Intel Core 2 Quad Q6600 - Kentsfield + D975XBX Bad Axe rev304

[eva2000]

Intel Core 2 Quad Q6600 - Kentsfield

This afternoon a courier dropped off a new toy for me to play with, an Intel Core 2 Quad Q6600 Kentsfield B3/Step 7 quad core cpu

QUPT, Week 32, 2006


Dropped it in a Intel D975XBX Bad Axe rev304 motherboard with latest 1441 bios and it is recognised as Intel Core 2 Quad Q6600 cpu. Ignore the processor speed of 3.72Ghz, the bios since day 1 has misreported the speed IF the fsb strap is changed from default to say 1067fsb or 1333fsb strap. Right now it's set at 1067fsb strap.


Temps look okay for idle pretty close to my E6600 and E6700 B1 cpus on same cooling. I used the new application method for AS5 listed here.


New feature in bios that i have never seen on D975XBX Bad Axe board when i used my E6600 and E6700 B1 cpus, a selectable cpu multiplier!

The default vcore is 1.275v for Coretemp VID but i had 1.3125v vcore set in bios on my previous session with E6700 cpu. With 1.275v vcore set in bios, vcore reads 1.242v with 43C idle cpu temps. The default settings with automatic multiplier:


The manual default multiplier settings (whether multipliers higher than default 9x work haven't tested yet):


The manual multiplier max and min multipliers are 40x and 5x respectively in bios:


As you can see the Intel D975XBX Bad Axe rev304 has no problem recognising the Intel Core 2 Quad Q6600 Kentsfield cpu in bios at least. I have enabled Enhanced Powerslope in bios which reduces vcore droop by as much as 50%

[eva2000]

System:

• Intel Core 2 Quad Q6600 Kentsfield QUPT B3 Step 7
• Corsair Hydrocool 200EX
• Intel D975XBX Badaxe rev304 1441 bios
• 128MB Gainward FX5200 PCI card
• 2x1GB Gskill PC6400PHU2 2GBHZ Micron D9GMH 0629 batch
• 3x320GB Seagate 7200.10 SATAII Matrix Raid 0
• 1x74GB Western Digital Raptor 8MB
• Pioneer DVD-RW
• FDD
• 1KW PCP&C SLI PSU

Cpuz v1.37 recognises the Intel Core 2 Quad Q6600 correctly but Crystalmark thinks the cpu is a Woodcrest, while CoreTemp v0.93 thinks the cpu is a Intel Xeon Q6600 Kentsfield.

Not sure if it's the new 1441 bios or Corsair Hydrocool 200EX water cooling setup, but bios idle temps of 43-45C are reflected in Coretemp windows idle temp of 36-40C which cpu1 core being 3C cooler than the other 3 cores. That's still 6-10C hotter than E6600 B1 at default 2400Mhz and default vcore of 1.242v vcore.

Room ambient temps = 27.6C - 28.5C





Quad Stressprime at idle shows coretemps of 36-40C:
(click image for larger version)


Quad Stressprime load shows cpu0/2 at 57C and cpu1/3 at 55C:
(click image for larger version)


Quad Systool 8M and 16M PI:


Everest Ultimate Edition v3.50
D975XBX Bad Axe has one outstanding issue that effects any cpu, that is when Sandra 2007 or Everest are launched, it resets the vcore to that of below stock value.

On my E6600 B1s default vcore = 1.25 (1.242v), when Everest/Sandra 2007 are launched, vcore is reset to 1.218v. Seems same occurs with Intel Core 2 Quad Q6600 cpu, default vcore = 1.275 (1.242v). When Everest is lauched, vcore resets to 1.15v

This is way below the lowest vcore setting available in bios of 1.275v (1.242v). As you can see windows idle temps dropped too.



Weird never noticed there's a DIMM temp sensor there before ? 1441 bios added ?

[eva2000]

Overclocking:
Looks like the D975XBX Bad Axe rev304 board might not be fully working with Intel Core 2 Quad Q6600 cpu. I ran Systool quad 16M Pi to test clocks with clockgen and from bios and managed

vFSB = default
vMCH = 1.628v
vDimm = 2.05v

- 2900Mhz at 1.242v (1.275v bios)
- 3003mhz at 1.277v (1.315v bios)
- 3060Mhz at 1.312v (1.35v bios)
- 3109Mhz is where trouble starts:
* 1.359v (1.4v in bios) = rebooted immediately in Systool quad 16M Pi tests.
* 1.359v vcore and raised vFSB to 1.395v = rebooted at 13% mark for Systool quad 16M Pi tests
* 1.406v (1.45v bios) with vFSB 1.395v = PASSED Systool quad 16M Pi tests.
* 1.453v (1.50v bios) with vFSB 1.395v = rebooted after 8-13% in Systool quad 16M Pi tests
* 1.476v (1.525v bios) with vFSB 1.395v = hangs at boot time just before windows logo

Raising vMCH to 1.685v from 1.628v allowed me to pass at
* 1.359v (1.4v bios vcore) + vFSB 1.395v + vMCH 1.685v = PASSED Systool quad 16M Pi tests (can even loop the test without problems).

