DFI UT X58-T3EH8 continued - Part 2

[cstkl1]

hmm something interesting

i can run qpi multi 12.

go figure

[ci2kla]

hmm something interesting

i can run qpi multi 12.

go figure

What do you mean?....

[cstkl1]

What do you mean?....

means i am running my proc at 4.3ghz bclk 205, dram at 2050 C7
uncore at 4.1ghz
but instead of qpi at 205 x 18 i am running 205 x12.

all i did was change the qpi link speed to disable

so far performance wise no difference from gc benchmark etc

will post later..
just sold off my gc's just a few minutes ago.

sourcing some new gc's hopefully atomics'

[Biker]

you'll get that if cpu vtt and vdimm aren't set optimally....

DFI has been able to duplicate this issue and are currently working on a fix....

Thanks for the tips and updates, guys. I haven't had more time to test as yet.

I have however come across a much more fundamental problem:

The board will not recognise my RAID card!

The card in question is the one in my sig (LSI megaraid 8708ELP) and it is not recognised at boot and does not show up in Windows either...

I have tried it in all slots (except the 4x slot since it is physically an 8x card and will not fit)... I have played with memory lowgap (up to 2560) with no change...

Has anyone got a RAID card to work in PCI-E slots 1, 3 or 4?

[cstkl1]

Thanks for the tips and updates, guys. I haven't had more time to test as yet.

I have however come across a much more fundamental problem:

The board will not recognise my RAID card!

The card in question is the one in my sig (LSI megaraid 8708ELP) and it is not recognised at boot and does not show up in Windows either...

I have tried it in all slots (except the 4x slot since it is physically an 8x card and will not fit)... I have played with memory lowgap (up to 2560) with no change...

Has anyone got a RAID card to work in PCI-E slots 1, 3 or 4?

tested one for office a areca pcie 8x two ssd on raid no issue. on the middle pciex 16 slot. aka pciex 3

[Biker]

tested one for office a areca pcie 8x two ssd on raid no issue. on the middle pciex 16 slot. aka pciex 3

Roger that.

Thanks for testing...

No joy in middle slot for me so:

Defective board? or

Board does not recognise 4x card? or

Something overlooked in bios? or

Something else going on....

Answers on a postcard .....

Edit: Just tested with an Adaptec host raid card that i had lying around and that works just fine so it is looking like a LSI / DFI incompatibility

Anyone got one of these working with a LSI Megaraid card?

[cstkl1]

Roger that.

Thanks for testing...

No joy in middle slot for me so:

Defective board? or

Board does not recognise 4x card? or

Something overlooked in bios? or

Something else going on....

Answers on a postcard .....

Edit: Just tested with an Adaptec host raid card that i had lying around and that works just fine so it is looking like a LSI / DFI incompatibility

Anyone got one of these working with a LSI Megaraid card?

well it also worked fine with my asus xonar d2x which is 1x.

[ChaosMinionX]

New board gets here tomorrow... lets hope this ones a little more cooperative.

[SoulsCollective]

Roger that.

Thanks for testing...

No joy in middle slot for me so:

Defective board? or

Board does not recognise 4x card? or

Something overlooked in bios? or

Something else going on....

Answers on a postcard .....

Edit: Just tested with an Adaptec host raid card that i had lying around and that works just fine so it is looking like a LSI / DFI incompatibility

Anyone got one of these working with a LSI Megaraid card?

Can confirm that a Highpoint RocketRAID and Areca card both work in all PCI-E slots.

Which BIOS version are you using? IIRC the changelog for one of the recent BIOS files on the DFI site mentioned PCI-E RAID card incompatibility fixes. However, it seems you're not alone

[Biker]

However, it seems you're not alone

Wow!

A hardware issue clone

Seriously though, it looks like that recent bios fix worked for Adaptec cards but LSI are no go. The thing is I have 2 important arrays on my LSI card and since I moved to X58 after my P35-T2R's NB started dying i need to to work

Apparently the 8888ELP card is causing issues to (saw it regarding P45 boards) so I was wondering if it is a PCI-E 2.0 issue (my LSI works flawlessly on PCI-E 1.0)...

What bios? Currently using the 16/06 release and have tested the Adaptec fix release....

[StimulatedBoy]

Been trying out BIOS 06-16 the last day and a half and cant get my system to work at same clock as I did with 05-08 without BSODing all over the place. Reverting to previous release...

[cstkl1]

Been trying out BIOS 06-16 the last day and a half and cant get my system to work at same clock as I did with 05-08 without BSODing all over the place. Reverting to previous release...

no diff for me.

but then again i did stress test the hellout of every clocks i did.

