i7 overclocking - Page 2

VJS · 2009-02-22, 03:43

Lars that cooler certainly looks like it's doing the job.

The difference between the motherboard temp and the CPU temp at idle is a good indication if your thermal contact between the cpu and heatsink is any good.

32-34C at idle is very good, I would think that would mean good contact.

Considering your Motherboard temperature is climbing into the near 50C, I would totally agree and say additonal air flow through the case is required.

Another possibility for you... is it possible to direct the heat disipated from the CPU out of the case directly?

On my U-120 is was able to create a partition out of boxing tape that directed heat directly out the top of the case.

CADavis · 2009-02-22, 07:52

Hey I just built an i7 system also, will be watching this thread and probably posting progress once I decide on a waterblock.

i7 920
DFI Lanparty DK X58-T3eH6
Super Talent 6GB Triple Channel CL8 DDR3-1333
4x 250GB Western Digital Caviar Raid Edition 1 (RAID0)
PC Power & Cooling S61EPS 610W

Windows 7 Beta1 64-bits

ltd · 2009-02-23, 17:21

It looks like the thermal paste from noctua needs a little bit more "burn in" time. The average temperature stayed the same but now there are no real peeks anymore. Peek CPU temperature is now 66°C and there is no strong reaction to the case/MB temperature anymore.

Still running with all bios setting to auto.

VJS · 2009-03-03, 04:20

Any additional progress or comments??? (Current Ghz and temps stable)

flouran · 2009-03-03, 05:52

Quote:

Originally Posted by VJS

Any additional progress or comments??? (Current Ghz and temps stable)

I haven't really read any other posts on this thread, but I believe that Intel is modifying the pins on the current i7s and will release them next year (probably cheaper, but most definitely improved). So, it may be in your best interests to wait until then...

ltd · 2009-03-03, 17:14

@vjs: To be honest i was to lazy to do any overclocking work. No promise when i really start to do it. But without overclocking the system is extremely stable.

@flouran: Intel has all the time some plans for newer/better cheaper systems. With that argument i would never be able to buy a new computer.

CADavis · 2009-03-17, 06:49

Just got a new waterblock for the i7 system. Currently testing overclocks on the i7 920 for about 4 hours, gotten up to 3.8ghz stable for 1 hour and then 1 error. All core temps are staying below 61 C with 100% cpu usage (using Real Temp 3.00, find it on xtremesystems.org). Bclk 190mhz x 20.0 = 3800mhz. 1.39v Vcore. DRAM is triple channel DDR3 running at 760mhz (1520mhz effectively) at 1.55v, 9-9-9-27-1T.

Guessing that the RAM is at fault since it is quite a bit above rated stock, and a little above the highest SPD values, gonna try going to Bclk x 06 for the ram, will mean at 190mhz the ram will be 570mhz (1140mhz effectively, below rated stock speed of 1333mhz).

If that doesn't work then I will keep the ram at Bclk x 06 and try increasing vCore a little.

Here's a couple pics for now:
http://i2.photobucket.com/albums/y20...-03-16_001.jpg
The watercooling and main parts.

http://i2.photobucket.com/albums/y20...-03-16_002.jpg
The top view.

http://i2.photobucket.com/albums/y20...-03-16_003.jpg
The hard drive, power supply, and cd-rom side.

http://i2.photobucket.com/albums/y20...36ghz1hour.png
The 3.6ghz stable for 1 hour.

VJS · 2009-03-17, 12:55

Nice Rad!!!

I'm supprise the cpu temp are that high with WC on these i7's...

Don't have one yet however. Also try to run a few DC's once you think it's stable. I would also agree that its memory since people claim stability at much higher temperature and CPU Ghz.

CADavis · 2009-03-17, 20:54

little update for you guys:
http://i2.photobucket.com/albums/y20...1ghz7hours.png
4.01ghz stable for 7+ hours and counting.

I tried 4.1ghz before this but the system froze pretty quickly. also the "system" temp. was getting up to 80 C so I installed a 70mm fan bolted to one of the waterblock alll-thread bolts, pointed right at the NB and PWM heatsinks. Dropped it back down under 60 C!

I'm going to let it keep testing until I get over 24 hours stable here, then maybe try Bclk +1 or +2mhz, but I'm happy with 4ghz. Probably try one tick lower vCore, and perhaps I might try tightening RAM timings but I'm not sure yet.

Quote:

Originally Posted by VJS

Nice Rad!!!

I'm supprise the cpu temp are that high with WC on these i7's...

It's pretty normal actually. With stock cooler at stock speeds, normal is about 75C+ on each core. At stock speeds my WC kept it around mid 40s, low 50s max. At 4ghz and 1.4v, this thing is putting out around 275 watts!

VJS · 2009-03-18, 02:10

Note sure how the i7's are but the Q6600's wouldn't crash but were unstable (prime unstable) above 60C in my past experience. (4 q6600's worth)

I'm planning on getting a i7 in the near future but might wait for the tock or is it tick??? I've seen similar temps on a U-120 not sure if you can comment on that.

Have you considered using a set of 140's? That much difference due to fan's on a WC setup as well as the chipset temperature makes me think your don't have enough airflow or water flow. But 275W holycow!!! Have you checked your water temperature post waterblock?

BTW is that a DIY heatcore and which car, looks nice and massive as well as decent hook-ups or taps.

Nice nice work.