Clockgen'ing from 3109Mhz to 3150-3200Mhz resulted in Systool quad 16M Pi test rebooting immediately regardless of vcore up to 1.453v. Similar, if i set in bios greater than +30% FSB adjustable i.e. +33% will result in hang just before windows logo loads.

For those that recall, originally D975XBX Bad Axe board came with max adjustable FSB in percentage terms at +30%. Then later bios allowed +50% max FSB. It seems with Q6600 cpu, i'm only getting stable FSB clock at +30% max so far with 1441 bios. Might need to try earlier bioses as well http://i4memory.com/showthread.php?t=2804 and see if it helps.

Hope it's the board/BIOS and not the cpu as this Q6600 was pretested by the seller to 3600Mhz 9x400fsb quad prime stable at 1.45v vcore on air on Asus P5B Deluxe and Asus P5W64 WS Pro motherboards

Update: Looks like Kentsfield might have a low FSB limit on D975XBX Bad Axe motherboard, Shamino only managed max 356FSB with QX6700 on D975XBX Bad Axe http://sg.vr-zone.com/?i=4098&s=7.

[eva2000]

Super Pi v1.50
Doing some stock cpu speed tests with no pi tweaks just to get an idea of how Intel Core 2 Quad Q6600 cpu runs at 9x266fsb = 2400Mhz at 1.242v with memory @266mhz 4-4-4-8.

Single 1M & 32M


Single 2M & 4M


Single 8M & 16M


Dual 1M


Quad 1M


Dual 8M


Quad 8M


Dual 32M


Quad 32M

[G H Z]

Nice chip, good luck and have fun with it eva.

[s e t h]

another low fsb wall on the new stepping kentsfields...this is getting really worrying
great testing as always...thanks...

[VictorWang]

ha, u need a P5B-DL Mod with bios=0804 for this chip

[happyhero]

I updated my bios from 604 to 804 with the latest update exe in windows and i got a BSOD did a reboot and set defaults and then changed to my faverate settings. It seems that the update was done good as i am in windows without any probs. is there any thing i should know?

[eva2000]

thanks guys

ha, u need a P5B-DL Mod with bios=0804 for this chip

Which mods are needed exactly ?

Super Pi v1.5 Overclocked
Starting off below the motherboard FSB limit of ~345FSB at 9x333FSB with 2:3 divider and 500mhz 4-4-4-8 on memory with 1067fsb strap strap.

@3001Mhz at 1.336v (1.375v bios)



Single 1M & 2M


Single 4M & 8M


Single 16M & 32M


Cinebench with FX5200 PCI video card

[-Sweeper_]

how much q6600 will cost? what about yorkfield? i could find one of these quad-core cpus for around ~U$600~U$700 at their launch? (sorry 4 my bad english)

[eva2000]

AFAIK seen Q6600 slated for Q1 2007 @US$851 I think.

Swapped from 2GB Gskill PC6400PHU2 HZ Micron D9GMH 0629 batch memory to 2GB OCZ PC2-8000 Titanium Alpha VX2 Micron D9GKX memory.



System:

• Intel Core 2 Quad Q6600 Kentsfield QUPT B3 Step 7
• Corsair Hydrocool 200EX
• Intel D975XBX Badaxe rev304 1441 bios
• 128MB Gainward FX5200 PCI card
• 2x1GB OCZ PC2-8000 Titanium Alpha VX2 Micron D9GKX
• 3x320GB Seagate 7200.10 SATAII Matrix Raid 0
• 1x74GB Western Digital Raptor 8MB
• Pioneer DVD-RW
• FDD
• 1KW PCP&C SLI PSU

Re-ran some of the single Super Pi v1.5 32M Pi times with stock @2400Mhz 9x266fsb with 2:3 divider and OCZ Titanium AlphaVX2 @400mhz 3-3-3-8 at 2.29v vdimm.



Super Pi 1M & 32M




(Due to higher async upclocked memory speed and tighter timings of 3-3-3-8, the 32M time is 98+ seconds faster than 1:1 266mhz 4-4-4-8 timings!)

Super Pi 8M & 16M


(Due to higher async upclocked memory speed and tighter timings of 3-3-3-8, the 16M time is ~45 seconds faster than 1:1 266mhz 4-4-4-8 timings!)

Dual 32M


(Due to higher async upclocked memory speed and tighter timings of 3-3-3-8, the dual 32M Pi time is ~150 seconds faster than 1:1 266mhz 4-4-4-8 timings!)

