It looks like it displays wrong values because the registers have not been unlocked.
Please run it with this command
ctchip34 ali1429
Then press enter
Then $ and C5 and enter
then read the values

byte_76 wrote on 2025-07-26, 14:09:

Disruptor wrote on 2025-07-26, 10:57:

Ah, IO-Redirection may not work with ALI 1429. You need to run the program manual and proceed the unlock by writing C5. Please t […]
Show full quote

Ah, IO-Redirection may not work with ALI 1429.
You need to run the program manual and proceed the unlock by writing C5.
Please test whether IO-Redirection to a file works or not.
Sorry again.

That seems to have worked.

@Disruptor, Using CTCHIP to troubleshoot this low performance x4 multiplier issue is a great idea.
Unfortunately, this didn’t work. All register values are FFh, so the registers Unlock didn’t work.

Checking the ALI1429.CFG file, I notice that the Unlock/Lock MACRO’s are incorrect.
I believe the correct MACRO’s should be:

MACRO OPEN =22h:00000011,23h:11000101
MACRO CLOSE=22h:00000011,23h:00000000

@byte_76
So please try again after replacing the MACRO’s in the ALI1429.CFG with the above.
(The .CFG files are text-files and can be edited with any text-editor.)

The past few weeks I have been working on an updated ALI1429G.CFG file for CTCHIPZ. Now that I have a good disassembly listing from the ATC-1762 VER-2.0 BIOS, I expect to find the meaning of some more register bit definitions.
I will post the updated ALI1429G.CFG file here when ready.

Jan

Disruptor wrote on 2025-07-26, 15:16:

It looks like it displays wrong values because the registers have not been unlocked. Please run it with this command ctchip34 al […]
Show full quote

It looks like it displays wrong values because the registers have not been unlocked.
Please run it with this command
ctchip34 ali1429
Then press enter
Then $ and C5 and enter
then read the values

Hi Disruptor,

I didn't see your reply when I wrote mine.
Yes, this should also work. 😀

Cheers, Jan

Disruptor wrote on 2025-07-26, 15:16:

It looks like it displays wrong values because the registers have not been unlocked. Please run it with this command ctchip34 al […]
Show full quote

It looks like it displays wrong values because the registers have not been unlocked.
Please run it with this command
ctchip34 ali1429
Then press enter
Then $ and C5 and enter
then read the values

New run attached with C5 and CFG with new values as per Jan's post.

I ran the app, as soon as it launched I typed $ and pressed enter and then typed C5 and pressed enter

Ok, it looks like the tool delivers plausible values now.

First the 100 MHz and 133 MHz settings look similar.

Which BIOS settings do you have configured?
I'd like to double check whether your BIOS settings match to the settings that tool has reported.
Please try to show all pages from BIOS.

As I have seen so far your configured settings are suboptimal.
But there are people at Vogons who can help you more.

Disruptor wrote on 2025-07-26, 18:11:

Ok, it looks like the tool delivers plausible values now. […]
Show full quote

Ok, it looks like the tool delivers plausible values now.

First the 100 MHz and 133 MHz settings look similar.

Which BIOS settings do you have configured?
I'd like to double check whether your BIOS settings match to the settings that tool has reported.
Please try to show all pages from BIOS.

As I have seen so far your configured settings are suboptimal.
But there are people at Vogons who can help you more.

Spoiler

INDEX=03h C5; UNLOCK, Please write C5 here else it won't work
03h 76543210 11000101; WRITE C5 HERE

*******************************************************************************
INDEX=12h 10; Hidden refresh
12h 4 1; Hidden refresh: Enabled

*******************************************************************************
INDEX=16h 05; DRAM Timings
16h 32 01; DRAM write timing (Slow, normal, fast, fastest): Normal
16h 10 01; DRAM read timing (Slow, normal, fast, fastest): Normal

*******************************************************************************
INDEX=17h 79;Cache stuff
17h 54 11; HITMJ sampling time (2t, 3t, 4t, 5t): 5t
17h 01 01; Cycle Check point (resvd, fast, normal, slow) : Fast

*******************************************************************************
INDEX=19h 5C; Cache timings
19h 7 0; Cache read timings (0WS, 1WS): 0WS
19h 6 1; Cache write timings (0WS, 1WS): 1WS

*******************************************************************************
INDEX=1Ah 0F; RAS-TO-CAS
1Ah 76 00; RAS-to-CAS delay (2t, 4t, 6t, resvd): 2t

