VOGONS


Reply 80 of 202, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++

Would front panel FSB or multiplier toggles in addition drop performance down the 386 territory?

All hail the Great Capacitor Brand Finder

Reply 81 of 202, by clueless1

User metadata
Rank l33t
Rank
l33t
gdjacobs wrote:

Would front panel FSB or multiplier toggles in addition drop performance down the 386 territory?

I've never done it, but I assume if it could be done, one could put a switch between 83 and 66 FSB. The jumper configs are pretty complicated though. Not sure how one would do that. Has anyone ever posted photos of a similar job they did? A visual of the wiring along with the jumper config would help.

On a practical note, my 486 already can get down to WC1 speeds.

What I find more interesting is the fact that CCD and DCD switches work with Ultima 7 when L1D does not.

The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks

Reply 82 of 202, by boxpressed

User metadata
Rank Oldbie
Rank
Oldbie

I had associated SETMUL with my K6-3+ build for so long that I didn't realize that it would work with my Cyrix 5x86-120.

I decided to see how slowly I could get the CPU to run. When it is running full-speed (3x @ 40MHz bus), it is close to a Pentium 90 in performance.

To begin, I jumpered the FSB to 25MHz so that the CPU ran at 75MHz.

Now I can slow my system down enough to do a proper recording of Heart of China for my wavetable thread. Faster CPUs allow you to hear only about 20-30 seconds of the intro music before going to the first scene.

I could do this all by jumper, of course, but SETMUL is so convenient.

Here are the Speedsys results if anyone in interested.

CYRIX 5X86-75                        SPEEDSYS

34.33
INT CACHE DISABLED 11.72
INT CACHE DISABLED + SETMUL 1 7.21
INT & EXT CACHE DISABLED + SETMUL 1 6.99



Reply 84 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t
boxpressed wrote:
I had associated SETMUL with my K6-3+ build for so long that I didn't realize that it would work with my Cyrix 5x86-120. .... […]
Show full quote

I had associated SETMUL with my K6-3+ build for so long that I didn't realize that it would work with my Cyrix 5x86-120.
....

CYRIX 5X86-75                        SPEEDSYS
34.33
INT CACHE DISABLED 11.72
INT CACHE DISABLED + SETMUL 1 7.21
INT & EXT CACHE DISABLED + SETMUL 1 6.99

Indeed, it is very cool that the Cyrix 5x86 can switch to and from the 1.0x multiplier from the command prompt. This way the Cx5x86 can substitute the common types of 486 processors: The original 486DX-33 at multi 1.0x, the 486DX/2-66 at multi 2.0x and the 486DX/4-100 at multi 3.0x. Though switching between 2.0x and 3.0x requires a reboot and one jumper to be toggled, or a case-switch attached to that jumper.
When I tested it on my 486: 66 MHz without L1 cache/No-Turbo is stil too fast for many classics like Death-Track, but the same at 33MHz made that game run properly. Compare that to the AMD 5x86 with 3.0x and 4.0x multi; Better off with the 1.0x, 2.0x and 3.0x options on the Cyrix.
(PS. Just talking about the normal Cyrix 5x86 models with a maximum multiplier of 3.0x, knowing there are rare 4.0x ones too)

SetMul v1.21 wil be up in the coming weeks. As the numbering indicates it is just a minor feature update: It will read from Pentium II/III registers the multiplier setting that was configured at bootup.
This is not as straightforward as it may seem, as there are about 5 different lookup tables depending on the processor model/stepping and multiplier readout value. Intel was obviously running out of possible values in their small 4-bit multiplier register and worked around that in different ways. I could not find proper documentation, but tests with my E.S. CPUs and an example from CrystalCPUID should suffice. Fun stuff. 😀

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 86 of 202, by clueless1

User metadata
Rank l33t
Rank
l33t
gerwin wrote:

SetMul v1.21 is available now in the 1st post of this topic.

Thanks, gerwin!

The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks

Reply 87 of 202, by sonicpro

User metadata
Rank Newbie
Rank
Newbie

Is there any limitation of using SetMul on Nehemiah processor installed on i815 Tualatin-compatible board? I do not think I have the right slotket for Nehemiah, though I have a Tualatin-compatible board manufactured by PCPartner.

Reply 88 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t

When the C3 Nehemiah processor functions on your motherboard, then it is all the same to SetMul. (Though the lowest possible speed settings may depend on the motherboard to some extend, as some systems lock when trying to set the nehemiah to a mere 200..266MHz.)

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 89 of 202, by m5215tx

User metadata
Rank Member
Rank
Member

I have a VIA C3-700AMHz processor + a socket 370 to slot 1 adapter with adjustable bus speeds (66MHz and 100MHz) + an Intel SE440BX-2 motherboard.

Can I mount my C3 on the socket 370 to slot 1 adapter and set it for a bus speed of 66MHz then install it on the SE440BX-2 motherboard and from MS-DOS 6.22 use SetMul to adjust the speed to play old and newer DOS games?

