Jump to content

Welcome to Geeks to Go - Register now for FREE

Geeks To Go is a helpful hub, where thousands of volunteer geeks quickly serve friendly answers and support. Check out the forums and get free advice from the experts. Register now to gain access to all of our features, it's FREE and only takes one minute. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more.

Create Account How it Works
Photo

DQS Training


  • Please log in to reply

#1
Grongle

Grongle

    Member

  • Member
  • PipPip
  • 15 posts
Hi,
I'm hoping to find someone who can tell me quite a bit about specific details regarding the pre-POST tests, and particularly when DQS Training is performed.

I built my computer last November, but I've never managed to get the POST time below 20 seconds, and doing even that well has been very rare. It has averaged 4˝ minutes to POST (i.e. BEFORE getting as far as the Bootscreen).

The system is a difficult one, using an AMD mobo that has a highly theoretical ability to handle 16GB of RAM. This system carries 8GB of DDR2-1000 which the AMD Athlon 5600+ sees as DDR2-800, and between the mobo, the RAM, and that CPU, the system is pretty tricky. I have 4 hard drives and a Corsair HX520W PSU. The PSU tests OK with my little TT Dr Power, plus it calculates as correct for the load by Corsair and NCIX. I recently RMAd the motherboard back to L.A., but received it back as "repaired" without a word of comment from Gigabyte.

A G.Skill RAM tech has told me the RAM must be functioning properly as my ONLY problem is the slow POST. Windows works fine. I'm running no changes to my CPU voltage, but I use
NB +0.3V = 1.400V (recommended by the G.Skill tech as required by the 8GB of RAM)
DDR2 +0.200V = 2.000V
Trfc 4 @ 327.5ns

I feel the RAM needs all the help it can get, and I like DQS to perform. I know little about it. I notice that DQS training seems to take a long time whenever I have made changes in
  • CMOS battery removed and replaced
  • BIOS flashed
  • voltages modified; Trfcs changed
  • hard drives removed or replaced

But I'd really like to know what is happening when "nothing" seems to be happening: when the computer is powered on from cold, and the fans spin and there are lights, but you would think the case was otherwise dead—UNTIL there is a POST beep, which can be quite a few minutes in happening. After a battery out-and-in + BIOS flash + setting all the BIOS voltages as above, I have had up to 48 MINUTES of "nothing happening" before that POST beep. The usual after a BIOS flash is 10 or 15 minutes.

Later, on subsequent cold startups, the POST time drops to its average of 4˝ minutes. Recently I have sometimes had it down to just below 2 minutes. I haven't seen 20 seconds since August—and I lost that when I flashed the BIOS from F8 to F10. Retrying F8, I never succeeded in getting it back, although I have meticulous records of all my settings.

If I knew more about the actual process of DQS training, or just where the POST is delaying, it would be very very helpful. Thanks.

Edited by Grongle, 09 October 2009 - 11:19 AM.

  • 0

Advertisements


#2
Neil Jones

Neil Jones

    Member 5k

  • Member
  • PipPipPipPipPipPipPipPip
  • 8,476 posts
I've never seen any board in the last fifteen years or so take over 20 seconds to complete POST.
Modern day boards typically complete POST in about three to five seconds.
I have to ask, why have you put up with this for nearly a year when it's not the norm?

What is your motherboard make and model please?
With regards to DGS Training, here's a page on it: http://www.techarp.c...ang=0&bogno=433 But I think you may be barking up the wrong tree.

Edited by Neil Jones, 09 October 2009 - 04:04 PM.

  • 0

#3
Grongle

Grongle

    Member

  • Topic Starter
  • Member
  • PipPip
  • 15 posts
Thanks, Neil. I know about the TechArp overview, but I'm looking for some quite specific details.

(Hm! System details should have been there. Good thing you mentioned it. GA-MA78GMS2H 1.1; just got it back from RMA (a futile exercise, as we all knew). The CPU is light for the system, and the RAM is demanding. The PSU tests OK and is rated OK by both Corsair and NCIX—who both stood to make more money if they had indicated a more powerful PSU. (NCIX said No, even when asked.)

@Neil, A few barks I've made along the way:

"It’s not like a software or BIOS or POST problem. It is like a hardware problem. You feel like there is nothing happening at all except for case fans and lights being on. Then suddenly the beep sounds and you wonder if it did take only two or three seconds after some lengthy time of the computer just sitting there and doing nothing, not even trying, at all."

and

"I don’t expect this process to involve gradually whittling down half-minutes and full minutes. No; I expect there to be a single factor which is exerting its influence over the entire boot process. I think it would be a mistake to reduce the timing piecemeal. What needs to be discovered is one definitive cause for such a long time from initial power ON until the BIOS display."

What I'm trying to do now is to determine exactly what is taking place when "nothing at all seems to be taking place" while the fans are on during the minutes before POST. The longest time before POST—and it was a perfect POST and a subsequent perfect Windows session, as it ALWAYS is—was some 48 minutes! Yes, 48 minutes! That was with DQS timing performing, and right after a CMOS battery removal/replacement and BIOS flash, with changed Trfc settings and NB and DDR2 voltages. If DQS were working all that time, it would make sense that I had given DQS the maximum work to do. I've stated above which situations seem to take DQS time.

On the other hand, I have also felt I was carefully timing nothing at all. This argument suggests DQS is NOT engaged yet and that the system is just sitting there for up to 48 minutes, and then, suddenly, the POST takes place in the very few seconds that POSTS always take place.

This problem has been discussed on various forums. Nobody has cracked it yet. And, yes, I have stripped down the system and tried it with ZERO drives and just one stick of RAM and tried different RAM sticks. I've tried SATA and PATA settings, too. The RAM tech was sure that faulty RAM would not allow smooth Windows operations, keeping in mind that the long time to POST is the ONLY problem on this system.

Restarts from Windows always work perfectly. This problem is exclusive to cold starts. There seems to be a strange predilection for VERY cold starts, as the data is definitely most reliable after the system has been off at least 3 hours. I know; you'd think that would make no difference.

I need to know what is happening before the POST.

The person who finally solves this will receive a free ice cream cone shipped to his door.

Oh, 2 big constraints: I don't have hardware to simply swap back and forth: at least $100 each or more for the usual suspects, such as mobo, CPU, PSU, RAM—nope; I don't have the available hardware just sitting around, nor the kind of money to gamble. (Note that the motherboard is just back from RMA. The PSU tests OK with TT Dr. Power.)

Again, thanks.
Again, I'm hoping to get some very detailed, specific information about DQS. Even ruling it OUT would be very helpful.

Edited by Grongle, 10 October 2009 - 07:45 AM.

  • 0






Similar Topics

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

As Featured On:

Microsoft Yahoo BBC MSN PC Magazine Washington Post HP