Oh my God!

Discussion relating to Socket 7 hardware.
Jim
K6'er Elite
Posts: 1745
Joined: Wed Jan 21, 2004 7:10 pm
Location: Toronto

Post by Jim »

CPU Cool works w/ WINXP, don't know for sure about 2K though.
Superpuppy 3
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
sylvan2626
Senior K6'er
Posts: 139
Joined: Tue Apr 05, 2005 3:25 pm
Location: Florida, USA

Post by sylvan2626 »

And the saga continues...

Got CPUCool and CPUFSB installed and in use.
Installed Coolbits.

To recap:
FSB running at 124MHz (most of the time - sometimes I switch to 112MHz and run with the board's L2 memory cache enabled). Discovered my SDRam modules are not PC133 - that may be what is keeping me from hitting 133MHz FSB. Gotta do something about that!

112MHz seems to be the upper limit (or at least of the choices that I have from CPUFSB) for running with the board's L2 cache enabled.

Multiplier set at 4.5.

Core voltage set at 2.1V (most of the time).

Overclocked my AGP card using CoolBits. The card is an nVidia-based card produced by MadDog (their label Champion MX 400 - the GeForce2 MX400 with 64MB SDR 133MHz). The GPU is set at 200MHz.
Currently overclocked as follows:
GPU originally 200MHz - now 251MHz
Card RAM originally 133MHz - now 220MHz

Screen resolution set to 1024 x 768, 32 bit color.

As the commerical says, "Zoom! Zoom!".
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

What does the SPD say for your RAM?
sylvan2626
Senior K6'er
Posts: 139
Joined: Tue Apr 05, 2005 3:25 pm
Location: Florida, USA

Post by sylvan2626 »

For the module in question (SPD from Everest):

Field Value
Memory Module Properties
Serial Number None
Module Size 256 MB (2 rows, 4 banks)
Module Type Unbuffered
Memory Type SDRAM
Memory Speed PC100 (100 MHz)
Module Width 64 bit
Module Voltage LVTTL
Error Detection Method None
Refresh Rate Normal (15.625 us), Self-Refresh

Memory Timings
@ 125 MHz 3.0-3-3-6 (CL-RCD-RP-RAS)
@ 100 MHz 2.0-2-2-5 (CL-RCD-RP-RAS)

Memory Module Features
Early RAS# Precharge Not Supported
Auto-Precharge Supported
Precharge All Supported
Write1/Read Burst Supported
Buffered Address/Control Inputs Not Supported
Registered Address/Control Inputs Not Supported
On-Card PLL (Clock) Not Supported
Buffered DQMB Inputs Not Supported
Registered DQMB Inputs Not Supported
Differential Clock Input Not Supported
Redundant Row Address Not Supported
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

See my PCChips PM.

You should be OK @124MHz, since this SPD shows -
Memory Timings @125 MHz: 3.0-3-3-6 (CL-RCD-RP-RAS)

You *might* be able to get the module to run at 133MHz with the above timing...if it happens to be a *good* module...

Of course, finding some PC133@2.0-2-2-X would also be nice. ;-)

Did we ever scope out what your L2 cache can run at (markings)?
Jim
K6'er Elite
Posts: 1745
Joined: Wed Jan 21, 2004 7:10 pm
Location: Toronto

Post by Jim »

You just got my attention. How does one go about scoping out what the L2(3) cache is capable of running at?
Superpuppy 3
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
sylvan2626
Senior K6'er
Posts: 139
Joined: Tue Apr 05, 2005 3:25 pm
Location: Florida, USA

Post by sylvan2626 »

KachiWachi,

Read the PM of Sept 10 last night and will reply to it. This is the board and components in question, though.

I bought those RAM modules - 4 of them - as advertised "PC133 fully compatiable with PC100 boards" and then the modules turn out to be PC125. I don't know whether I'm more irratated at them for misrepresenting a product or me for not checking them sooner (I'm beyond my product return period). I was wanting to try 133MHz FSB.

Anyway... No on the L2 chip question. The chip's markings are as follows:

EliteMT
LP61G 6464AF-5
9926S N5E03NB (I believe that "N5E03NB" contains a "zero" rather than the letter "O".)

I have been using memory timings @ 124MHz of 2.0-3-3-6 some of the time and 3.0-3-3-6 some of the time.

