VOGONS


First post, by Vendest

User metadata
Rank Newbie
Rank
Newbie

Looking for setting recommendations due to poor performances shown in benchmark such as Speedsys and LSST.
Motherboard is a TMC TI5VG+, 100 MHz FSB and 128mb of RAM
Paired with an AMD K6 2+ 500 MHz and the bios is patched to support this CPU.

I may have set a DIP switch wrongly or could the CPU be defective?

Attachments

Reply 1 of 12, by Gmlb256

User metadata
Rank l33t
Rank
l33t

Try running K6INIT, perhaps it could be that write allocate feature isn't being properly configured for the K6-2+ CPU.

VIA C3 Nehemiah 1.2A @ 1.46 GHz | ASUS P2-99 | 256 MB PC133 SDRAM | GeForce3 Ti 200 64 MB | Voodoo2 12 MB | SBLive! | AWE64 | SBPro2 | GUS

Reply 2 of 12, by Vendest

User metadata
Rank Newbie
Rank
Newbie
Gmlb256 wrote on 2023-11-20, 17:45:

Try running K6INIT, perhaps it could be that write allocate feature isn't being properly configured for the K6-2+ CPU.

Thanks, I loaded K6INIT and tried Speedsys again but there are no changes.
I wanted to rule out a possible issue with the motherboard so I installed a spare MMX200 and launched Speedsys again.
Reading is much slower but test seems to be more consistent. To be frank I am not sure if the results are correct for MMX 200

Thanks again for sharing about K6INIT, I'll run some Demo benchmarks to check.

Attachments

Reply 3 of 12, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie

For comparison, here's my Epox MVP3G-M with a K6-2+ with the cache enable mod. Different motherboard, higher clock speed, higher bus speed, and more L2 cache; it isn't an apples to apples comparison, but your system should be closer to this than it is.

Attachments

"I'd rather be rich than stupid" - Jack Handey

Reply 4 of 12, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie
Vendest wrote on 2023-11-20, 17:23:

I may have set a DIP switch wrongly or could the CPU be defective?

It's very hard to believe that it's the CPU's fault.
I would say that it's most likely related to the motherboard itself (possibly Jan's patched BIOS does not work properly on all MB revisions?)

1 x PLCC-68 / 2 x PGA132 / 5 x Skt 3 / 9 x Skt 7 / 12 x SS7 / 1 x Skt 8 / 14 x Slot 1 / 5 x Slot A
5 x Skt 370 / 8 x Skt A / 2 x Skt 478 / 2 x Skt 754 / 3 x Skt 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k

Reply 5 of 12, by CoffeeOne

User metadata
Rank Oldbie
Rank
Oldbie
Vendest wrote on 2023-11-20, 17:23:
Looking for setting recommendations due to poor performances shown in benchmark such as Speedsys and LSST. Motherboard is a TMC […]
Show full quote

Looking for setting recommendations due to poor performances shown in benchmark such as Speedsys and LSST.
Motherboard is a TMC TI5VG+, 100 MHz FSB and 128mb of RAM
Paired with an AMD K6 2+ 500 MHz and the bios is patched to support this CPU.

I may have set a DIP switch wrongly or could the CPU be defective?

Yes, it is clearly too slow.
As others said, 100% not a defective cpu. It is either motherboard and Bios or Bios only.

"and the bios is patched to support this CPU."
What does that mean exactly? Which Bios did you use?

Reply 6 of 12, by Vendest

User metadata
Rank Newbie
Rank
Newbie

The patched bios is from http://www.steunebrink.info/k6plus.htm TI5VG1J2.ZIP.
Designed to enable support for SS7 motherboard with K6-2+ and K6-3+.

What puzzles me is the writing performance with both K6-2+ and MMX 200.

I must also state that the motherboard was damaged during shipping when I received it several months ago, I fixed 2 traces with botch wires.
It is possible I missed to fix a trace.

Reply 7 of 12, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

usb card?

WT/WB?

disable system bios cachable

ensure write allocate is on

turbo button

just throwing some things at the wall, see if anything sticks.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 8 of 12, by CoffeeOne

User metadata
Rank Oldbie
Rank
Oldbie
Vendest wrote on 2023-11-22, 02:27:
The patched bios is from http://www.steunebrink.info/k6plus.htm TI5VG1J2.ZIP. Designed to enable support for SS7 motherboard wit […]
Show full quote

The patched bios is from http://www.steunebrink.info/k6plus.htm TI5VG1J2.ZIP.
Designed to enable support for SS7 motherboard with K6-2+ and K6-3+.

What puzzles me is the writing performance with both K6-2+ and MMX 200.

I must also state that the motherboard was damaged during shipping when I received it several months ago, I fixed 2 traces with botch wires.
It is possible I missed to fix a trace.

You are right, values are also too slow for the MMX 200.
So it is not an issue about K6 2+. You could continue testing with the original Bios (to exclude one possible problem) and the MMX 200.
Did you reset all Bios to "Setup default" already?

Reply 9 of 12, by Vendest

User metadata
Rank Newbie
Rank
Newbie

Thanks for the recommendations, I've checked accordingly but there is no improvement.

Sphere478 wrote on 2023-11-22, 04:18:
usb card? WT/WB? disable system bios cachable ensure write allocate is on turbo button […]
Show full quote

usb card?
WT/WB?
disable system bios cachable
ensure write allocate is on
turbo button

I removed any pci cards but the AGP video card, made sure to disable system bios cacheable and enabled Cache Rd+CPU Wt Pipeline
That board does not have a turbo button (turbo LED only).

CoffeeOne wrote on 2023-11-22, 06:46:

You are right, values are also too slow for the MMX 200.
So it is not an issue about K6 2+. You could continue testing with the original Bios (to exclude one possible problem) and the MMX 200.
Did you reset all Bios to "Setup default" already?

I flashed the original bios that the board initially came with and installed the MMX 200.
Reset the bios to default and then done some adjustment about cache and ram.

The motherboard was in a boot loop when I first received it, it worked after fixing some traces. Couple of months later I may need to pull out the microscope once more 😀

Reply 10 of 12, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Is there a way to tell if wt/wb is in wb mode in software?

Alternatively you could do the mod

I don’t think this is it, but I’m kinda at a loss here, this is odd

K6-2/3+ Success on Asus P5A 1.06 - simple mod

Probably not the problem.

Try a different video card

Does it happen in straight dos and windows?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 11 of 12, by Vendest

User metadata
Rank Newbie
Rank
Newbie
Sphere478 wrote on 2023-11-22, 18:20:
Is there a way to tell if wt/wb is in wb mode in software? […]
Show full quote

Is there a way to tell if wt/wb is in wb mode in software?

Alternatively you could do the mod

I don’t think this is it, but I’m kinda at a loss here, this is odd

K6-2/3+ Success on Asus P5A 1.06 - simple mod

Probably not the problem.

Try a different video card

Does it happen in straight dos and windows?

Not sure how to check wt/wb in software nor there is any settings in the bios.
I soldered a pull-up resistor between vcc3 and wt/wb but there's no change in speedsys.
Tried with different PCI and AGP video card, no dice.
Thanks for your help, I think there is a bigger issue with the motherboard.

I have yet to try in Win9x, is there a speedsys equivalent for windows? or should I use 3dmark?

Reply 12 of 12, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Yeah see what 3d mark does

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)