance = Standard
System Memory Multiplier = 4.00D
Memory Frequency = 1600
DRAM Timing = Manual
CAS Latency = 8
RAS to CAS Delay = 7
RAS Precharge = 6
tRAS = 21
tRRD = 7
Rank Write to Read = 7
Write to precharge = 7
Refresh to Act = 72
Read to precharge = 7
Static tREAD value = 7
Static tREAD Phase Adjust = 0
CMD = 2N
DDR3 Overvoltage = +0.3V
FSB Overvoltage = +0.2V
MCH Overvoltage = +0.3V
Loadline calibration = enabled

And place a fan above the memory to keep them cool.
Let me know if it worked for you.

Quote:

---

Originally Posted by BenchZowner

Use the PINK DIMM Slots &
Try these settings for a start:

CPU Clock Ratio = 7x
CPU Host Control = Enabled
CPU Host Frequency = 400MHz
PCI Express Frequency = 105MHz
Performance Enhance = Standard
System Memory Multiplier = 4.00D
Memory Frequency = 1600
DRAM Timing = Manual
CAS Latency = 8
RAS to CAS Delay = 7
RAS Precharge = 6
tRAS = 21
tRRD = 7
Rank Write to Read = 7
Write to precharge = 7
Refresh to Act = 72
Read to precharge = 7
Static tREAD value = 7
Static tREAD Phase Adjust = 0
CMD = 2N
DDR3 Overvoltage = +0.3V
FSB Overvoltage = +0.2V
MCH Overvoltage = +0.3V
Loadline calibration = enabled

And place a fan above the memory to keep them cool.
Let me know if it worked for you.

---

Thanks so much Bench with ur settings everest load cpu unless crashes, when i put q6600 at x9 i got crash and now i raise cpu voltage to 1,5, and seems to be ok. I make some tests and after i tell u the result.
Following my actual settings:

Code:

---

GIGABYTE GA X48T DQ6 Clock Settings

MB Intelligent Tweaker (MIT)
Robust Graphics Booster...........: Auto
CPU Clock Ratio...................: 9
CPU Frequency.....................: 3,600
CPU Host Clock Control............: Enable
CPU Host Frequency(Mhz)...........: 400
PCI Express Frequency (Mhz).......: 105
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00D
Memory Frequency(Mhz).............: 1,600
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS#  (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 6
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 7
Rank Write to READ Delay (tWTR)...: 7
Write To Precharge Delay (tWR)....: 7
Refresh to ACT Delay (tRFC).......:  72
Read To Precharge Delay (tRTP)....: 7
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............:  Auto
Command Rate (CMD)................: 2T
Clock Driving and Skew Control
CPU/PCIEX Clock Driving Control...: 900
CPU Clock Skew Control............:  Normal
(G)MCH Clock Skew Control.........: Normal
System Voltage Control () = Base Voltages
DDR3 OverVoltage Control...(1.50v): +0.30
PCI-E OverVoltage Control.,(1.50v): 0.2
FSB OverVoltage Control....(1.10v): 0.30
(G)MCH OverVoltage Control,(1.45v): 0.3
CPU GTLREF1 Voltage*.......(67%)..: Auto
CPU GTLREF2 Voltage*.......(61%)..: Normal
Loadline Calibration..............: Enabled
CPU Voltage Control...............: 1.5000v
Normal CPU Vcore..................:
Note: * Hidden Features Press [CTRL] + [F1] from main Menu to Access
Advanced Bios Features
CPU Multi-Threading...............: Enabled
Limit CPUID Max to 3..............: Disabled
No-Execute Memory Protect.........: Disabled
CPU Enhanced Halt(CIE)............: Disabled
CPU Thermal Monitor 2.............: Disabled
CPU EIST Function.................: Disabled
Virtuaalization Technology Disabled

---

A question, what about pink slots???
Now i'm using the greens

The pink slots overclock better, while the other pair have problems reaching high frequencies and may require more relaxed timings in comparison with the pink ones.
So do yourself a favor and switch over to the pinkies :p:

Drop the PCI Express Overvoltage to normal, there's no benefit at all from raising the PCIe Voltage, and there's more harm to it as you're putting more stress on the boards power circuitry.

