LeFlash wrote on 2021-01-25, 07:32:

I doubt there's an easy generic mod. You have to check to find a trace where your board sends the 3.3V to the CPU and cut there.
The other components that need 3.3V won't be happy with 2.5V. Every board is different in layout.

Exactly.
It depends on the board´s layout. That´s why I had chosen the GA586DX, it already has a diode in the IO voltage line of each processor (probably for decoupling) and I just unsoldered one end and put a second diode in series. The Vcore mod is no big deal either, just change the values of the resistors at the regulators accordingly.
The heat dissipation is about the same as when using regular Pentium MMX 233. The voltage drop is higher (more heat) but the current the Tillamook causes is lower (less heat).
I calculated all this back in the days, made the mods but - due to lack of time- never managed to put 2 Tillammoks in for a test.
Later I gave this board away, I still have one, but this is in use running a neat NT4 machine as FTP server (with regular MMX CPUs).

majestyk wrote on 2021-01-25, 08:12:

Exactly. It depends on the board´s layout. That´s why I had chosen the GA586DX, it already has a diode in the IO voltage line o […]
Show full quote

LeFlash wrote on 2021-01-25, 07:32:

I doubt there's an easy generic mod. You have to check to find a trace where your board sends the 3.3V to the CPU and cut there.
The other components that need 3.3V won't be happy with 2.5V. Every board is different in layout.

Exactly.
It depends on the board´s layout. That´s why I had chosen the GA586DX, it already has a diode in the IO voltage line of each processor (probably for decoupling) and I just unsoldered one end and put a second diode in series. The Vcore mod is no big deal either, just change the values of the resistors at the regulators accordingly.
The heat dissipation is about the same as when using regular Pentium MMX 233. The voltage drop is higher (more heat) but the current the Tillamook causes is lower (less heat).
I calculated all this back in the days, made the mods but - due to lack of time- never managed to put 2 Tillammoks in for a test.
Later I gave this board away, I still have one, but this is in use running a neat NT4 machine as FTP server (with regular MMX CPUs).

I have a tyan tomcat coming in the mail can I test smp between a tillamook and a 233mmx

trying to think of a way to test the theory without buying another tillamook 🤣

H3nrik V! wrote on 2021-01-25, 07:36:

gonzo wrote on 2021-01-24, 18:16:

Is there real no chance to get a multi of 4,5x to reach 300 MHz at 66 FSB?? […]
Show full quote

Is there real no chance to get a multi of 4,5x to reach 300 MHz at 66 FSB??

I have an AIR 54TDP here with HX-chipset (Dual-Socket-7; https://stason.org/TULARC/pc/motherboards/A/A … tium-54TDP.html) - it boots fine with L2 enabled and one or two (YEAHHH) Tillamook(s) , but at the moment only at 233 (1,5x=3,5x 66 MHz). Still to test possible undocumented FSB-options (over 66 MHz) by myself...
The Vcore is maybe very high (only split-voltage-option for P55C -> I assume only for 2,8 V...), but for the moment it boots stable, still without any OS (DOS etc.).
Until know no more tests made (no really time to do), but nice to have...

Does that board recognize both CPUs? It's pretty awesome if this is actually possible? And what about Operating System?

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS 🙁
-> Speedsys returns a CPU-result of about the level of a 486DX2-50 (at 233 MHz Tillamook-speed).
-> CheckCPU says "L1 Cache disabled", too.
-> the L1 can be enabled with SetMul -> unfortunately it works with DOS/9x/ME only 🙁

So, is there a possibility to enable the L1-Cache under 2000/XP, because of the using of both CPU-sockets?
Or, much better, is it possible to enable the L1-cache per some "wire-trick" as shown here for L2-cache, before entering any NT-system?
If not possible, the dual-Tillamook-project dies very soon, I'm afraid... 🙁

This could be due to the issue of cache-integrity.
It was outlined here quite convincing:

http://www.vcfed.org/forum/showthread.php?737 … mook-s&p=608177

I´m sure "Eudimorphodon" meant L1 cache when speaking of L2 cache, since he writes about the integrity of the "on die caches" which must be L1. (PI MMX lack a L2 cache.)