I still plan to find some true PC133 modules, just to see if the system will hit 133MHz FSB.

Currently running at 5.5 x 115MHz FSB with no (mobo) L2.
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

What about the TAG chip?

Jim - The same way you do for any other part...check out the datasheet. The tricky part is finding them. :wink:
sylvan2626
Senior K6'er
Posts: 139
Joined: Tue Apr 05, 2005 3:25 pm
Location: Florida, USA

Post by sylvan2626 »

I would expect that to be around the two L2 chips, but I am not sure what I am looking for. Any clues, like typical shape or markings?
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

Long and skinny SOJ or TSOP (usually).

I couldn't find a datasheet, but a -5 would be 5ns, or 200 MHz. So @133MHz, you should be OK...unless the TAG is not that fast (needs to be 7.5nS or better for 133MHz).

Motherboard layout can also influence this, so just because the chips can do it, doesn't mean that the motherboard trace layout can. Hopefully the designer took this into account when he laid out the PCB.
smoke
Newbie K6'er
Posts: 18
Joined: Sun Feb 27, 2005 6:25 am

Post by smoke »

Uranium235 wrote:I just find it a little odd that Epox would give the K6+ CPU's the correct recognition in their BIOS's but not expand the Write Allocation feature for them. I use Rawpower to enable WA on my board but it would be convenient if it was handled in the BIOS.
I had always thought my epox mvp3g5 was enabling WA. I remember way back to when I was using the k6-iii+ 450 in a tyan s1590. I loaded a util to enable WA via config.sys/autoexec.bat as many did at the time. When WA was enabled sisoft sanra memory benchmark showed significant increase vs. not having WA enabled. Later on I picked up an mvp3g5 rev 1.4 (2000-12-21 BIOS). I did some testing and found that trying to enable WA using the previous method was not having any effect on benchmark scores. With no K6 specific tweaks at all I was getting 195-200 MB/s (memory & FSB @ 100MHz CAS 2), which if I remember correctly was basically about the max you could achieve on a mvp3 chipset with a non-overclocked 100MHz FSB. This is why I just assumed that WA was enabled. What benchmark are you using that shows a difference between having WA enabled/disabled?

Uranium235 wrote:Sylvan, here is the only known link I have for the 10/26/01 Epox Germany Beta BIOS:
http://www.geocities.com/steves1977uk/biosmods/
I found it by accident a few weeks ago while searching this forum. It is the G5BIOS.BIN at the link and has been slightly modified by Sandells to unlock some hidden memory settings.
I dont understand. Does sandell's work for epox, or is this just a BIOS he modified and epox happens to be hosting it? Also, AFAIK the last official BIOS for the mvp3g2/5 is dated 2000-12-21. Was this BIOS what sandell's used as a starting point for his BIOS you linked to?
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

Getting the CPU recognized by the BIOS is pretty easy...maybe two lines of code.

Enabling WA is another matter all together. They simply may not have had room in the code to do this...so they didn't.
Uranium235

Post by Uranium235 »

I recently tested my EP-MVP3G5 with a K6-2+ 570. Write Allocation was enabled for the K6-2+ and the CPU was correctly recognized by name. The Write Allocation problem only seems to be with my K6-3+ 400ATZ , very odd.

Nearly every benchmark was affected in some way with Write Allocation disabled. I ran CPUmark99, Sandra, Wintune 98, MDK2 and Final Reality.

The Epox Germany Beta BIOS was modified by Sandells with a program called Modbin which is widely available. Ive used it myself many times. Sandells doesn't work for Epox. I don't believe this BIOS version was ever available to the general public, the origin is a little unclear.
Uranium235

Post by Uranium235 »

Maybe this old post can better explain the origin of the 10/26/2001 MVP3G5 BIOS:

http://www.k6plus.com/index.php?name=PN ... ght=mvp3g5

I don't see any mention of what specific fixes or optimizations were made with this Beta BIOS. This is all I know.
User avatar
KachiWachi
K6'er Elite
Posts: 507
Joined: Wed Sep 21, 2005 10:53 am
Location: Pennsylvania, USA

Post by KachiWachi »

Uranium235 - Is the K6-III/+ being recognized as such?
Post Reply