[Factotum]

hi guys!
my new Ci7 920 D0 Batch#3845B260 end new WC setup:
WC:
[Laig 3.2+XSPC ResTop] -> [XSPC RX360+3xYate Loon 1440RPM] -> Sapphire (Core i7 920) -> MCW30 (x58) -> SW16 (Digital PWM) -> MCW60r2 (GTX285)



ambient - 24
water IDLE - 26.3
water Load - 32

1HR Prime Blend:

[unclewebb]

Factotum: You might be having a problem with Turbo throttling at full load. That's when your +1 Turbo will rapidly cycle on and off so your CPU multiplier will bounce back and forth from 21 to 20. That is likely why RealTemp is reporting 20.8 for your multiplier.

Try running i7 Turbo and keep an eye on what it shows for the Calculated Multiplier. Run Prime 95 Small FFTs and see if it is steady at 21.000 or not.
http://www.fileden.com/files/2008/3/3/1794507/Turbo.zip

If your Turbo is cutting out at full load then there should be an option in your bios to turn this off.

[Factotum]

Factotum: You might be having a problem with Turbo throttling at full load. That's when your +1 Turbo will rapidly cycle on and off so your CPU multiplier will bounce back and forth from 21 to 20. That is likely why RealTemp is reporting 20.8 for your multiplier.

Try running i7 Turbo and keep an eye on what it shows for the Calculated Multiplier. Run Prime 95 Small FFTs and see if it is steady at 21.000 or not.
http://www.fileden.com/files/2008/3/3/1794507/Turbo.zip

If your Turbo is cutting out at full load then there should be an option in your bios to turn this off.

done



bios settings:

CPU Feature
Set VR Current Limit Max: Enabled
Thermal Management Control: Disabled
EIST Function: Enabled
CxE Function: Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled

***** Logical Processor Setting *****
Intel HT Technology: Enabled
Active Processor Cores: All 
PPM Function: Enabled
Turbo Mode Function: Enabled
1 core Max Turbo Ratio: 22x
2 core Max Turbo Ratio: 21x
3 core Max Turbo Ratio: 21x
4 core Max Turbo Ratio: 21x
CPU Non-Turbo Clock Ratio: 20x

what wrong?

[unclewebb]

While Prime95 Small FFTs is running, the Calculated Multiplier should stay steady at 21.000. If i7 Turbo 6.1 shows that it is constantly less than 21.000 then that is a sign of Turbo throttling. Your bios looks like it is setup correctly. Try turning on the Log File feature of i7 Turbo so you can go back and see if the multiplier drops down for 1 second which might be reporting error or if it is constantly changing and dropping below 21.000. Post a piece of your log if you see a lot of numbers in a row below 21.000.

[Factotum]

@unclewebb
5 min Prime95 Small FFTs

and logfile: http://cp.people.overclockers.ru/cgi...i7TurboLog.zip
anomaly?

[unclewebb]

Factotum: How about reducing your overclock down to about 190x21 and run Prime95 Small FFTs again and see if this problem goes away?
Based on your bios settings, your CPU shouldn't be throttling but it looks like it is trying to.

Edit: Before doing that maybe try turning off Everest. This might be interfering with i7 Turbo.

[Factotum]

Edit: Before doing that maybe try turning off Everest. This might be interfering with i7 Turbo.

confirmed!!! Turbo/RealTemp have conflicts with Everest
5min Prime Small FFTs w/o Everest:


LogFile: http://cp.people.overclockers.ru/cgi...TurboLog_1.zip

many thanks @unclewebb

[Dami3n]

Testing my new toy,


Low volts, but high temp. I think its time to watercooling

Edit-Promising,

[Dami3n]

I received today my dominator gt, and reading the elpida graveyard post, i have fear to kill them too fast. What vtt is usually needed with elpida chips?

[ci2kla]

Testing my new toy,


Low volts, but high temp. I think its time to watercooling

Edit-Promising,

Wow that temp is rather high IMO...by the way what batch of yours (3520)?

Guys I have a problem...Tried 0508 and 0616 BIOS and problem still exists.
[NO OVERCLOCKED] When I tried using PCIE Clock 100Mhz my GTX275 won't work and device manager claimed it can't active the nVIDIA driver however it worked @ 101Mhz+...I wonder if it's an BIOS issue?

[Dami3n]

3845A935.

I have no problems with pci and my 295. Im using 0508 official bios.

[eva2000]

I received today my dominator gt, and reading the elpida graveyard post, i have fear to kill them too fast. What vtt is usually needed with elpida chips?

Depends on mem and bclk frequencies and cpu multiplier used - http://www.xtremesystems.org/forums/...d.php?t=227580

could be anywhere between 1.31-1.65v cpu vtt needed as it can be individual cpu dependent as well as motherboard dependent

[Blacky]

I need some help for my friend, he got his new setup today (DFI X58 + G.SKill Trident + i7 920 D0) and the mobo is only showing 2 cores instead of 4 (@BIOS on CPU-Z etc), we tried everything and still shows 2, any idea? please any help would be appreciated...