Sphere478 wrote on 2021-01-24, 20:19:

so is it safe to connect red yellow and green bridges? just to get it over with and that makes it work in all boards? socket 5/7 […]
Show full quote

so is it safe to connect red yellow and green bridges? just to get it over with and that makes it work in all boards? socket 5/7 assuming proper voltage? just so I understand. or can they hurt on some boards?

so I just got mine and it has the 4x multi bridge already installed does that mean that it will over ride any setting on the motherboard or does that mean I need to remove all multi jumpers from the motherboard? will they conflict? does this mean that the only way to get another setting is to remove the bridge?

we need a tillamook interposer 🤣 I hate modifying these chips but gotta do whatcha gotta do I guess.

also, is smp in any way possible on these with a mod or has it been totally removed?

The green connection will let the tillamook work where it usually doesn't? I mean motherboards like p5a and similar

gonzo wrote on 2021-01-25, 11:54:

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS :-( -> Speedsys returns a CPU-result […]
Show full quote

H3nrik V! wrote on 2021-01-25, 07:36:

gonzo wrote on 2021-01-24, 18:16:

Is there real no chance to get a multi of 4,5x to reach 300 MHz at 66 FSB?? […]
Show full quote

Is there real no chance to get a multi of 4,5x to reach 300 MHz at 66 FSB??

I have an AIR 54TDP here with HX-chipset (Dual-Socket-7; https://stason.org/TULARC/pc/motherboards/A/A … tium-54TDP.html) - it boots fine with L2 enabled and one or two (YEAHHH) Tillamook(s) , but at the moment only at 233 (1,5x=3,5x 66 MHz). Still to test possible undocumented FSB-options (over 66 MHz) by myself...
The Vcore is maybe very high (only split-voltage-option for P55C -> I assume only for 2,8 V...), but for the moment it boots stable, still without any OS (DOS etc.).
Until know no more tests made (no really time to do), but nice to have...

Does that board recognize both CPUs? It's pretty awesome if this is actually possible? And what about Operating System?

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS 🙁
-> Speedsys returns a CPU-result of about the level of a 486DX2-50 (at 233 MHz Tillamook-speed).
-> CheckCPU says "L1 Cache disabled", too.
-> the L1 can be enabled with SetMul -> unfortunately it works with DOS/9x/ME only 🙁

So, is there a possibility to enable the L1-Cache under 2000/XP, because of the using of both CPU-sockets?
Or, much better, is it possible to enable the L1-cache per some "wire-trick" as shown here for L2-cache, before entering any NT-system?
If not possible, the dual-Tillamook-project dies very soon, I'm afraid... 🙁

Are you running dual tillamook and it is showing two graphs under nt/2000/xp?

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

But of course interposers are hard to find these days.

Dual AMD K6 computer

majestyk wrote on 2021-01-25, 12:36:

This could be due to the issue of cache-integrity. It was outlined here quite convincing: […]
Show full quote

This could be due to the issue of cache-integrity.
It was outlined here quite convincing:

http://www.vcfed.org/forum/showthread.php?737 … mook-s&p=608177

I´m sure "Eudimorphodon" meat L1 cache when speaking of L2 cache, since he writes about the integrity of the "on die caches" which must be L1. (PI MMX lack a L2 cache.)

Oooo good link

Sphere478 wrote on 2021-01-25, 18:05:

Are you running dual tillamook and it is showing two graphs under nt/2000/xp? […]
Show full quote

gonzo wrote on 2021-01-25, 11:54:

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS :-( -> Speedsys returns a CPU-result […]
Show full quote

H3nrik V! wrote on 2021-01-25, 07:36:

Does that board recognize both CPUs? It's pretty awesome if this is actually possible? And what about Operating System?

