Jump to content
Sign in to follow this  
Guest Darien_merged

How can we tell the auto DS and Slew Settings

Recommended Posts

Guest Darien

like the title says... how can we tell the Drive Strength and Drive Slew Rate that the bios uses when set to auto.

 

we can use wpcredit but how?

Share this post


Link to post
Share on other sites
Guest Darien

We can tell the Drive Strength and Drive Slew Rate that the bios uses when set to auto by doing a register dump with cpu-z or by using wpcredit

 

using wpcredit or cpu-z you need to look under the device/section labeled 01EC

 

with wpcredit hit choose device and choose the one that shows 01EC

or

with cpu-z reg dump go to this section

RAM Memory Controller

bus 0 (0x00), device 0 (0x00), function 4 (0x04)

Common header

Vendor ID 0x10DE

Model ID 0x01EC :)

Revision ID 0xC1

PI 0x00

SubClass 0x00

BaseClass 0x05

Cache Line 0x00

Latency 0x00

Header 0x80

PCI header

Subvendor ID 0x10DE

Subsystem ID 0x0C17

Int. Line 0xFF

Int. Pin 0x00

Dump

0 1 2 3 4 5 6 7 8 9 A B C D E F

00 DE 10 EC 01 00 00 20 00 C1 00 00 05 00 00 80 00

10 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

20 00 00 00 00 00 00 00 00 00 00 00 00 DE 10 17 0C

30 00 00 00 00 00 00 00 00 00 00 00 00 FF 00 00 00

40 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

50 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

60 DD 22 DD 22 BB 44 BB 44 AA 11 AA 11 99 22 99 22

70 44 44 44 44 F0 00 00 00 7E 3E 80 0F BB 44 AA 11

80 44 44 00 00 00 00 00 00 00 00 00 00 00 00 00 00

90 35 35 33 33 31 31 31 31 13 13 12 12 11 11 11 11

A0 00 00 00 00 56 C6 00 00 00 00 00 00 5F 05 A0 00

B0 BC B7 98 C8 8C 85 97 56 AC C8 97 B8 8C 86 96 55

C0 78 06 56 07 41 03 00 00 4A B8 DA 00 00 00 00 00

D0 FF 20 FF 20 FF 20 FF 7F 00 00 00 00 00 00 00 00

E0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

F0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

The important part of this the last part where it says Dump, and then a big matrix of Hex code... lines 60 and 70 are the ones you need

60   DD 22 DD 22 [BB 44] [BB 44] AA 11 AA 11 99 22 99 22 

70   44 44 44 44 F0 00 00 00 7E 3E 80 0F [BB 44] AA 11

 

the portion I put in brackets corresponds to the 6 settings (3 drive strength and 3 drive slew)

 

I'm going to double check the translation and post the results here in a second:

 

OK they are read backwards.. so mine are: 4-11 4-11 4-11

 

so the last brachet is bb 44, reverse that 44 = 4 and BB = 11, so the first slots strength-slew is set to 4-11

 

the middle bracket is the same as the last bracket so the second on is the same = 4-11

 

the last slot is the first bracket bb 44, reverse that 44 = 4 and BB = 11, so the first slots strength-slew is set to 4-11 ( just like the other two)

 

so mine are: 4-11 4-11 4-11

 

thx to NeoArthur and medianOcer for their posts... please correct me if I got this wrong... this is a HUGE help to know where to start from... thx

 

I'm running 2 x 512MB OCZ 4000vx Gold RAM in Dual Channel..

Share this post


Link to post
Share on other sites

Np, sorry you ended up answering your own thread/post, but it looks like you

found a thread where we were dealing with that.

 

Best to you.

 

PS: I was just going to address your query, but you got it already. Nice job. :nod:

Share this post


Link to post
Share on other sites

I'll have a new batch of pcr's ready in a few days at most ;).

I still have to look throuh one more PCI-SIG data sheet to check for header type 01h code 1st.