Sounds like it might work unless I am missing something.

Roland MT-32 (old), CM-32LN, SC-55, SC-88VL, MT-120, SD-35, SD-20, SD-80, SD-90
Yamaha TG100, TG300, MDF2, MU15, MU100, MU2000EX + PLG150-DR + PLG150-PF + PLG150-VL
KORG NS5R, X5DR
AKAI SG01k
KAWAI GMega
KETRON SD2

Reply 90 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t

It should work.
In this topic: VIA C3 Nehemiah vs. Coppermine(-128) and Tualatin. The VIA C3 compatibility of the SE440BX-2 Motherboard is demonstrated. With BIOS version P13 that is.
For earlier VIA C3 models 'Samuel' and 'Ezra', like yours, most Slotkets will work.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 91 of 202, by m5215tx

User metadata
Rank Member
Rank
Member

I am using my VIA C3-700AMHz processor with a VIA chipset based motherboard I acquired so the CPU is fully recognized and used properly. I have MS-DOS 6.22 and I have run SetMul and I see all the options and have been experimenting with it. Is the CPU stable after its clocked down and features disabled? When I run Privateer the cutscene seems to get a little out of sync. At first I thought I had made the CPU too slow but when I had it running at about the speed of a 486DX2 66MHz I still noticed the sync issues. My 386DX 40MHz does not have this issue so something is up.

Roland MT-32 (old), CM-32LN, SC-55, SC-88VL, MT-120, SD-35, SD-20, SD-80, SD-90
Yamaha TG100, TG300, MDF2, MU15, MU100, MU2000EX + PLG150-DR + PLG150-PF + PLG150-VL
KORG NS5R, X5DR
AKAI SG01k
KAWAI GMega
KETRON SD2

Reply 92 of 202, by Kamerat

User metadata
Rank Oldbie
Rank
Oldbie
m5215tx wrote:

I am using my VIA C3-700AMHz processor with a VIA chipset based motherboard I acquired so the CPU is fully recognized and used properly. I have MS-DOS 6.22 and I have run SetMul and I see all the options and have been experimenting with it. Is the CPU stable after its clocked down and features disabled? When I run Privateer the cutscene seems to get a little out of sync. At first I thought I had made the CPU too slow but when I had it running at about the speed of a 486DX2 66MHz I still noticed the sync issues. My 386DX 40MHz does not have this issue so something is up.

Can you tell us what graphic and sound card you're using and how is the monitor connected?

DOS Sound Blaster compatibility: PCI sound cards vs. PCI chipsets
YouTube channel

Reply 93 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t

@m5215tx

The CPU is not unstable, since an unstable CPU would have locked up. What particular SetMul switch causes the problem, is it the downclocking or the disabling of caches?
Note that this SetMul tool is experimental, it is based on CPU options mentioned in the old datasheets and observations of the apparent results. It is what it is.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 94 of 202, by m5215tx

User metadata
Rank Member
Rank
Member

Upon further examination I don't think SetMul is the issue here so for now please disregard this inquiry. I think I have some other hardware issue here so I may open a different thread elsewhere for some assistance with this. Thanks for the feedback.

Roland MT-32 (old), CM-32LN, SC-55, SC-88VL, MT-120, SD-35, SD-20, SD-80, SD-90
Yamaha TG100, TG300, MDF2, MU15, MU100, MU2000EX + PLG150-DR + PLG150-PF + PLG150-VL
KORG NS5R, X5DR
AKAI SG01k
KAWAI GMega
KETRON SD2

Reply 95 of 202, by lvader

User metadata
Rank Member
Rank
Member

Been playing with a 5x86 build, the turbo button doesn't seem to make much difference so I tried it on the multiplier jumper. Unlike other Pentiums, here it changes the multiplier on the fly so the turbo button switches between 66 and 100 MHz and with setmul I can switch to 33 MHz as well.

Reply 96 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t
lvader wrote:

Unlike other Pentiums, here it changes the multiplier on the fly

3x Multi to 2x Multi without reboot? Had not tried that, but that is neat!

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 98 of 202, by MERCURY127

User metadata
Rank Member
Rank
Member

run SetMul on my P3 800EB (6x133) and get output^
multiplier 6.0
cpu frequency 939 🤣
something wrong?
mobo GA-6VXC7-4X (VIA 694X & 686A)
details here https://www.gigabyte.com/Motherboard/GA-6VXC7-4X-rev-50#sp
and here http://www.motherboard.cz/mb/gigabyte/GA-6VXC7-4X-P.htm
and here NO overclock at now!

Reply 99 of 202, by gerwin

User metadata
Rank l33t
Rank
l33t

Does it do this consistently at every run? SetMul just does a read-TSC directly from the CPU, waits a bit, and does another read-TSC. Speed in MHz is calculated from the TSC change over that time period.
Just tested it on Tauleron Pentium III running at 8.0x133=1066MHz. SetMul gets it right every time here...
I have a similar VIA board in storage but not set-up for testing.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul