DRAM Driving Strength Profile... wottizzit??

[Chucko]

I've been running a (long and tedious) series of Memtest86+ runs to see if I can get my current rig to run 8GB of RAM at 1066MHz. Tonight I ran into a situation I hadn't seen before. Usually if I'm pushing the limit of the system in one dimension or another, it will get a bunch of errors, I'll make an adjustment, and it'll be solid again. Not this time. It failed hard before I figured out what to tweak.

System specs: Core 2 Quad Q6600, EP45-DS3R mobo w/ F10 BIOS, 4 x 2GB OCZ Reaper HPC 1066MHz DDR2 RAM. I have a good Linpack stable overclock of 3.0GHz (FSB 333MHz), 800MHz RAM at stock voltages (or lower).

So starting from 266MHZ FSB, 266MHz memory latch, and 3.00A divider, I've been gradually increasing the FSB (and RAM clock) and changing whatever needs to be tweaked to get Memtest86+ stable again. I had things running nicely at 315MHZ FSB and decided to bump it up to 320.

To my surprise, it failed to post. I tried again, same result. Back it down to 315, everything's happy, runs for 2 hours without error. Back up to 320, nada. Bump up the MCH Core voltage to 1.05V (still below stock), same thing.

What fixed it was bumping the DRAM Driving Strength Profile on both channels from 800MHz to 1066MHz. Note that the MCH core is still at 0.95V as I write!

So this is a roundabout way of asking:

Exactly what does the Driving Strength Profile control, anyway?

[linderman]

I would love to know that myself !

[grishkafool]

Suprised that you haven't recieved an answer yet.... Lsdmeasap, I too am wondering what the Driving Strength Profile is... I found this reference, but I don't really understand it in relationship to my EP45-UD3P.

[Psycho101]

Taken from an Anandtech article, slightly altered to keep it reader friendly, and not too bogged down in technical stuff:

"Sometimes called driving strength. This feature allows you to control the memory data bus' signal strength. Increasing the drive strength of the memory bus can increase stability during overclocking. DRAM drive strength refers to the signal strength of the memory data line. A higher number means a stronger signal and is generally recommended for an overclocked module to improve stability."

Put more simply, it controls the strength of signal the board uses to communicate with the RAM and also the strength of how the RAM communicates with the board and installed components.

Think of it like having a conversation. You speak at different volumes depending on background conditions. In a quiet room (akin to a non overclocked PC) you can talk quite softly. In a room with alot of background noise, like a nightclub (akin to a highly overclocked system with all components on or at the edge of what they can do) you might need to shout.

Although increasing driving strength can help, it all depends on the type of memory modules used. As a general guide, setting the driving strength to either the exact memory speed (800 when RAM is running 800-DDR, 1066 when at 1066-DDR etc) or the higher of the two settings when overclocking (eg if at 920-DDR, 1066 drive strength may help).

Failing to post on an overclock for me always gets me thinking it might be the RAM, so I review my settings and see what might need changing. Driving Strength can be important, and you never can tell if your modules are sensative to this setting or not. Also, the number 1 cause of no-post and reboot loops is Static tRead (damn I HATE tRead, just the mention of it gives me a tight chest )

[Lsdmeasap]

tRD is often set to low by Auto as well which causes many issues for first time builders without knowledge or access to different modules

Source link for above article please? Thanks

I had always thought this >>
Driving Strength Profiles just manually (Or really Automatically) sets a pre-defined set of timings for all the rest of the settings on that page, per the Driving Strength profile you choose.

But reading your above quote I think that may all just be wrong. I can say the info you did link does not apply to this setting as there is no 1-8 choice, so I am leaning toward the Anandtech article being correct of course.

[Psycho101]

The quote from my post is from several sources all pointing to an Anandtech article written by "Adrian Wong", but having searched Anand I can't find it. Not even a Google cached page. Maybe you'll have better luck.

The numeric settings discussed are options on DFI boards. The Gigabyte BIOS most probably implements drive strength in the same way, but describes it as speed settings such as 800, 1066 etc. Which implementation is correct, I don't know. However with both drive strengths available to adjust, and more settings, I would conclude that DFI has more options for fine tuning. Gigabyte seem to have chosen to hide settings behind memory speed designations, who knows what each setting changes.

I would definately have thought though, just from the name that it's signal strength that's altered, not latencies and timings. To address memory at higher and higher speeds, I'm sure stronger switching strengths must be required, as the switching itself will be happening for less time.

[Lsdmeasap]

Ya, seems he is gone? And the site he moved to as well after his initial site
Adrian's Rojak Pot - Performance is our middle name!

This is all I could find from that info, aside from the new Version 10+ Book you can now Purchase : ) This is from Revision 6.5
Chipset Features Setup

Version 7.0 rev 2 is here in cache, but I cannot find 10 anywhere but for purchase
The Definitive BIOS Optimization Guide

Buy it here (Or of course find torrent)
The Definitive BIOS Optimization Guide - Google Product Search

Some things are mentioned about Drive strength there, but not like you quoted exactly so maybe what you found is info from the new $$ Book you must buy?