Thx again Bench.
Now... i set to default pci/ex voltage but i put cpu on 1,55v to keep windows alive after 20minutes and over of stress.
Following actual most stable settings:

Code:

---

GIGABYTE GA X48T DQ6 Clock Settings

MB Intelligent Tweaker (MIT)
Robust Graphics Booster...........: Auto
CPU Clock Ratio...................: 9
CPU Frequency.....................: 3,600
CPU Host Clock Control............: Enable
CPU Host Frequency(Mhz)...........: 400
PCI Express Frequency (Mhz).......: 105
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00D
Memory Frequency(Mhz).............: 1,600
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS#  (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 6
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 7
Rank Write to READ Delay (tWTR)...: 7
Write To Precharge Delay (tWR)....: 7
Refresh to ACT Delay (tRFC).......:  72
Read To Precharge Delay (tRTP)....: 7
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............:  Auto
Command Rate (CMD)................: 2T
Clock Driving and Skew Control
CPU/PCIEX Clock Driving Control...: 900
CPU Clock Skew Control............:  Normal
(G)MCH Clock Skew Control.........: Normal
System Voltage Control () = Base Voltages
DDR3 OverVoltage Control...(1.50v): +0.30
PCI-E OverVoltage Control.,(1.50v): Normal
FSB OverVoltage Control....(1.10v): 0.30
(G)MCH OverVoltage Control,(1.45v): 0.3
CPU GTLREF1 Voltage*.......(67%)..: Auto
CPU GTLREF2 Voltage*.......(61%)..: Normal
Loadline Calibration..............: Enabled
CPU Voltage Control...............: 1.5500v
Normal CPU Vcore..................:
Note: * Hidden Features Press [CTRL] + [F1] from main Menu to Access
Advanced Bios Features
CPU Multi-Threading...............: Enabled
Limit CPUID Max to 3..............: Disabled
No-Execute Memory Protect.........: Disabled
CPU Enhanced Halt(CIE)............: Disabled
CPU Thermal Monitor 2.............: Disabled
CPU EIST Function.................: Disabled
Virtuaalization Technology Disabled

---

First image showing stress test.

Now this is my new bizarre problem.
As show in the first image, speedfan receive a cpu temperature of 75c° (sometimes 80), then i put thermal sensors inside cpu bracket and the thermal monitor show me 49c°.
I heard that q6600 sensor simetimes work wrong... is this my case?
Following the sensors images.
Thx again for the great help.

What's the main difference between rev. 1.0 and 1.1 because i can buy the 1.0 here in the netherlands (mobo is only thing i need to finish my rig :p: )

And i have 3 disks, 2x Raptors 74GB and 1x 36GB Raptor, if i would like to run Raid0 on the 74GB disks i want to use the ICH9 for it and should i use the ICH9 also for the single 36GB disk or is it better to use je JMicron next to the Raid0 on the ICH9?

Tips are welcome...

Quote:

---

CPU Voltage Control...............: 1.5500v

---

:lol: I said you might want to up it a notch from the 1.4125v i quoted , wasn't particularly expecting that sort of hike :)
No wonder it's getting a bit toasty, even on water.:0

Quote:

---

Originally Posted by X4er0

What's the main difference between rev. 1.0 and 1.1 because i can buy the 1.0 here in the netherlands (mobo is only thing i need to finish my rig :p: )

---

wait for rev1.1
that will be true X48 platform

Quote:

---

Originally Posted by dinos22

wait for rev1.1
that will be true X48 platform

---

Thnx m8, and do you also have an answer to my ich9 and jmicron scenario?

Maybe i'll go for 2x74GB Raptor in Raid0 on the ICH9 and single 36GB Raptor on the JMicron paired with my SATA DVD-RW

Quote:

---

Originally Posted by Supershanks

:lol: I said you might want to up it a notch from the 1.4125v i quoted , wasn't particularly expecting that sort of hike :)
No wonder it's getting a bit toasty, even on water.:0

---

U think 1,55 is too mutch? any idea what i can change coming back to 1,4125 and no getting crash?

Any idea on my thermal sensors problem?

Quote:

---

Originally Posted by X4er0

What's the main difference between rev. 1.0 and 1.1 because i can buy the 1.0 here in the netherlands (mobo is only thing i need to finish my rig :p: )

---

I'm not going into many details...but the major one is... fixed burnt mosfets issue when using high voltages :D

For the rest...I'll have to check with my contact in GB.

