C2D Overclocking & EIST

[lbax]

Thanks to Wusy for the great C2D Overclocking Guide!!!

During the preparation phase, The Guide instructs to;
8. Disable ‘C1E’
9. Disable ‘EIST’

There is no indication at the end of the Guide if these functions can be reinstated. I have seen conflicting discourse if these functions can be used on OC'ed C2Ds.

Wusy, can you provide some insight regarding using these functions with OC'ed C2Ds? Risks & roadblocks.

THANK YOU!

 

[etienc]

I found out something interesting...

On my Gigabyte DS3, if I manually set the clock multiplier in BIOS, the EIST feature in BIOS suddenly disappears.

The even stranger part is that when I reset Fail-Safe defaults, the EIST feature does NOT re-appear (unless I re-flash the BIOS).

What the heck is up with that??

 

[levicki]

I found out something interesting...

On my Gigabyte DS3, if I manually set the clock multiplier in BIOS, the EIST feature in BIOS suddenly disappears.

The even stranger part is that when I reset Fail-Safe defaults, the EIST feature does NOT re-appear (unless I re-flash the BIOS).

What the heck is up with that??

I had such bugs with Asus and Gigabyte BIOSes.

C1E is just like regular C1 power state just with lower power consumption hence E (for Enhanced).

 

[rquinn19]

Something I needed to add in my guide as well, TM2 (Thermal Monitor2), that goes hand in hand with C1E which lowers voltage when CPU is idle (but kept at same speed if EIST is disabled). Extreme bad news when an overclocked CPU is in idle...

Is that the F4 BIOS you're using? (I thought they fixed almost all the bugs :? )

when should we be expecting an updated guide?

 

[lbax]

Extreme bad news when an overclocked CPU is in idle...

Yes it is. EIST, TM2 & C1E are all very desirable functions to have operating effectively with OCed processors. Having the CPU running full speed++ at idle is not my idea of a livable solution. I am not willing to trade away power, heat & noise management for the short bursts of performance that I use. I can only hope that BIOS updates &/or CPU steppings will smooth out the utilization of these functions.

Wusy, Thanks once again for The Guide & your service to all of the C2D community.

 

[crow_smiling]

During the preparation phase, The Guide instructs to;
8. Disable ‘C1E’
9. Disable ‘EIST’

There is no indication at the end of the Guide if these functions can be reinstated. I have seen conflicting discourse if these functions can be used on OC'ed C2Ds.

RMClock will allow you to re-enable EIST and C1E, at least on the P5B Deluxe with the current BIOS. It should hopefully work with other boards to.

Edit. I’ve just tested it with a second board that uses the older 945G chipset and it works with that to.

 

[crow_smiling]

EIST, TM2 & C1E are all very desirable functions to have operating effectively with OCed processors. Having the CPU running full speed++ at idle is not my idea of a livable solution. I am not willing to trade away power, heat & noise management for the short bursts of performance that I use.

If you are happy to over-clock using the stock VCore range, then you can still use all these features whilst doing so.
I know it’s almost sacrilegious to some of you for me to suggest over-clocking at stock VCore, but C2D does pretty well at stock voltage and doing this allows a balance between Speed, Power consumption, heat and noise to be achieved. There’s more than one way to over-clock.

 

[etienc]

Yes, the EIST "vanishing act" is with the F4 BIOS.

I am not willing to trade away power, heat & noise management for the short bursts of performance that I use.

This is why I've been considering just biting the bullet and running my C2D E6400 at stock speed. It's nice idling at 32C rather than 38-40C. Had my FSB clocked to 400Mhz, which of course helps in memory benchmarks, but not sure if it's actually making a difference in real-world apps (I mainly do gaming). It kills me to run the FSB at 266 though, just seems so slow :|

RMClock will allow you to re-enable EIST and C1E, at least on the P5B Deluxe with the current BIOS. It should hopefully work with other boards to.

I'll have to try out this utility, even though I don't like when software controls my OCing. What version do we need to use for the C2Ds? (I think I remember seeing a regular and beta?)

 

[crow_smiling]

RMClock will allow you to re-enable EIST and C1E, at least on the P5B Deluxe with the current BIOS. It should hopefully work with other boards to.

I'll have to try out this utility, even though I don't like when software controls my OCing. What version do we need to use for the C2Ds? (I think I remember seeing a regular and beta?)
I’m using version 2.1. You aren’t really using software for over-clocking; you are just using it to change the Speedstep settings.