 

Then I'll release them :).

Gonna have stuff like Secondary Latency Timer, min grant and max latency, latency(the last 3 I allready had but it's more accurate now).

Also has those drive and slews :).

 

Tons of goodies, not that much for like new timings or anything but the 1st 00-3F in offsets, those are getting a major overhual.

Share this post


Link to post
Share on other sites

:), I based the the info off some official manuals.

 

The devices that have 00h for offset 0Eh, I used this manual for the 1st 30 offsets:

PCI Local Bus Specification Revision 2.2.

The ones with 01h for that offset I'm gonna use this manual, I just got this one:

PCI-to-PCI Bridge Architecture Specification Revision 1.2.

(I can't belive you have to be a PCI-SIG member to dl these : at least I found them elsewhere heh)

 

Makes the pcr's much more accurate :), and it serves as a basis for new pcr's in the future.

As others could dl the things, and use the begining of them for the 2 types of devices I made pcr's for(Which would be a good idea for others to do that's for sure, would make alot of those pcr's alot better).

 

The devices that have 80h for that one offset I dn though, I'm not sure what that's about.

However I do know the standered 00h-0fh are allways identical so I'll go about it that way.

 

This is all so we don't have extra latency timings and io addresses, etc etc, in the wrong devices.

 

That and I can truthfully say what latency is actually what on what device.

Instead of not knowing, ie that 1Bh offset on some devices which some say was a latency of some sort, it actually was but only on certain types of devices.

The min grant and max latency settings are the same way, only on certain types of devices.

 

 

Anyways since I got the 2 manuals now I can finish it up within another day or so.

I gotta go over my buddies today sometime, his nephew died :.

So I should probably get some sleep.

I worked all night of all things, on the tcc5 drive and slew rates and got them down perfectly for the most part...

 

Since this thread is sorta about drives and slews I'll share my info here, I'm to lazy lol...

 

 

Here's my results from my beta bios, which is'nt even close to being down, I mean I barely touched it so far(old romsip, very old ver of my mods, but diff bios and etc, bpl, oem regs, mp mod, blah).

 

 

;;;;;; Drive Strengths & Slew Rates :

 

TCC5 : (Slot 1 : Prefers Slack Timings, Slot 3 : Any Timings)

1-1 (Prefers Tight Timings?)

10-1 (Prefers Slack Timings?)

 

ADATA : 3-11?

Drive Strengths : 1, 2, 3, 4, 10, 11

Slew Rates : ?

 

;;;;;;

 

;;; 278 :

 

CPU Multiplier : 5X (Memtest Stable)

VCore : 1.45v (Memtest Stable)

CPU Speed : 1390Mhz

 

Memory Speed : 556Mhz

 

Slot 3 : TCC5 (Memtest Stable In Slot 1 As Well)

 

North Bridge Voltage : 1.8v (Memtest Stable)[TCC5]{SC}

VMEM : 2.8v (Memtest Stable)[TCC5]{SC}

 

;; Timings :

 

4-4-2-3 (Memtest Stable)[TCC5]{SC}

 

CAS: 3 (Memtest Stable)[TCC5]{SC}

CMD: 2 (Memtest Stable)[TCC5]{SC}

 

9-14 (Memtest Stable)[TCC5]{SC}

 

0-0-2-3-3-6-3 (Memtest Stable)[TCC5]{SC}

 

X-X-X-X-1-1 (Memtest Stable)[TCC5]

 

;;;

 

;;; 275 :

 

CPU Multiplier : 8X (Memtest Stable)(S&M Stable)

VCore : 1.475v (Memtest Stable)(S&M Stable)

CPU Speed : 2200Mhz

 

Memory Speed : 550Mhz

 

Slot 3 : TCC5

 

North Bridge Voltage : 1.8v (Memtest Stable)(S&M Stable)[TCC5]{SC}

VMEM : 2.8v (Memtest Stable)(S&M Stable)[TCC5]{SC}

 