*******************************************************************************
INDEX=1Eh F3; VESA
1Eh 4 1; VESA master cycle ADSJ (normal, delay): Delay
1Eh 3 0; VESA L2 Cache Read (normal, long) : Normal
1Eh 2 0; VESA L2 Cache Write (normal, long): Normal

*******************************************************************************
INDEX=20h 04; Clocks divider, weird, proceed with caution
20h 210 100; At Bus Clock (7.19Mhz, CLK2/4, /5, /6, /8, /10, /12, resvd): CLK2/8
20h 654 000; Polling clock (14.318Mhz, CLK2, /2, /3, /4, 28.636, rsv, rsv): 14.318Mhz

*******************************************************************************
INDEX=22h 80;Recovery / ISA write cycle
22h 2 0; ISA write cycle (0WS, 1WS): 0WS
22h 1 0; Onboard IO Recovery (Disabled, Enabled): Disabled
22h 0 0; IO Recovery (Disabled, Enabled): Disabled

*******************************************************************************
INDEX=25h C0; ISA timings, slow refresh
25h 76 11; 32 Bit ISA wait time (2t, 4t, 6t, 8t): 8t
25h 54 00; Slow Refresh (Disaled, 30us, 60us, 120us): Disabled
25h 32 00; 16 Bit ISA memory command (0WS, 1WS, 2WS, 3WS): 0WS
25h 10 00; 16 Bit ISA command (0WS, 1WS, 2WS, 3WS): 0WS

*******************************************************************************
INDEX=27h 30; Stuff
27h 54 11; LDEVJ Chk point delay (Disable, 1CLK2, 2CLK2, 3CLK2): 3 CLK2
27h 1 0; Internal ADS delay (Disabled, Enabled): Disabled
27h 0 0; Local ready syn mode (SYN, BYPASS): SYN

I deliberately left the BIOS settings on defaults (Loaded BIOS defaults) while messing with the system. The only changes I made were to ensure that the L1 and L2 cache are enabled and set them to WB. (And a couple of other settings that have no influence on performance.)

I thought we were trying to figure out why Speedys is reporting such unusual results at 133MHz.

For my own sanity, with the system at 133MHz, I tuned the BIOS settings and ran Speedsys again. The CPU score increased to 19.03. The cache "Moving" result increased to around 165MB/s while the "Reading" result was around 125MB/s and "Writing"was around 25MB/s.

These are quite good results for the cache tests while the CPU score is obviously way too low. So tuning the BIOS does not fix the result in Speedsys.

Doom (640x480) scores 42.01 FPS.

Quake (640x480) scores 12.2 FPS.

Overall, the system seems to perform okay. It's not amazingly fast but it's also not terrible either.

Understood.
That's a significant increase in SpeedSys' CPU score.
And your game scores look okay.

This chipset is known as not the best performer.
I have no idea what is the reason for the low SpeedSys CPU score.

Chkcpu wrote on 2025-07-26, 17:23:

The past few weeks I have been working on an updated ALI1429G.CFG file for CTCHIPZ. Now that I have a good disassembly listing from the ATC-1762 VER-2.0 BIOS, I expect to find the meaning of some more register bit definitions.
I will post the updated ALI1429G.CFG file here when ready.

Jan

Hi Byte_76,

The updated ALI1429G.CFG file for CTCHIPZ is ready now.
I‘ve put it up in Rav’s original thread about the ALi1429 chipset, for anyone who likes to tinker with their ALi1429 board.
Re: Overclocking a TI486DX2

Cheers, Jan

Chkcpu wrote on 2025-08-03, 19:03:

Hi Byte_76, […]
Show full quote

Chkcpu wrote on 2025-07-26, 17:23:

The past few weeks I have been working on an updated ALI1429G.CFG file for CTCHIPZ. Now that I have a good disassembly listing from the ATC-1762 VER-2.0 BIOS, I expect to find the meaning of some more register bit definitions.
I will post the updated ALI1429G.CFG file here when ready.

Jan

Hi Byte_76,

The updated ALI1429G.CFG file for CTCHIPZ is ready now.
I‘ve put it up in Rav’s original thread about the ALi1429 chipset, for anyone who likes to tinker with their ALi1429 board.
Re: Overclocking a TI486DX2

Cheers, Jan

Hi Jan,

Thank you for this info.

I'll send an new output using the updated CFG file as soon as I have a chance to play with the board again.