Wait, found 10.6 in google cache, seems to be at techarp but the site ie messed up?
Tech ARP - The Definitive BIOS Optimization Guide

It worked for a few but not now, but nothing mentioned anything like Gigabytes options or your quotes about settings, similar but not the same so I guess you may have found something he wrote that is not part of the book?

MD Driving Strength

Common Options : Hi, Lo / High, Low

Quick Review

This BIOS feature offers simplified control of the memory data bus' driving strength.

The default value is Lo or Low. With heavy DRAM loads, you might want to set this feature to Hi or High.

Due to the nature of this BIOS feature, it is possible to use it as an aid in overclocking the memory bus. Your memory module may not overclock as well as you want it to. By raising the driving strength of the memory bus, it is possible to improve its stability at overclocked speeds.

However, this is not a surefire way of overclocking the memory bus. All you may get at the end of the day is increased EMI and power consumption.

Please note too that increasing the memory bus drive strength will not improve the performance of your memory subsystem.

Therefore, it is recommended that you leave the MD Driving Strength at its default Lo or Low setting. Set it to Hi or High only if you have a heavy DRAM load or if you are trying to stabilize an overclocked memory module.

Tech ARP - MD Driving Strength

Memory DQ Drive Strength

Common Options : Not Reduced, Reduced 15%, Reduced 30%, Reduced 50%

Quick Review

This BIOS feature allows you to reduce the drive strength for the memory DQ (data) pins. It does not allow you to increase the drive strength because it has already been set to use the maximum drive strength by default.

When set to Not Reduced, the DQ drive strength will remain at full strength.

When set to Reduced 15%, the DQ drive strength will be reduced by approximately 15%.

When set to Reduced 30%, the DQ drive strength will be reduced by approximately 30%.

When set to Reduced 50%, the DQ drive strength will be reduced by approximately 50%.

Generally, you should keep the DQ driver at full strength if you have multiple memory modules. The greater the DRAM load, the more DQ drive strength you need. But no matter how many modules you use, AMD recommends that you set this BIOS feature to Not Reduced if you are using a CG or D revision Athlon 64 or Opteron processor.

However, if you are only using a single memory module, you can reduce the DQ drive strength to improve signal quality and possibly achieve higher memory clock speeds. If you hit a snag in overclocking your memory modules, you can also try reducing the DQ drive strength to achieve higher clock speeds, even if you are using multiple memory modules.

AMD recommends that you reduce the DQ drive strength for Revision E Athlon 64 and Opteron processors. For example, the DQ drive strength should be reduced by 50% if you are using a Revision E Athlon 64 or Opteron processor with memory modules based on the Samsung 512Mbits TCCD SDRAM chip.

Tech ARP - Memory DQ Drive Strength

Close, but of course Gigabytes exact settings would not be covered as the likely create their own combination of things.



I asked you for your source link because I have always found and suggested 1066 worked better too even at 800Mhz, so I just wanted to read what this similar mind as mine found and if he had anything better to share. Think I will have to grab this torrent and see what all he has to offer.

Actually just found it in mediafire on google, file name is as follows should be on page one if you want it >>
bios Breaking_Through_the_BIOS_Barrier.pdf

I assume that route is likely not the paid version though, but who knows? No revision mentioned in the PDF so dont know how old it is either but looks to be the 2004 version as that is all I find at my normal (Free) places I frequent

[Psycho101]

Thanks for finding the link.

Strange thing is... you know how threads you reply to get automatically added to "watched" and you get an email when there is a reply? Well your post set off Live Mail's Phishing filter

I guess one of those links, maybe the one that links to the paid for article is flagged as suspicious. I guess it's wise as you say, not to pay for the pdf, as you may not get what you pay for.

It seems strange that there isn't much info on driving strength. All the info there is, seems about as vague as both Gigabytes and MSI's (Missing Something Inside) explanations. Both just tell you the values you can set, nothing more. I suppose they just don't want to "blind us with science".

Tracing the exact quote back to a link on a couple of forums labeled as a clickable "Adrian Wong" lead to the Rojak Pot site, but it seems to be all geared up to sell ebooks.

Rojak Pot.... what a strange name. Sounds like some kind of cannabis curry

[Lsdmeasap]

Thats funny about your email, I use windows Live mail desktop myself! I have all notifications here disabled myself, you can imagine why I am sure.

See if this sets it off (Sprint's new Now Network commercial says it should hit the spam filter )>>>

Miracle Banana Diet

Just incase it needs a URL linked one
Miracle Banana Diet

Or
miraclebananadiet.com - Under construction


Rojak Pot is explained in the PDF, I could not copy it here so you will just have to grab it to see. It is a big long story at the end of the PDF. I didn't read it, very busy and it seemed longer than my current attention span

[Psycho101]

Nope, that post went straight through. Mail filter seems a little unreliable too. I have had obvious SPAM get through on occasions. I get loads of MSN Photo Album things from random womens names, things like "Hi sex. Good see you again. Here my naked holiday photos. Cum see me long time." and it lets them straight through. I never follow the links though..... honest.