[eva2000]

Time for some 3dmark/pcmark action with 2x 256MB BFG 7800GTX SLI with 85.96 SLI drivers

System:

• Intel Core 2 Quad Q6600 Kentsfield QUPT B3 Step 7
• Corsair Hydrocool 200EX
• Intel D975XBX Badaxe rev304 1441 bios
• 2x 256MB BFG 7800GTX SLI 85.96 drivers
• 2x1GB OCZ PC2-8000 Titanium Alpha VX2 Micron D9GKX
• 3x320GB Seagate 7200.10 SATAII Matrix Raid 0
• 1x74GB Western Digital Raptor 8MB
• Pioneer DVD-RW
• FDD
• 1KW PCP&C SLI PSU

Done with Intel Core 2 Quad Q6600 at stock 2400Mhz 9x266fsb and 2GB OCZ PC2-8000 Titanium Alpha VX2 @400mhz 3-3-3-8.

2x 256MB BFG 7800GTX SLI @494(546)/1370

click image for full screen shot

PCmark05


3dmark06 & 3dmark05


3dmark2003


3dmark2001


Aquamark3

[eva2000]

Still with 2x1GB OCZ PC2-8000 Titanium Alpha VX2 Micron D9GKX this time with Intel Core 2 Quad Q6600 @9x333fsb = 3001Mhz and 2:3 divider for 500mhz 4-4-3-4 with tightened memset v3.0 timings.

Single 1M & 8M


Single 32M


(~18 seconds faster than Gskill PC6400PHU2 HZ same cpu clock but memory at 500mhz 4-4-4-8 and no memset tweaks)

[GWillakers]

Just a question and a comment:

Q: It seems you consider the reseting of Vcore to a value below stock to be a fault of the D975XBX. It seems to me that this would be more of a fault of Everest / Sandra. No?

Comment: You wondered if the Temperature reported for the Dimm in Everest was somehow attributed to 1441. I don't know Everest, but I believe it is reporting a value for System Zone 2, which is measured between the Dimm Banks. The value it reports for "Motherboard", I believe would correspond to what Intel refers to as System Zone 1, which is measured close to the processor package, under the Heatsink Fins.

[eva2000]

Yeah just found the D975XBX temperature zone docs and zone 2 is near dimm slots..

as to fault, everest runs fine on my Asus P5W DH without vcore resets.. so is it solely a everest issue or D975XBX ?

Sweet Q6600 clocked at 3001Mhz (9x333fsb) with memory 1:1 @333mhz 3-3-3-8 and 2x 256MB BFG 7800GTX SLI @473(525)/1350 managed to break all my previous 7800GTX SLI personal best scores for 3dmark06

3dmark06 = 9,633
http://service.futuremark.com/compare?3dm06=620460



Higher we go

3dmark06 = 9,787
http://service.futuremark.com/compare?3dm06=620581



Look at that cpu score

[thorgal]

Consider yourself lucky EVA2000 I only manage 280FSB (yes, you read right ) with my Bad Axe and QX6700B1, while the chip does 4Ghz on a P5B-del (had it up to 408x10, but wasn't really stable).

Maybe it's because I got an older rev. 302 modded Bad Axe, but this was extremely disappointing nonetheless. Multiplier change does work however, up to 13, above no go, even on Phase.

[eva2000]

ouuuuch 280fsb... dam then i really do consider myself lucky!

what was max fsb on P5B DX for QX6700 before and after full volt mods ?

[thorgal]

ouuuuch 280fsb... dam then i really do consider myself lucky!

what was max fsb on P5B DX for QX6700 before and after full volt mods ?

I did not test it before the vmods, so I can't help you there.

To be honest, I've not even found the max FSB with my Kents on lower multi's, lack of time I'm afraid. My step 5 Conroe got up to 458x10 on single stage on this mobo. I'm finishing up with a DDR2 roundup review at the moment and try to get more data for you later.

[eva2000]

I did not test it before the vmods, so I can't help you there.

To be honest, I've not even found the max FSB with my Kents on lower multi's, lack of time I'm afraid. My step 5 Conroe got up to 458x10 on single stage on this mobo. I'm finishing up with a DDR2 roundup review at the moment and try to get more data for you later.

