Detecting incorrect Ram amounts

[tktk88]

I am experiencing different amounts of Ram being detected off and on typically at cold boot. I have 10 megs in the computer Sometimes windows and the bios will see 2 or 6 meg of ram and not the whole 10 meg. Typicaly after a reboot or 2 and the computer being warm it detects all of the ram. CPU-z will always see all of the ram even if windows and the bios is showing less than the total of 10 that is installed.

My Ram is ADATA Gaming Series DDR3 1600 AX3U1600GB2G9-BG Settings per label on Ram CL 9-9-9-24 Voltage 1.55V~1.75V

 

[Petey7]

It sounds like either a problem with the RAM or the motherboard. I would try flashing the BIOS with the latest version to start with.

 

[A Guy]

My Ram is ADATA Gaming Series DDR3 1600 AX3U1600GB2G9-BG Settings per label on Ram CL 9-9-9-24 Voltage 1.55V~1.75V

How do you have 10 Gigs of Ram? how many sticks and what size?? You should be using 3 sticks in triple channel on your setup. Use slots 2, 4, and 6 for triple channel. A Guy

 

[Petey7]

My Ram is ADATA Gaming Series DDR3 1600 AX3U1600GB2G9-BG Settings per label on Ram CL 9-9-9-24 Voltage 1.55V~1.75V

How do you have 10 Gigs of Ram? how many sticks and what size?? You should be using 3 sticks in triple channel on your setup. Use slots 2, 4, and 6 for triple channel. A Guy

I was wondering about that. I also noticed in the OPs, system specs it list 12GB of RAM

Did a RAM stick go bad recently?

 

[smarteyeball]

Disappearing RAM is a common occurrence with the x58 chipset.

There are many potential causes:

► Faulty RAM Module/s
► Faulty Dimm Slots
► Improper CPU seating (pad/pin contact between CPU and Heatsink)
► Incorrect RAM timings/voltages
► Aggressive QPI Link Data Rate
► Electrical Shorting
► etc

Did one of your modules die?

What speed are they set to run at ie 1066mhz, 1333mhz, 1600mhz etc and have you manually entered the voltages and timings?


Things to try:

AI Tweaker:

DRAM Frequency - set this as close to 1600mhz as possible - keep it under 1600 for now if you have to.

Manually enter the Ram timings in the ' DRAM Timing Control 1st Information' - 9-9-9-24


It may also be necessary to change your CR (command rate) from 1N to 2N (2nd information - DRAM Timing Mode)


Set your DRAM voltage to 1.64 and since you have virtually all DIMM slots filled, up your QPI/DRAM voltage to 1.300 - 1.35



(To eliminate the QPI Link, set the QPI Link Data rate to the slowest speed available)



If after manually entering the timings and voltages it still fails to detect all modules at POST, remove the 'now 4GB' kit and test with only the one remaining 'complete' 6GB kit.


If it still fails to detect all 6GB - then you'll have to start the laborious process of trying each module individually.

If any individual sticks fail to POST - then you've got a dud module/s.

If the machine POSTs with each stick, take a 'known good' stick and POST the machine from the other bootable slots

(*NB not all slots will allow the Machine to POST. ie A1, B1, C1 will allow the machine to POST with a single module - A2,B2,C2 will not)

In this instance leave a 'known good stick' in a bootable slot and add another 'known good stick' and see if it detects 4GB in all of the slots.

If it doesn't, you may have a dud DIMM/Mobo.


This is a 'quick' elimination method as you don't have to sit through memtest86+ runs - yet.



.

 

[tktk88]

Thanks for your replies.........

This is a new machine and as questioned per my system specs I do have 12 Gig of Ram you can see all my posts in the Crashes and BSOD section upon arrival I was having Multiple BSOD which turned out to be a bad ram stick so it is removed waiting for replacement Ram in the mean time I am running the machine with the 10 gig that I have............computer is running fine just experiencing the wierd not seeing all the Ram issue, this issue seems to be isolated to a cold boot after a reboot or 2 and the computer being warm all the ram is recognized by the bios and windows, at any given time CPU-Z always sees the ram despite windows and my bios does or not..............Do you guys suggest just running the 6 gig in the 2, 4, 6 slot until the replacement Ram arrives or is 10 gig ok?? I will try the suggested remedies this evening as always i appreciate your help!