;; Timings :

 

4-4-2-3 (Memtest Stable)(S&M Stable)[TCC5]{SC}

 

CAS : 3 (Memtest Stable)(S&M Stable)[TCC5]{SC}

CMD : 2 (Memtest Stable)(S&M Stable)[TCC5]{SC}

 

9-14 (Memtest Stable)(S&M Stable)[TCC5]{SC}

 

0-0-2-3-3-6-3 (Memtest Stable)(S&M Stable)[TCC5]{SC}

 

X-X-X-X-1-1 (Memtest Stable)(S&M Stable)[TCC5]

 

;;;

 

I can say for sure that the drives and slews are not inpacted by other timings.

However I do reccomend you try using all sorts of combinations of speeds and timings to "eliminate" the bad values ;).

That's actually how you doit, you must find speeds where some values get only a few errors maybe, and the others get over a 100 or so errors in memtest.

Then you keep the good ones, and try again at a diffrent speed or main timing setup.

(You cannot say dump the ones that fail but only with a few errors and only keep the ones that pass a few times then fail, you'll figuer it all out eventually, you must be keep all possible good values)

You keep doing this over and over until you eliminate all of the bad ones.

It's very time consuming, it took me all night to get through one stick.

 

I can say another thing too.

Drive strengths have no inpact on validity of drive slews and vice versa.

Meaning you can find a somewhat decent drive strength, and verfiy correctly, bad drive slews.

But only to a certain extent, because you may be using a bad drive or slew in the 1st place, in which case you may have to swap to another of good value to verify.

 

In the end, I had to verify the last of them at 279mhz, which was not fun, figuering out what ones only got a 1000+ errors, and ones that got inf errors.

I ran the whole set of good values 4 times to make sure, becaue this was where I had to verify them at the end, there was no other speed or timing area where I could at this point.

The only ones that lasted where 10 and 1 for drive strengths, all the drive slews failed eventually as well except for 1.

 

In the end, after testing all of that I tested the slots with the 2 possible combinations.

That's how I found that slot 1 is bad for tight timings, at least for my samsung.

Ie 4-4-2-3 0-0-2-3-3-6-3 @ 279.

However slot 3 did well, only getting 400 errors or something.

Even better then what slot 1 did with slack timings, like 8-4-4-4 0-0-6-6-3-6-6.

 

When I used slack timings, 10 did better sometimes, rarely.

Most of the time 1 did better.

 

Slot 3 and 1 are equal when slack timings are in place, however slot 1 beats out slot 3 with tight timings, example:

Slot 1 slack vs. slot 3 tight = slot 3 wins anyways.

When both are slack, they are about equal.

This is only with that one stick though.

 

I have to go through most of this again with the adata :.

I got some of it down though as you can see above.

 

Slew rate of 10 is better then 11 by a long shot though I can say, for my tcc5 anyways.

I may set that one as default for my bios...

 

Of course I tested at multible timings, 5-2-2-2, or cas2.5, or cmd 1.

Matter of fact, last time I messed with cmd 1, I got it 277 memtest stable of all things.

I have not tested it since I finished though.

I did re-check cas2.5 though, passed 240, however I noticed my bios is not up to that, it's the only spot where I get warm boot issues.

 

 

It's not hard though, just time consuming, kinda like reading one of my posts lol.

Just verify a bad timing, take it off your list, and keep doing that.

Eventually you'll have your ideal setup for those 2 things, drives and slews.

In the mean time, between testing, just use what you think is best out of those timings that are still on the list.

At least you''ll know what not to use ;).

I'm a bit surprised I figuered it out, but heck, I needed to.

 

Got me that 275 stable lol, and of all things, 278 memtest stable, did'nt check windows.

I have yet to check higher then 8x for 275 though.

 

Edit:

Oh, using a higher drive strength then slew is valid as well, I allways wondered about this one.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×