Quote:

---

Originally Posted by X4er0

And i have 3 disks, 2x Raptors 74GB and 1x 36GB Raptor, if i would like to run Raid0 on the 74GB disks i want to use the ICH9 for it and should i use the ICH9 also for the single 36GB disk or is it better to use je JMicron next to the Raid0 on the ICH9?

---

You can use any configuration on the ICH9R.
And of course the one you've mentioned.
ICH9R > JMicron ;)

Quote:

---

Originally Posted by Valdo

U think 1,55 is too much? any idea what i can change coming back to 1,4125 and no getting crash?

Any idea on my thermal sensors problem?

---

Well, the inside of the CPU will always be hotter than the heatspreadder's edge... and the DTS signals are more trustworthy in this case.
You can try using lower Vcore to see if your CPU can work stably with less voltage.
Try 1.425V, and if it's unstable move on to 1.45V, and slowly to 1.475V ( if 1.45V is unstable again ).

Quote:

---

U think 1,55 is too mutch? any idea what i can change coming back to 1,4125 and no getting crash?

---

for 9x400 yes:) I suggested using 1.425v as a start point. If that's not stable go to next voltage up, prime or test again, until it doesn't BSOD or primes ok. If you do it that way i reckon you'll be stable before you get to 1.55v. :)

Your not comparing like with like with the thermal sensor. For the voltage your using 75C is not unreasonable . I'd use Coretemp to track your cpu core temps. The value of your thermal sensors is going to be widely different as it's quite remote from the temps in your cores. I have an example that might highlight the issue. Using Everesyt I get the Temp of my NB, this is around 51C , now i have a sensor like yours fastened to the top of the NB this is reading 35.5C The 1st values from the internal CPU sensors measure the heat being generated. The 2nd value measure the heat being dissapated. In the case of your CPU Sensor it is fairly remote from the action.

Just get your vcore sorted & your temps will sort themselves out.

Quote:

---

Well, the inside of the CPU will always be hotter than the heatspreadder's edge... and the DTS signals are more trustworthy in this case.
You can try using lower Vcore to see if your CPU can work stably with less voltage.
Try 1.425V, and if it's unstable move on to 1.45V, and slowly to 1.475V ( if 1.45V is unstable again ).

---

ok i try next step on monday, if i remember good i got a crash with 1,5 but i'm NOT<---(edit) sure atm (many different tries today ;)).
Thx again for all the help

Quote:

---

if i remember good i got a crash with 1,5

---

which might mean that your memory is unstable as we talked about in the other rthread.

What is the main difference between Ver 1.0 and 1.1, is it different kind of capacitors or ......??

Quote:

---

Originally Posted by X4er0

What is the main difference between Ver 1.0 and 1.1, is it different kind of capacitors or ......??

---

Already told ya.
Posted above

Well I get My GIGABYTE GA X48T today. Can't figur out why i can't run my Corsair TWIN3X2048-1800C7DFIN at 1800mhz. could some one tell me why???

? 1 more Q why just i run Prime95 at Default cpu speed with out the test failing ???

Darknezz please use this form

Quote:

---

GIGABYTE GA X48T DQ6 Clock Settings

MB Intelligent Tweaker (MIT)
Robust Graphics Booster...........: Auto
CPU Clock Ratio...................: 9
CPU Frequency.....................: 3,600
CPU Host Clock Control............: Enable
CPU Host Frequency(Mhz)...........: 400
PCI Express Frequency (Mhz).......: 105
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00D
Memory Frequency(Mhz).............: 1,600
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS# (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 6
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 7
Rank Write to READ Delay (tWTR)...: 7
Write To Precharge Delay (tWR)....: 7
Refresh to ACT Delay (tRFC).......: 72
Read To Precharge Delay (tRTP)....: 7
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 2T
Clock Driving and Skew Control
CPU/PCIEX Clock Driving Control...: 900
CPU Clock Skew Control............: Normal
(G)MCH Clock Skew Control.........: Normal
System Voltage Control () = Base Voltages
DDR3 OverVoltage Control...(1.50v): +0.30
PCI-E OverVoltage Control.,(1.50v): Normal
FSB OverVoltage Control....(1.10v): 0.30
(G)MCH OverVoltage Control,(1.45v): 0.3
CPU GTLREF1 Voltage*.......(67&#37;)..: Auto
CPU GTLREF2 Voltage*.......(61%)..: Normal
Loadline Calibration..............: Enabled
CPU Voltage Control...............: 1.5500v
Normal CPU Vcore..................:
Note: * Hidden Features Press [CTRL] + [F1] from main Menu to Access
Advanced Bios Features
CPU Multi-Threading...............: Enabled
Limit CPUID Max to 3..............: Disabled
No-Execute Memory Protect.........: Disabled
CPU Enhanced Halt(CIE)............: Disabled
CPU Thermal Monitor 2.............: Disabled
CPU EIST Function.................: Disabled
Virtuaalization Technology Disabled