Per the Asus motherboard manual I have placed the ram in sequence it shows for the amount I have currently............

 

[tktk88]

The motherboard bios has already been flashed to the latest which for this Asus board is 0808

 

[tktk88]

Today my cold boot would only detect 2 gig of the 10 gig I applied the following suggestions:

1 I set the DRAM Frequency to 1333 for now

2 All Ram timings have been entered manually 9-9-9-24

3 Changed CR to 2N from 1N

4 Set DRam Voltage to 1.64 and QPI/DRAM voltage to 1.35

5 Set QPI Link to lowest speed

On reboot the machine is now detecting 10 gig of Ram in Bios and Windows..........this is the typical pattern after it is warm typically all ram is detected, I guess tomorrows cold boot will be the test...........

Any other things I should be looking at or changing at this point??

As always thanks for your help............

 

[smarteyeball]

While it's detecting the full 10GB at 'loose settings', I'd do a memtest86+ run. 8 full passes should be sufficient enough to uncover any apparent dud dimms/modules.

This should help differentiate between potential hardware faults vs settings issues.


(Let it run overnight as 8 passes will take around 8hrs - although if you get errors quickly, halt the test. Then you have the 'pleasure' of removing/swapping modules to test dimm/moudules slots.)



If there are zero memtest errors and it still consistently maintains the full 10GB detection on a cold boot with the current settings, you can slowly start changing some of the settings back. However leave the DRAM volts at 1.64v as you'll need this to run @ 1600mhz

►Start by upping the mhz back to 1600mhz. *NB the UCLK frequency should be double what the DRAM frequency is. AUTO should do that automatically, but it's probably worthwhile setting it manually.

► Set the QPI link up one notch.

► *Optional* Set the CR back to 1N (although there really isn't that much noticeable difference in terms of speed between 1N and 2N, and with a full 10/12GB 2N is better for stability )

►After that you can start lowering the QPI/DRAM volts until it either fails to detect the full amount, or becomes unstable. Although 1.35v is perfectly fine.

If after adjusting any of these settings results in detection loss, go back to a lower setting.

(Depending on the board/RAM - 6GB kits are 'guaranteed' to run at the listed specs ie 1600mhz @ 9-9-9-24 1N, but are 'not guaranteed' to run at those timings in a 12GB config, hence the need to loosen some timings / increasing voltages, particularly the QPI/DRAM voltage etc)



However, if there are zero memtest errors with the full 10GB detected (basically eliminating faulty hardware), but it still occasionally fails to detect the full amount - I'd start looking at other culprits like the CPU cooler tension which affects the CPU pads / socket pin contact etc.

 

[tktk88]

Todays update back to the same thing with out having made any changes to the suggested setting from my previous post at todays cold boot the system in bios and windows would only detect 2 of the 10 gig of ram.....reading some of your others suggestions ie. 6 gig guaranteed to run at 1600 in these settings and not 12 would it smarter to just go to 6 gig and not have these issues not really sure about the answer here at this point I use the gomputer for gaming interent and work from home at time.......i do alot of digital photography but nothing porofessional and I am not using photshop 64 bit to take advantage of the ram.....not really wanting to surrender but not sure if the efforts are worth it either at this point...in looking at comparisions in gaming performance not much is gained from 12 gig over 6 from what I can tell......................

 

[Petey7]

I don't think you would see much performance lost at all by going to 6 gigs. I doubt you ever use that much even with gaming and Photoshop.

 

[tktk88]

By going to the 6 gig should I expect more stability?? From what I can gather from reading these issues are prone to exisit when exceeding the 6 gig amount and prone to issue with the x58 chipset..................if i go the 6 gig route should I set everything in the bios for the memory back to auto or keep the current settings or something different.............

 

[Petey7]

Thats something smarteyeball would have to answer. All I can say is, you have to be running several extremely complex programs to need more than 6GB.

 

[tktk88]

Ok thanks for that hopefully he will read this and give me his input..................

 

[smarteyeball]

By going to the 6 gig should I expect more stability?? From what I can gather from reading these issues are prone to exisit when exceeding the 6 gig amount and prone to issue with the x58 chipset..................if i go the 6 gig route should I set everything in the bios for the memory back to auto or keep the current settings or something different.............

Sorry for the delay.

Unfortunately, even 6Gb configs suffer this problem. However, more than 6Gb can exacerbate the issue.

