Intel Core i5 & i7 Sandy Bridge Overclocking and Feedback

[PaulieG]

Get the Thermaltake Fro its easily the best aircooler.

No, it's not. Don't get me wrong, the Frio is a decent cooler. I tested it for a couple of days. However, it does NOT compare to the Venomous X or D14. Not even close. I believe when compared in reviews it falls behind the Megahalems too which is a half step behind the coolers I've already mentioned.

Did you just do the Multi and turn off Turbo ?.

This was all turbo. Went back and set the multi to a fixed ratio after I tested 4.8 with turbo. Reasonably stable either way. I was playing around a bit. I can get it to Windows at 5.1ghz on 1.41v, but not for very long.

Oh, and just a friendly reminder...no double posting. Please edit your post if you need to add something. Thanks!

 

[da20valve]

Frio

I have to disagree, I use to use the Venomous X and swapped to the Frio and the Frio was cooler, on another note i did a custom build for a customer today at work and used the Mega Shadow, (PT-MegaShadow) and i have to say wow at 4.8 (4.9Ghz Turbo) this thing didnt get over 56 degrees with 1.365v CPU and 1.9 PLL. I am very impressed with all the PT products.

 

[PaulieG]

I have to disagree, I use to use the Venomous X and swapped to the Frio and the Frio was cooler, on another note i did a custom build for a customer today at work and used the Mega Shadow, (PT-MegaShadow) and i have to say wow at 4.8 (4.9Ghz Turbo) this thing didnt get over 56 degrees with 1.365v CPU and 1.9 PLL. I am very impressed with all the PT products.

It's interesting that your results were quite different than most users. Maybe you had a VX with a defective base or something. It happens. Per most reviews and my own personal experience, the VX is a much better cooler than the Mega. I say that after being a huge Mega fan since it was released. However, I put my Mega on my back up rig after a 5c load temp drop with the VX. The Mega Shadow is really no better than the Mega either. Nickel plating does very little to increase heat dissipation. The only 2 air coolers that may cool better than the VX right now are the D14 and Silver Arrow. Both of which are much larger than the VX.

I'd love to see a screen shot of that 56c on full load at that vcore.

Let's move back on topic....you'd think it was a heatsink thread, and it's partially my fault. Maybe another thread to compare these heatsinks?

 

[LAN_deRf_HA]

If your load temps are 56c at 4.8 something is amiss. I mean basically everything you've said since you entered this thread has made no sense, contradicting the experiences of just about everyone. Are you sure your monitoring programs are up to date? Can you describe your stress testing procedures? And why on earth would you run a 1.9v pll?? The only way I can figure that makes sense is if you're doing this in a shop directly under the air vent of a powerful ac unit.

 

[LifeOnMars]

Here's a question for ya peeps. I just did a Prime run for an hour at 4.7GHz. I had Real temp and HW Monitor running as well. 3 of the cores had done 78 tests, one of them, only 58. There were no errors at all. Was it that core powering HW monitor and real temp or just a case of a failing core which would have eventually shown errors?

 

[da20valve]

I already sent the machine out but i will be doing another 4.8ghz and a Mega shadow on Monday, so i will post screen shots, also next week i have 2 of the same systems to build 1 with a Frio and 1 with a H70 so i will post them too.

 

[Undead46]

Couldn't get 4.8 stable at 1.375V on mine. :/
Had to use 1.38V which I don't want to use for 24/7...

So dropped down to 4.7 and currently running 1.37V, gonna keep going lower till unstable.

 

[Frizz]

Just bought and paid for a pre-binned i7 2600k from sno.lcn I can't wait to play with it!

 

[LifeOnMars]

Here's a question for ya peeps. I just did a Prime run for an hour at 4.7GHz. I had Real temp and HW Monitor running as well. 3 of the cores had done 78 tests, one of them, only 58. There were no errors at all. Was it that core powering HW monitor and real temp or just a case of a failing core which would have eventually shown errors?

Anyone?

Random, I think I will be going for a 2600K after seeing how impressive these 2500k's are. Let us know how it goes mate

 

[da20valve]

Couldn't get 4.8 stable at 1.375V on mine. :/
Had to use 1.38V which I don't want to use for 24/7...

So dropped down to 4.7 and currently running 1.37V, gonna keep going lower till unstable.

You have to add PLL voltage, 90% of the time if you add PLL voltage you can DECREASE your CPU voltage, Alot of people on here are running huge CPU voltages and its just not nessecery.
BTW with a aftermarket cooler 1.38V is totally exceptable to run 24/7,

Also on your chip you will probably find that 4.7ghz might be the sweet spot (alot of the 2600k's are 4.7-4.8Ghz stable), most chips these days overclock fairly well with a certain voltage, then they require a SH!T Load more to get the extra 200Mhz and its just not worth it. Just Leave it at 4.7 reduce voltage till its unstable then increase it by 2 increments (that will make sure its not running on the bare minimum).