VJS · 2009-03-18, 02:15

O.K. last comment you can tell I'm interested!

Is that pump, pumping to the waterblock or to the rad?

I know there are varying stances but my stance is pump to the rad from the CPU. (Fluids are not compressable) and pumps add heat. Just curious.

2009-03-03, 17:14	#17
ltd Apr 2003 2²·193 Posts	@vjs: To be honest i was to lazy to do any overclocking work. No promise when i really start to do it. But without overclocking the system is extremely stable. @flouran: Intel has all the time some plans for newer/better cheaper systems. With that argument i would never be able to buy a new computer. Last fiddled with by ltd on 2009-03-03 at 17:14

2009-03-17, 06:49	#18
CADavis Jul 2005 Des Moines, Iowa, USA 2·5·17 Posts	Just got a new waterblock for the i7 system. Currently testing overclocks on the i7 920 for about 4 hours, gotten up to 3.8ghz stable for 1 hour and then 1 error. All core temps are staying below 61 C with 100% cpu usage (using Real Temp 3.00, find it on xtremesystems.org). Bclk 190mhz x 20.0 = 3800mhz. 1.39v Vcore. DRAM is triple channel DDR3 running at 760mhz (1520mhz effectively) at 1.55v, 9-9-9-27-1T. Guessing that the RAM is at fault since it is quite a bit above rated stock, and a little above the highest SPD values, gonna try going to Bclk x 06 for the ram, will mean at 190mhz the ram will be 570mhz (1140mhz effectively, below rated stock speed of 1333mhz). If that doesn't work then I will keep the ram at Bclk x 06 and try increasing vCore a little. Here's a couple pics for now: http://i2.photobucket.com/albums/y20...-03-16_001.jpg The watercooling and main parts. http://i2.photobucket.com/albums/y20...-03-16_002.jpg The top view. http://i2.photobucket.com/albums/y20...-03-16_003.jpg The hard drive, power supply, and cd-rom side. http://i2.photobucket.com/albums/y20...36ghz1hour.png The 3.6ghz stable for 1 hour. Attached Thumbnails Last fiddled with by CADavis on 2009-03-17 at 07:05

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
overclocking	wildrabbitt	Hardware	16	2015-03-12 02:12
Overclocking	Primeinator	Information & Answers	12	2013-09-16 16:20
Does overclocking help?	esqrkim	Hardware	7	2010-06-26 08:15
What is overclocking?	JuanTutors	Hardware	39	2004-06-26 21:06
What is overclocking?	JuanTutors	Hardware	1	2004-06-22 00:10

2009-02-22, 03:43	#12
VJS Dec 2004 13×23 Posts	Lars that cooler certainly looks like it's doing the job. The difference between the motherboard temp and the CPU temp at idle is a good indication if your thermal contact between the cpu and heatsink is any good. 32-34C at idle is very good, I would think that would mean good contact. Considering your Motherboard temperature is climbing into the near 50C, I would totally agree and say additonal air flow through the case is required. Another possibility for you... is it possible to direct the heat disipated from the CPU out of the case directly? On my U-120 is was able to create a partition out of boxing tape that directed heat directly out the top of the case.

2009-02-22, 07:52	#13
CADavis Jul 2005 Des Moines, Iowa, USA 252₈ Posts	Hey I just built an i7 system also, will be watching this thread and probably posting progress once I decide on a waterblock. i7 920 DFI Lanparty DK X58-T3eH6 Super Talent 6GB Triple Channel CL8 DDR3-1333 4x 250GB Western Digital Caviar Raid Edition 1 (RAID0) PC Power & Cooling S61EPS 610W Windows 7 Beta1 64-bits

2009-02-23, 17:21	#14
ltd Apr 2003 2²×193 Posts	It looks like the thermal paste from noctua needs a little bit more "burn in" time. The average temperature stayed the same but now there are no real peeks anymore. Peek CPU temperature is now 66°C and there is no strong reaction to the case/MB temperature anymore. Still running with all bios setting to auto.

2009-03-03, 04:20	#15
VJS Dec 2004 13×23 Posts	Any additional progress or comments??? (Current Ghz and temps stable)

2009-03-17, 12:55	#19
VJS Dec 2004 13×23 Posts	Nice Rad!!! I'm supprise the cpu temp are that high with WC on these i7's... Don't have one yet however. Also try to run a few DC's once you think it's stable. I would also agree that its memory since people claim stability at much higher temperature and CPU Ghz.

2009-03-18, 02:10	#21
VJS Dec 2004 13×23 Posts	Note sure how the i7's are but the Q6600's wouldn't crash but were unstable (prime unstable) above 60C in my past experience. (4 q6600's worth) I'm planning on getting a i7 in the near future but might wait for the tock or is it tick??? I've seen similar temps on a U-120 not sure if you can comment on that. Have you considered using a set of 140's? That much difference due to fan's on a WC setup as well as the chipset temperature makes me think your don't have enough airflow or water flow. But 275W holycow!!! Have you checked your water temperature post waterblock? BTW is that a DIY heatcore and which car, looks nice and massive as well as decent hook-ups or taps. Nice nice work.

2009-03-18, 02:15	#22
VJS Dec 2004 13×23 Posts	O.K. last comment you can tell I'm interested! Is that pump, pumping to the waterblock or to the rad? I know there are varying stances but my stance is pump to the rad from the CPU. (Fluids are not compressable) and pumps add heat. Just curious.