For example, I had a similar issue where my BIOS would only detect 4Gb out of 6GB when I initially built this rig.

In my case it was a combination of CPU cooler tension and timings. I now get a 100% detection, but can replicate it with too aggressive timings etc.

***

Firstly, did you manage to do a memtest run with the full 10GB?

Or even any memtest runs?

Without eliminating dud modules/dimms, we could be fighting an uphill battle.


There's also no need to surrender the full 12GB completely, but for the time being I'd put it on the back burner and concentrate on getting a consistent 6GB @ 1600mhz detection.

If that can be achieved, you can add the extra 6GB later.

So the next step would be to run the 6GB kit only in tri-channel @ 1333mhz, same voltages (1.64v/1.35v) with the same slow QPI link etc.

Also, with the DRAM Frequency, 1st information, change the tRFC to 88 (to further loosen the timing)

If it's still failing to detect on a cold boot, you're going have to start looking at other things like adjusting the tension (screws) on your cooler block. Or even removing the cooler and re-seating the CPU etc.

If the screws are too tight, too loose or unevenly tensioned, this can cause uneven CPU pad/pin contact which is one known issue.

I'm not sure whether that cooler has a plastic/metal backplate. Uninsulated Metal on metal has also been known to cause 'shorts' which can affect detection.




Alternatively, is the machine still under warranty?

 

[tktk88]

Yes this is a brand new computer from CyberPower...........not 30 days old yet at deleivery i was having all kinds of BSOD issues which the guys helped me figure out that I had a bad ram stick thus why I am running the 10 gig instead of 12 for the time being.................new ram arrived but was not matching not even the same brand..........I called CyberPower at them them know I was not at all happy with that..........after some discussion with them on Friday they are now sending a new motherboard and 12 gig of ram.............so I am waiting now for that to arrive.......I think at this point with all the issues and this being new its time to replace the motherboard and the ram as well.......going through all this has made me think I should have just gone with 6 gig instead of the 12

 

[tktk88]

Also, yes i ran memtest overnight with the 10 gig in the machine it tests fine............

 

[smarteyeball]

Curious that it did not detect any errors - that tends to suggest that the board and RAM are probably okay :huh:

But hopefully whatever 'gremlin' that is the cause gets wiped out during the mobo/RAM swap.


I assume that since CyberPower are sending you the mobo/RAM - you are doing the swap yourself?

(I don't know if you've built a rig yourself, so please ignore if you already know all the following, this is for just in case you haven't)

Before rebuilding:

► Inspect the new mobos socket for any bent pins prior to CPU insulation

► Inspect the pads of the CPU for uneven contact spots (too see if the current cooler placement was uneven)

► *Optional* Assemble the board outside the case on a non-conductive surface ie wooden or glass table, Motherboard box etc - not the anti-static bag that the mobo comes in, or any cloth surfaces (eliminates potential shorting issues)

► Ensure that the coolers backplate is insulated (if it has rubber/paper washers etc - use them)

► Clean the existing thermal compound off the cooler block and CPU and re-apply fresh TIM (do not re-use the existing applied stuff)

► Use a 1/3/2/4 tightening method when attaching the cooler. ie Lightly tighten one screw at a time in a diagonal pattern, never over tightening one particular screw. (Top left, bottom right, Top right, Bottom left) When all four screws start to 'bite' - don't over-tighten them.

► When putting the mobo in the case, ensure that the standoffs are used - they can be plastic, brass, or even part of the board (don't screw the mobo straight into the case)

► If CyberPower are doing the rebuild - tell them to DO IT RIGHT


And do not regret going 12GB - there are more people successfully running 12GB than there are with issues. Currently, you've been unlucky - but don't let it dishearten you - you should nail it in the end


If you need any further help when the new mobo/RAM arrives, feel free to PM me.

Best of luck mate.

 

[tktk88]

On cold boot memtest does show errors not once it is warm i think I failed to mention that, however, after detecting all 10 gig and running memtest with the computer warm it does not indicate any errors even running it for 6 to 7 passes that part seems somewhat strange to me, do you thunk that still indicates a memeory issue or voltage and timing due to it being isolated to cold boot only?

Also thanks for the advice on the motherboard and ram replacement

 

[tktk88]

If I ran CPU-Z and posted a copy to you would that help you to look at what is going on?