---

to show us your current settings m8

Thanks do i need any extra kooling on the chipset. Damn it gets hot.. Well I got mess around with it and got it it too run 4050mhz and ram at 1600 but i seemed to crash random. Ran Prime on it for 8 hours but random crashes in gaming. I will get back too you on how it went..

Well I got it running at 4050mhz

MB Intelligent Tweaker (MIT)
Robust Graphics Booster...........: Auto
CPU Clock Ratio...................: 9
CPU Frequency.....................: 4,050
CPU Host Clock Control............: Enable
CPU Host Frequency(Mhz)...........: 450
PCI Express Frequency (Mhz).......: 105
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00D
Memory Frequency(Mhz).............: 1,800
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS# (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 6
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 7
Rank Write to READ Delay (tWTR)...: 7
Write To Precharge Delay (tWR)....: 7
Refresh to ACT Delay (tRFC).......: 72
Read To Precharge Delay (tRTP)....: 7
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 2T
Clock Driving and Skew Control
CPU/PCIEX Clock Driving Control...: 900
CPU Clock Skew Control............: Normal
(G)MCH Clock Skew Control.........: Normal
System Voltage Control () = Base Voltages
DDR3 OverVoltage Control...(1.50v): +0.50
PCI-E OverVoltage Control.,(1.50v): Normal
FSB OverVoltage Control....(1.10v): 0.30
(G)MCH OverVoltage Control,(1.45v): 0.3
CPU GTLREF1 Voltage*.......(67%)..: Auto
CPU GTLREF2 Voltage*.......(61%)..: Normal
Loadline Calibration..............: Enabled
CPU Voltage Control...............: 1.4240v
Normal CPU Vcore..................:
Note: * Hidden Features Press [CTRL] + [F1] from main Menu to Access
Advanced Bios Features
CPU Multi-Threading...............: Enabled
Limit CPUID Max to 3..............: Disabled
No-Execute Memory Protect.........: Disabled
CPU Enhanced Halt(CIE)............: Disabled
CPU Thermal Monitor 2.............: Disabled
CPU EIST Function.................: Disabled
Virtuaalization Technology Disabled

I well start testing some time later today..
Thanks a hole lot for your help..

Darknezz

your VTT(fsb Overvoltage) is a bit high means your running at 1.40v which is high for a 65nm cpu . you need to be at about 1.25v with high fsb +0.10v should do fine there.

Try
CPU GTLREF1 Voltage*.......(67%)..: Auto ---> -3
CPU GTLREF2 Voltage*.......(61%)..: Normal---> -3

You might be able to tighten tras secondary timings a bit. Try tweaking in memset.

on 1800 with same main timings as you
I'm at :-

Precharge Delay (tRAS)............: 18
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 5
Rank Write to READ Delay (tWTR)...: 5
Write To Precharge Delay (tWR)....: 7 (I'll try 7 I've been on 9~10)
Refresh to ACT Delay (tRFC).......: 65
Read To Precharge Delay (tRTP)....: 5
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 1T

I've also run with a bit more
(G)MCH OverVoltage Control,(1.45v): 0.375

My NB is on water now , but prevuiusly was running with a 60mm fan

luck:)

Supershanks Iam run a 45nm cpu but i know what you mean.. I have an old swiftech cpu cooler i am make to a nb cooler so that mit help, on the nb temp. Thanks for your help.. How did you get the fan to say there tape?? the swiftech has a 14mm thick cooper buttom. that should work.

