The system firmware has changed the processor's memory type range registers (MTRRs) across a sleep state transition (S4). This can result in reduced resume performance.

Ehhh... These canned responses don't hold water. I've seen this error on this forum multiple times, as well spread throughout the internet.

People are clicking "THEY NEED THIS ANSWER TOO"

This is an issue.

Hi, I discovered the cause of my problem - Intel Rapid Start Technology (IRST).

Basically there is a hidden hibernation partition (different to hiberfil.sys) used by IRST.

The IRST is set to activate by default (on my system) after 120 minutes.

So after 2 hours of sleeping, drive activity would commence and the active user session in RAM would be written to the Hibernation Partition (8GB in my case).

The problem came during resume, when pressing the power button, it failed to restore the session from the SSD to the RAM, and I got a failed boot.

A second press cold booted the system to a fresh desktop with any unsaved work lost.

A workaround in my case was to disable IRST either in the bios or in the systray utility.

Ultimately the solution was a new motherboard, which was put in by Dell yesterday.

Sleep / IRST & Hibernate all work fine now.

Interestingly hiberfil.sys (6.6GB) is ONLY used for booting in this setup:

Sleep - RAM

Extended Sleep (IRST) - hibernation partition

Re-boot - is now (under windows 8 with 'fast startup' enabled) effectively a 'cold boot' ie read from disk

Shut down / PowerOn - is now effectively a 'warm boot' as it reads from hiberfil.sys

So there is some disk redundancy with the 2 systems (IRST and 'Windows Hibernate'), you could effectively utililize just one system with little difference in performance.

To be honest I am hard pressed to measure any difference in speed between any of the following:

1. a hiberfil.sys assisted boot
2. a IRST 'resume'
3. a Windows Hibernate resume

All of these are getting my system to a login prompt in about 6 secs from power button press.

At the moment I have all of them enabled, but may change to reclaim disk space.

Not sure most of that was relevant to you!

G :)

 Couple of other points:

S0 = on

S1, S2, S3 are various forms of sleep, depending on manufacturer, generally RAM is powered all else is off.

S4 is hibernate or IRST

S5 is off / shut down

Hi I have a similar issue with Sleep in Windows 8 Pro (64-bit).

Normal shut down / boot is fine.

Hibernate is fine

Only sleep is a problem.

If sleeping for a short period with the powerlight pulsing the system resumes fine. However if left for an extended period, e.g. overnight. Then the powerlight is out.

Pressing the power button result in failure to boot - powerlight comes on briefly, then POST, then dies.

Second press boots the system, stating in the log that the system failed to shut down cleanly.

So basically in the sleep condition at some point the system 'powers down' ' 'dies', this is neither hibernate (which is set to never) or shutdown. And neither has the battery depleted causing a low power / critical level condition causing hibernate as the battery will be close to full (>90%) on a successful boot.

Obviously any open documents / work is lost in this condition as power to RAM is lost and hibernate does not occur.

I have tried disabling 'fast startup'

BIOS and all drivers are latest

Windows updates all installed.

Have tried a restore from factory image and re-install of all the above and problem remains. 

Any help appreciated.

Dell Vostro 3360, Core i5, 8GB RAM, 128GB SSD

hiberfil.sys is 6.6GB

 

Considering the age of the thread and to avoid confusion, I would recommend creating a new thread so that the community can focus on that thread. Refer to the following link prior to posting a question so that you may provide all the relevant information, for us to help you:

http://support.microsoft.com/kb/555375

Mr. Kumar [support engineer],

If you feel that the original Question has been confused by a multitude of voices; please say so. At least try to re-direct the thread. Your suggestion to begin a new Thread for the Question simply adds to the confusion And avoids addressing the original Question.

This Community Forum IS* MS's Customer Support !!!

Attention MS Community Forum leaders:

"We have a Problem".

At least 14 MS users have come here hoping for an answer to a Question. !!

None have received a valid answer.

The "Moderator" of the thread only adds to the confusion by suggesting the thread be close. This suggestion is now removed or deleted; yet the Moderator remains silent.

So far, 14 MS users have checked the 'Me Too' Box confirming they 'had' the same Question.

a Post by Mr. Kumar, a Community Star and MS Support Engineer, neither helps nor helped *no one. Yet it remains his sole response MS Forums users.  One would hope that another MS-C/S person might pick up this thread and provide an answer to the question; or at least add a new perspective.

*The original poster deserves an answer, and I assume may still need an answer. Does he have one?

 If not, then this "Question" and this *Thread should stay 0pen, un-answered, while Pleading for support.

where is MS? !!!"

"I have this problem too!!"

10.11.2013