It's hard to explain in one go, but the basic concept just like with laptop (except it's OC'ed this time) is to force down in multiplier and test voltage at every notch(multiplier) until you reach 6x. It's an extremely time consuming session as you're trying to find the lowest possible voltage for every multiplier at the set FSB.
As mentioned above, that's only possible IF RMClock provides more than 1.3525V

I suggest disabling all the intermediary multiplier values from within RMClock and just use the min and max ones available. It saves time and the intermediary ones aren’t often used that much anyway.
You can still go through this procedure regardless of RMClock not being able to over-volt; you are just limited to ~1.1x – 1.32x volts.

The way crow_smiling described with stock voltage would defeat the purpose of pure control overclocking. (Assuming RMClock does not support more than 1.3525V)

I don’t know what the term ‘pure control overclocking’ means, but the method that I outlined above is only useful and necessary in a limited number of scenarios.
E.g. if you enable C1E and EIST in the BIOS and overclock an E6300 to 3GHz using stock VCore, you will need a FSB of 429. This will mean at idle you will have 6 * 429 = 2.57GHz. The default VCore at idle with Speedstep is ~1.15V, which is very likely not enough for the CPU to be fully stable at 2.57GHz.
By using the method that I outlined above, the system will boot at 1.32xV and be stable and once RMClock is loaded it will re-enable EIST/C1E and set stable voltages.
This process shouldn’t be needed with an E6600 as the idle clock speed would only be 2GHz if the max speed was again set to 3GHz; 1.15V should be fine for that clock speed.
With an E6400 it is marginal. I tested an E6400 using this method, it was overclocked to 2.9GHz and at idle it was at 2.17GHz 1.15V and it wasn’t 100% stable. It’s the luck of the draw of course as other chips may do better.

 

[crow_smiling]

As you said down at the buttom, luck is something I do not take into account.

When it comes to over-clocking luck plays a part, some chips are just plain better than others, if you get one of those you are lucky, capiche!

If you don't like then don't bother even to overclock. You are in a proper overclocking forum afterall. All comments here are strictly limited to proper techniques of overclocking not half-arsed ones.

As I hinted at before, I think you are an overclocking dogmatist. There are many styles of overclocking and unless a moderator tells me that my ideas aren’t relevant here, I’ll keep posting. Please keep your opinions as to what is valid as an overclocking style to yourself. Not everyone is into the Mussolini style of overclocking.

Please remember what forum you're posting in right now. This is not a playground. it was I who introduced the concept of using RMClock into TGForumz, but it was intended only for the CPU forum not to be mentioned or used in this OC-CPU forum.

Did you write RMClock, are you a moderator? If the answer to both these questions is No, then why do you think I would be vaguely interested in your opinions on this issue? I’m not, so please keep them to yourself.

 

[crow_smiling]

I’m not, so please keep them to yourself.

This is a question for my guide. No one asked for your ameture comments.
You didn’t seem to know the answer to the question and when I pointed out what the answer was, this seems to have bugged you. Is this why you are so prissy!

No one knows everything, who cares, I’m sure you know more about C2D than I’d care to know, but it doesn’t mean you know everything; that’s life.

 

[crow_smiling]

As I said, keep your none serious suggestions in Black_Knight_MC's thread please.

Ja, mien fuehrer.

 

[jjohan35]

despite all the girly bickering, i like crow_smiling's suggestion and ideas. they explained a lot to me regarding intel's chipset and the limitations.

deflate the egos.

 

[Blouge]

>There is only one style and only one type of overclocker. You're either one or you're not and are a tweaker.

Someone who increases FSB by 1 MHz has both tweaked and overclocked their PC and is consequently a tweaker and an overclocker. Similarly, someone who decreases FSB by 1 MHz is a tweaker and an underclocker.

My current PC is E6600 at 8x400 = 3.2GHz and I leave the options C1E, EIST, SpeedStep, and Execution Bit checking turned on. It's stable with air-cooling and 1.3125 vcore - way less than the suggestion in the guide (1.4) which exceeds Intel's max spec. I also use minimal voltages for NB, SB, and MCH. I generally like the guide, but I regret having needlessly jumped in above max spec voltage for everything.

 

[jjohan35]

Someone who increases FSB by 1 MHz has both tweaked and overclocked their PC and is consequently a tweaker and an overclocker. Similarly, someone who decreases FSB by 1 MHz is a tweaker and an underclocker.

and someone who increases their voltage by .0001 is an overvolter, and someone who decreases their voltage by .0001 is an undervolter. there is no such thing as a 'pure overclocker'.