Quote:

---

Supershanks Iam run a 45nm cpu

---

Yes:D that's why I muntioned it :lol:
I used double sided tape to secure the fan , it's a 60mm.

Quote:

---

I have an old swiftech cpu cooler

---

:lol: Mine is also a recycled CPU Cooler :D -
Alphacool Cape KC42-X2 Pro

I used these for the PWM area Enzotech MOS-C1 as the thermalright mosfet coolers are too wide.

luck:)

The Ones you have on the Mossfet can't buy them in Denmark :( But i am going to use a Thermalright HR-09U (Type 2) for the mossfet and a EnzoTech SLF-1 - 35mm for the sb..

As far as i know none of the Thermalright mossfet coolers fit it's the width thats too big. if you look at the Technical Specs you'll see that all models have a common width of 15.5mm. , about 3mm too wide.
Thermalright originallty had the HR-09 S/U type 2 Mosfet Cooler as compatible, we got that removed after around 3 of us had shelled out for them :D

You might be able to file down the width , but it might be a long job :shrug:

Well then i well have fine something for it. Do you have the spec width 12.5mm. but how long is it we have cooper were i work and a hole lot of milling machines..

Supershanks I found a fan at work 40mm 24volt but but it's going to run at 12volt It's for the nb.. :) Hope it works..

sorry darkness not got the width will try & get it for you.
Yep 40mm should be good, at least give you an idea of impact. a fan has. :)

Edit block I think needs to be 84mm, Pin to pin is about 94mm

Ok thanks I made some Mossfet cooling at work 12.5mm and 101mm long. What the most anyone has got the ram to run on this mobo i am running them at 1800mhz right now

Well i got the Ram to run at 1900mhz:shocked: write now At 2.00volt :clap:

MB Intelligent Tweaker (MIT)
Robust Graphics Booster...........: Auto
CPU Clock Ratio...................: 8.5
CPU Frequency.....................: 4,030
CPU Host Clock Control............: Enable
CPU Host Frequency(Mhz)...........: 475
PCI Express Frequency (Mhz).......: 105
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00B
Memory Frequency(Mhz).............: 1,900
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS# (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 6
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 5
Rank Write to READ Delay (tWTR)...: 5
Write To Precharge Delay (tWR)....: 7
Refresh to ACT Delay (tRFC).......: 65
Read To Precharge Delay (tRTP)....: 5
Static tREAD Value (tRD)..........: 7
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 1T

Nice going m8 :up:

I am going too see if I am get the ram too 2000mhz But how much volt du you think the ram would need.

Well this Mobo is just great I have play around with it for some days now. I hit 2000mhz on my Memory today :yepp: Thanks for all the help everyone has given me :clap:

I'm having some trouble stabilizing the system as well. I have a QX9770 with a Scythe Mugen cooler with 1x 120mm fan in a Lian-Li 1000 aluminum case (exhaust an is about 2" from the heatsink working in tandem with the airflow. I also slapped a 40x40x20mm fan at 0.5" height from the northbridge part of the mobo heatpipes.

I have 4x2gb OCX 7-7-7-24-2T rated DDR3-1600 installed.
Currently it is at DDR2-1333 8-7-7-20-2T.
It runs stable at DDR2-1640 8-7-7-20-2T under memtest.

I have the following settings:

FSB 400
Multiplier 9.5x = 3.8GHz
Vcore = 1.3875V
Vfsb = +.25
Vnb(mch) = +.325
Vmem = +.4 -> 1.92V is within spec (up to 1.95V)
mem div at 3.33D (1333)
CPU drive = 900mV
no skew on cpu or mch
other (seconary) memory settings auto except for 2T see above for primary

Using Prime95 with 4x instances and affinity I can run small FFT for 2h 30m (so far) no errors CUP temp 80C = 175F (yes it is high but this machine won't be stressed like this, plus it was stable...)

Using memtest I get no errors on an overnight run at DDR3-1640 (see above) so I set memory to DDR3-1333 to eliminate memory as the problem.

When I run Blend 1800Mb for each Prime95 it dies on core #2 and #4 within about 10 minutes. Temp only gets to about 74C. I think accordng to HWMonitor the NB is getting to 63C.

Normally I woul have water cooled this machine but my office didn't want "plumbing".