Ah DDR2 review.. on your sig

Looking forward to that review

Decided to see how well tight 3-3-3-8 timings compare to high clocked 4-4-4-8 timings. Only difference is 4-4-4-8 timings were done on 2GB Gskill PC6400PHU2 HZ Micron D9GMH 0629 batch @500mhz 4-4-4-8 and tight 3-3-3-8 timings were done with 2GB OCZ PC2-8000 Titanium Alpha VX2 Micron D9GKX @417mhz 3-3-3-8 at 2.39v

System:

• Intel Core 2 Quad Q6600 Kentsfield QUPT B3 Step 7
• Corsair Hydrocool 200EX
• Intel D975XBX Badaxe rev304 1441 bios
• 128MB Gainward FX5200 PCI card
• 3x320GB Seagate 7200.10 SATAII Matrix Raid 0
• 1x74GB Western Digital Raptor 8MB
• Pioneer DVD-RW
• FDD
• 1KW PCP&C SLI PSU

vs



Intel Core 2 Quad Q6600 Kentsfield quad core cpu was clocked on both tests at 9x333FSB = 3001Mhz.

2:3 divider 500Mhz 4-4-4-8 vs 4:5 divider 417mhz 3-3-3-8



vs



Results

500mhz 4-4-4-8 (left) vs (417mhz 3-3-3-8 (right)

1M
16.907s vs 16.812s

2M
40.750s vs 40.750s

4M
1m 34.454s vs 1m 34.000s

8M
3m 31.156s vs 3m 31.157s

16M
7m 44.281s vs 7m 44.141s

32M
16m 48.438s vs 16m 47.562s

Amazing that 417mhz 3-3-3-8 timings not only kept up with but also beat 500mhz 4-4-4-8 timings in 4 out of 6 of the Super Pi times. Especially in 32M Pi test! I suspect it's the tighter timings but also due to tighter default tRD value. According to memset v3.0, 417mhz 3-3-3-8 had by default tRD = 5 whereas 500mhz 4-4-4-8 had tRD = 7 by default.

Probably explains why PC6400 rated DDR2 memory in 3-3-3-8 default timings usually cost as much if not more than PC8000 rated memory due to it's performance as well as the time and cost it takes for the hand testing and binning process!

[thorgal]

That is very interesting info you've got there EVA2000 ! Strangely enough, your results do not represent my findings so far. I do not know the figures by heart, but on my P5B-del PC8000C4 always outperformed PC6400C3... I hope to show this to people in the upcoming review (don't want to hijack your thread now )

Edit : my testing is done on Conroe by the way

[thunderstruck!]

Very nice eva. What is max prime stable? How does it compare to your E6600 ES?

[eva2000]

thanks thorgal, looking forward to that review

thunderstruck, badaxe board is fsb limited for kentsfield at 340-345fsb so not going any higher for now with max 9x multiplier .... this Q6600 was pretested on P5B DX at 3600mhz 9x400fsb 1.45v quad stressprime on air...

I have P5B DX on order next week heh

[eva2000]

To be more thorough, I also tested OCZ PC2-8000 Titanium Alpha VX2 at same 500mhz 4-4-4-8 timings and default bios boot up as the Gskill PC6400PHU2 HZ and compared them to the existing results.



Gskill PC6400PHU2 HZ D9GMH @500mhz 4-4-4-8 (left)
vs
OCZ PC2-8000 Titanium Alpha VX2 @417mhz 3-3-3-8 (middle)
vs
OCZ PC2-8000 Titanium Alpha VX2 @500mhz 4-4-4-8 (right)

1M
16.907s vs 16.812s vs 16.766s

2M
40.750s vs 40.750s vs 40.578s

4M
1m 34.454s vs 1m 34.000s vs 1m 33.609s

8M
3m 31.156s vs 3m 31.157s vs 3m 30.218s

16M
7m 44.281s vs 7m 44.141s vs 7m 42.282s

32M
16m 48.438s vs 16m 47.562s vs 16m 43.828s

Plotting the graph two different ways shows:





Looks like the Micron D9GKX based OCZ PC2-8000 Titanium Alpha VX2 is just all round faster than the Micron D9GMH based Gskill PC6400PHU2 HZ 0629 batch memory for both 4:5 417mhz 3-3-3-8 and 2:3 500mhz 4-4-4-8. The difference is small, but for those Super Pi junkies like me, it looks like interesting times a head for my Super Pi adventures with this Micron D9GKX based OCZ memory

But as folks will ultimately also factor in price, the Gskill PC6400PHU2 HZ Micron D9GMH are probably a better bang for your buck being on average 35-45% cheaper in Australia

[thorgal]

Again great work EVA2000

The differences you point out also come back in my testing: Teamgroup PC5300 "333" vs Teamgoup PC6400 "333" : the latter with GKX chips performs a little better than the former with the same speeds, but scale much higher. You'd also have to pay double for the PC6400C3 of course.

For now : teamgroup GMH go up to about 1050-1070cas4, GKX go to 1150cas4 (actually I'm testing cas 5 orthos at 1250Mhz as we speak, so far no prob).

All testing with max. of 2.4VDDR

[eva2000]

Yeah this is my first set of D9GKX but looks like they're really binned for speed/performance