Overclock.net banner

What do you think 45nm chips will take?

  • 1.3 -1.4 Vcore

    Votes: 122 20.0%
  • 1.4 -1.5 Vcore

    Votes: 238 39.1%
  • 1.5 -1.6 Vcore

    Votes: 249 40.9%
1 - 20 of 343 Posts

·
Banned
Joined
·
4,710 Posts
Discussion Starter · #1 ·
Right guys, My test subject arrived today, A Pentium Dual Core E6300.

This chip is based on the 45nm 'Wolfdale' core and has 2mb of L2 cache,

Batch Number : Q911B009
Stepping : SLGU9 ( R0 )

Now the whole point of me doing this is so we have some evidence of just what Vcore these 45nm chips can run with 24/7 witout running into some form of performance degradation or in the worst case, The chip actualy failing completely.

I shall ONLY be adjusting Vcore and not PLL which would kill the chip in minutes.

Asking many people on this forum on what they think is the safe max voltage for a 45nm chip will result in loads of different answears, FACT is that no one really knows. We all know what Intels spec sheet says on the matter but I think that Intel are being very conservative.

I think with the right cooling they should be fine with a Vcore between 1.5 and 1.6v for 24/7 use. Afteral if a chip is said to last 20 years with its stock Vcore and running it at 1.5-1.6 Vcore reduces that life to 5 years does it matter? By the time that 5 years comes around the chip will be too dated to be of any use anyway and if the extra Vcore only reduces the life span to 10 years then it's even better.

My Watercooling Loop :

XSPC Delta V3 water block
Laing DDC Ultra with XSPC reservoir pump top
Thermochill PA 120.3 ( The updated model with 15mm fan spacing )
3x Xigmatek 120mm Oranage/Black white LED fans in push

The Test :

Thinking of a test that would show that the chip had degraded was difficult but in the end I came up with a simple test of using Linx.

The plan is to overclock the chip as far as it will go with around 1.55Vcore while still keeping the temps at a 'reasonable' level, then run Linx for an hour until I hit a 100% stable overclock with that voltage, while at the same time making a note of the temps during the stress test.

Then at set intervals ( I thinking every 2 weeks ) run linx again to check if stabilty still remains, If the chip does show to be unstable then I would imagine that this shows that the excess voltage has degraded the chip slightly. Also temps will be noted and compared to the previous test.

I will then Increase Vcore to try and restore stabilty and make a note of the new Vcore that was needed to regain stability.

I'm open to sugestions on other tests that may show results better then mine


So can an ADMIN stickie this thread and make it an 'OFFICIAL' ?

Testing and overclocking starts tomorrow when my new fluid arrives and my loop is bled and ready to rock.

Any sugestions guys?
 

·
Premium Member
Joined
·
10,389 Posts
Good luck! I am betting much over 1.6 volts for more than a few minutes will result in some degradation. I am looking forward to this!
 

·
Premium Member
Joined
·
4,776 Posts
Wooo!!! Awesome dude. +Rep for you.

Even my cat smells the epic coming off my monitor.
 

·
Banned
Joined
·
4,710 Posts
Discussion Starter · #10 ·
Quote:


Originally Posted by mtbhrd
View Post

Maybe start off at 1.45v?

Otherwise, I can't wait for the results, though its gonna be awhile...

They'll be up by dinner time tomorrow IF I get my fluid for my loop...


Really really pisst me off that it didn't come today
 

·
Premium Member
Joined
·
4,776 Posts
Quote:


Originally Posted by mtbhrd
View Post

Maybe start off at 1.45v?

Otherwise, I can't wait for the results, though its gonna be awhile...

I second this suggestion. There are several people already running 1.4v+ through their 45nm. And Intel states that is the absolute max, even though it is well out of the operating area.

On second thought, maybe 1.5 would be better. Test the "unknown ground". Test beyond Intel's specs.

--->@almighty I'm trying to find the thread where a guy hit 6GHz. Maybe you can tinker with his settings. If I find it I'll pm it to you.<---
 

·
Banned
Joined
·
4,710 Posts
Discussion Starter · #13 ·
Quote:


Originally Posted by crashdummy35
View Post

I second this suggestion. There are several people already running 1.4v+ through their 45nm. And Intel states that is the absolute max, even though it is well out of the operating area.

Some people also claim that 1.55Vcore is the absolute max.....

I like 1.55v, It's a nice number
 

·
Premium Member
Joined
·
2,865 Posts
I've run my e8500 w/+1.7v @ -5ºC ambient water and she's still going strong.............that was only suicide runs..but still...
 

·
Registered
Joined
·
603 Posts
Sounds very interesting!

Just one question to the OP:
From the original Linx test to the next stability test (2 weeks after),

- Is the chip going to be running 24/7?
- What load will you give the chip?

I suggest loading with SMP folding, which will also detect if the setup becomes unstable.
 

·
Banned
Joined
·
4,710 Posts
Discussion Starter · #20 ·
To answear a couple of questions :

1. My computer WONT be on 247 and hence the chip wont be on 24/7, the whole point of this is to test it in the REAL world, not many people leave there computers on 24/7. My computer is on more then average though during the day.

2. Load will vary but again, just normal load, Games, movies...etc...etc... and then Linx when it comes to checking it every 2 weeks.

3. 1.6 Vcore just aint going to happen, again this is to simulate running high Vcore in everyday situations, 1.6 Vcore is as already mensioned for suicide runs and not for daily running. Alhtough if she is still running ok i might suicide it


On a side note these have been doing close to 5Ghz with round about 1.55Vcore over in the Pentium Dual core E6300 thread over at XS
 
1 - 20 of 343 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top