So... it looks like the CPU/Memory interface (NB) is the culprit, how to I stabilize it?

Thanks.

Quote:

---

Originally Posted by MagnumMan

I'm having some trouble stabilizing the system as well. I have a QX9770 with a Scythe Mugen cooler with 1x 120mm fan in a Lian-Li 1000 aluminum case (exhaust an is about 2" from the heatsink working in tandem with the airflow. I also slapped a 40x40x20mm fan at 0.5" height from the northbridge part of the mobo heatpipes.

I have 4x2gb OCX 7-7-7-24-2T rated DDR3-1600 installed.
Currently it is at DDR2-1333 8-7-7-20-2T.
It runs stable at DDR2-1640 8-7-7-20-2T under memtest.

I have the following settings:

FSB 400
Multiplier 9.5x = 3.8GHz
Vcore = 1.3875V
Vfsb = +.25
Vnb(mch) = +.325
Vmem = +.4 -> 1.92V is within spec (up to 1.95V)
mem div at 3.33D (1333)
CPU drive = 900mV
no skew on cpu or mch
other (seconary) memory settings auto except for 2T see above for primary

Using Prime95 with 4x instances and affinity I can run small FFT for 2h 30m (so far) no errors CUP temp 80C = 175F (yes it is high but this machine won't be stressed like this, plus it was stable...)

Using memtest I get no errors on an overnight run at DDR3-1640 (see above) so I set memory to DDR3-1333 to eliminate memory as the problem.

When I run Blend 1800Mb for each Prime95 it dies on core #2 and #4 within about 10 minutes. Temp only gets to about 74C. I think accordng to HWMonitor the NB is getting to 63C.

Normally I woul have water cooled this machine but my office didn't want "plumbing".

So... it looks like the CPU/Memory interface (NB) is the culprit, how to I stabilize it?

Thanks.

---

I have same problem, i'm running E8500 so 9.5x400 wih same mem 2x2GB OCZ 1600EB on 7-7-7-24 2T.

It's the 4GB, most people can get it higher but use 2GB we have 2x2GB tried everything and with the help of benchowner i have these settings:

1.35 loadbalance on
1.85V mem
+0.25FSB
+0.35MCH

Go over to the larger GA-X48T-DQ6 thread on this forum, I posted all my settings which are now stable.

I'm confused

Quote:

---

So... it looks like the CPU/Memory interface (NB) is the culprit, how to I stabilize it?

---

Quote:

---

Go over to the larger GA-X48T-DQ6 thread on this forum, I posted all my settings which are now stable.

---

Have you sorted it or not ??
if yes where please :shrug:

I've been runnung my mch at a minimum of around +0.375 & have been up to +0.475 for 550 FSB

I have had on occasion the same symptom running prime which usually takes a dive as soon as small ffts start using blend. just upping vcore isn't the answer. I'm on water for my nb, any mch volts >0.475 don't add anything.

I'm running FSB at stock speed 1600, I experienced hard lockups if the FSB voltage was +.30V or higher; I experienced hard lockups if the NB was +.40V or higher.

The key to getting the memory stable was to set all the "secondary" memory timings (below the 7-7-7-20 section). Manually set each one to be +1 over the "Auto" setting on the left. If it says 5, put it to 6. I set the one that is normally 60 to 75. Set 2T command rate. This plus +.20 FSB and +.30 MCH I believe stabilized it.

You should definitely first run memtest for about 6 hours after changing the secondary settings; once that passes then make sure you can run 6 hours of Small FFT in Prime95 with affinity on every core, and then try blend...

Quote:

---

I experienced hard lockups if the FSB voltage was +.30V or higher

---

yep that makes sense on 65nm vtt (fsb) is around 1.35 to get 400~450 which is +0.25v in gigabyte over voltage.

on 45nm the setting should be lower.

Quote:

---

I experienced hard lockups if the NB was +.40V or higher.

---

As my NB is on water , that probably tallies with my 0.475v max , as i said anything above that was wasted. that means a real nb voltage of 1.25 + 0.50 = 1.75v is tops.

agree re 2ndry timings

luck:)

what bios u use??
can u post the template for 1000 8-7-7-21 rams??

what rams u use ??

All the info is on the "official" thread in the forums here.

http://www.xtremesystems.org/forums/...&postcount=323

dark, your settings ?