GD80/UD6 and 750 at 5.2G

[Massman]

Time to update this thread. What has happened in the last days:

 

- Released Core i5 article

- Released LGA1156 motherboard round-up

- Retested UD6 for F1-OC

- Retested GD80 with new bios for F1-OC

 

As this is a thread about the boards' OC capabilities, I won't comment on the findings in both articles. Check them out if you want, don't if you don't want to.

 

Hey guys,

 

As we haven't recieved our retail DFI P55 motherboard, I've just started to pretest my retail 750 on the Gigabyte UD6 ... you know: just in case!

 

Pushing i5 750 on the UD6

 

Results speak for themselves

 

 

 

 

Most cool of this session might actually be the validation I've been able to obtain:

 

http://valid.canardpc.com/show_oc.php?id=715625

 

A few things to note:

 

- You're only as strong as your weakest core. Core #1 does 5132MHz (as you'll see later) with only 1.64V and probably -60°C, but I need to decrease the temperature to -70°C to get my other cores to idle stable at that frequency. Core #0 is most definitly the weaker of all four.

- If you're pushing for BCLK, you need to focus on pushing the temperature. I drop in temps to -85°C/-86°C for the 251MHz BCLK validation, but that's really the limit for this configuration. The chip is, by the way, not at all benchmark stable at that temperature.

 

Pushing i5 750 on the GD80

 

As you might know, I wasn't totally convinced by the GD80 as it has some issues with voltage under extreme cooling. A couple of days ago I figured out what was the big issue, fixed it and now I can actually run at 1.6V+ VTT. First the result:

 

 

Notes:

 

- Although I did spend some time tweaking the memory timings on the UD6, it seems that the board is genuinly slower than the GD80. Other people contradict this finding so maybe I'm just lucky, but on air cooling I can do 6 seconds better on the GD80 at 4100MHz (8m37 versus 8m43). I can post the screenshots here if you want, just let me know.

- If you plan to use 1.6V VTT, increase the PCH 1.05V to 1.3V. I've asked MSI to make this happen automatically, but this is the key to make it happen. The reason is really simple: PCH is integrated on uncore die, so ...

- My initial observation of the GD80 being better in terms of clock per volt seems to be correct. With the UD6 I'm maxing out at 5070MHz, with the GD80 at 5132MHz using the same volt and temp conditions.

[Massman]

Last update before the break.

 

- Tested UD6 more than a couple of times this week; same story: BCLK capabilities for 32M is inferior compared to the GD80. Efficiency at high clocks is inferior as well. It might be an issue with my configuration/OS/skill as other people are in fact producing quite good scores.

- GD80 is back to HQ to get a couple of updates, for instance PCI-e clock improvements.