 

[Undead46]

You have to add PLL voltage, 90% of the time if you add PLL voltage you can DECREASE your CPU voltage, Alot of people on here are running huge CPU voltages and its just not nessecery.
BTW with a aftermarket cooler 1.38V is totally exceptable to run 24/7,

Also on your chip you will probably find that 4.7ghz might be the sweet spot (alot of the 2600k's are 4.7-4.8Ghz stable), most chips these days overclock fairly well with a certain voltage, then they require a SH!T Load more to get the extra 200Mhz and its just not worth it. Just Leave it at 4.7 reduce voltage till its unstable then increase it by 2 increments (that will make sure its not running on the bare minimum).

What's a recommended PLL?

 

[da20valve]

1.8V is standard so anything under or including 1.88-1.9v, people are running higher but i dont recommend it, Unless you have a good case with good airflow and a good mainboard like the UD3, UD4, UD5 or good Asus P8P67. Depending on your mainboard the Early 1155 sockets didnt support PPL Overvolting, only the later ones did, the new B3's overvolt it on there own (when turbo kicks in) But they are 1.8v standard.



I see yours is a B3 so when you set the PLL voltage make sure you ENABLE THE PLL OVERVOLT in the bios (standard is AUTO).
Start of with something reasonable like 1.85v or around that then work around that figure.

 

[LifeOnMars]

I'm preferring you alot more tonight than earlier....no hard feelings? What do you reckon about my little puzzler earlier regarding prime?

 

[Frizz]

Anyone?

Random, I think I will be going for a 2600K after seeing how impressive these 2500k's are. Let us know how it goes mate

lol I just wish I had my 2nd 6950 to bench it in crossfire. Its too bad but that is all I'd have used the extra GPU power for anyway.

I got one that is apparently stable at 5.4ghz with 1.48vcore, I guess I can use this as a goal for my crappy overclocking skills as in If I can't reach it we'll know why

 

[da20valve]

By any chance was it the 4th core that was lacking. ?

 

[LifeOnMars]

By any chance was it the 4th core that was lacking. ?

No da it was the 3rd core Just seems a bit strange. I have been gaming hard with it all night at the same clock and not one bit of bad news to report. Normally Bad Company 2 flags a bad overclock straight away.

 

[da20valve]

Do yo have all the Turbo mulipliers set to the same on all cores. ?

 

[Undead46]

Do yo have all the Turbo mulipliers set to the same on all cores. ?

You should set them to all cores.

 

[da20valve]

I know, but sometimes people cant get a stable overclock so they set the multi to 47 THEN they set the TURBO MULTI to 48 on the first 3 cores and auto on the 4th core, That way your still getting 4.7Ghz across all cores and 4.8Ghz on 3 of the cores when turbo kicks in.

 

[LAN_deRf_HA]

1.8V is standard so anything under or including 1.88-1.9v, people are running higher but i dont recommend it, Unless you have a good case with good airflow and a good mainboard like the UD3, UD4, UD5 or good Asus P8P67. Depending on your mainboard the Early 1155 sockets didnt support PPL Overvolting, only the later ones did, the new B3's overvolt it on there own (when turbo kicks in) But they are 1.8v standard.



I see yours is a B3 so when you set the PLL voltage make sure you ENABLE THE PLL OVERVOLT in the bios (standard is AUTO).
Start of with something reasonable like 1.85v or around that then work around that figure.

I guess it's just your wording but you make it sound like pll overvolt needs to be on to adjust the pll voltage. They're two separate things. You can drop pll down to reduce temps without affecting stability and flip on overvolt for a higher multi.

 

[LifeOnMars]

Do yo have all the Turbo mulipliers set to the same on all cores. ?

Yes its just set to 47, no individual core differences.

 

[da20valve]

No sorry it may be worded incorrectly, I was just stating that when increasing PLL voltage that PLL OVERVOLT SHOULD be set to ENABLE when running high clocks with a big Turbo Multi.

Either way if your increasing the PLL VOLTAGE on the new B3's overvolt should be enabled.

 

[LAN_deRf_HA]

Speaking of, so pll overvoltage was on for all or a lot of those 4.8ghz clocks?

 

[Undead46]

Well, I was only able to get 4.7GHz stable at 1.38Vcore and 1.9V PLL.
Couldn't get 4.8GHz without going over 1.38V... :/

Currently stressing my RAM OC @ 2133MHz 1.65V 7-10-8-27-1N.
So far it's going okay.

 

[LAN_deRf_HA]

1.47GHz stable at 1.38Vcore

Man, you have the worst chip in the history of chips
Guessing you meant 4.7 and 4.8