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS 🙁
-> Speedsys returns a CPU-result of about the level of a 486DX2-50 (at 233 MHz Tillamook-speed).
-> CheckCPU says "L1 Cache disabled", too.
-> the L1 can be enabled with SetMul -> unfortunately it works with DOS/9x/ME only 🙁

So, is there a possibility to enable the L1-Cache under 2000/XP, because of the using of both CPU-sockets?
Or, much better, is it possible to enable the L1-cache per some "wire-trick" as shown here for L2-cache, before entering any NT-system?
If not possible, the dual-Tillamook-project dies very soon, I'm afraid... 🙁

Are you running dual tillamook and it is showing two graphs under nt/2000/xp?

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

But of course interposers are hard to find these days.

Dual AMD K6 computer

majestyk wrote on 2021-01-25, 12:36:

This could be due to the issue of cache-integrity. It was outlined here quite convincing: […]
Show full quote

This could be due to the issue of cache-integrity.
It was outlined here quite convincing:

http://www.vcfed.org/forum/showthread.php?737 … mook-s&p=608177

I´m sure "Eudimorphodon" meat L1 cache when speaking of L2 cache, since he writes about the integrity of the "on die caches" which must be L1. (PI MMX lack a L2 cache.)

Oooo good link

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

Sphere478 wrote on 2021-01-25, 18:05:

Are you running dual tillamook and it is showing two graphs under nt/2000/xp? […]
Show full quote

gonzo wrote on 2021-01-25, 11:54:

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS :-( -> Speedsys returns a CPU-result […]
Show full quote

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS 🙁
-> Speedsys returns a CPU-result of about the level of a 486DX2-50 (at 233 MHz Tillamook-speed).
-> CheckCPU says "L1 Cache disabled", too.
-> the L1 can be enabled with SetMul -> unfortunately it works with DOS/9x/ME only 🙁

So, is there a possibility to enable the L1-Cache under 2000/XP, because of the using of both CPU-sockets?
Or, much better, is it possible to enable the L1-cache per some "wire-trick" as shown here for L2-cache, before entering any NT-system?
If not possible, the dual-Tillamook-project dies very soon, I'm afraid... 🙁

Are you running dual tillamook and it is showing two graphs under nt/2000/xp?

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

But of course interposers are hard to find these days.

Dual AMD K6 computer

majestyk wrote on 2021-01-25, 12:36:

This could be due to the issue of cache-integrity. It was outlined here quite convincing: […]
Show full quote

This could be due to the issue of cache-integrity.
It was outlined here quite convincing:

http://www.vcfed.org/forum/showthread.php?737 … mook-s&p=608177

I´m sure "Eudimorphodon" meat L1 cache when speaking of L2 cache, since he writes about the integrity of the "on die caches" which must be L1. (PI MMX lack a L2 cache.)

Oooo good link

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

Don’t worry that it doesn’t know what it is it just means that the code that the cpu is sending isn’t something the bios has in it’s database.

Here is my k6 3 on a motherboard that prettymuch runs fine on it. Just doesn’t say the name and speed right. It’s actually running at 450mhz but bios is just confused on speed and name report. It doesn’t hurt anything.

Continue with tests as long as you think voltage is proper.

There are a lot of people who will be building a dual tillamook if you can succeed including myself. The second you tell me that you got two graphs in task manager I’m ordering another one.

Sphere478 wrote on 2021-01-25, 18:59:

Don’t worry that it doesn’t know what it is it just means that the code that the cpu is sending isn’t something the bios has in […]
Show full quote

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

Sphere478 wrote on 2021-01-25, 18:05:

Are you running dual tillamook and it is showing two graphs under nt/2000/xp? […]
Show full quote

Are you running dual tillamook and it is showing two graphs under nt/2000/xp?

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

But of course interposers are hard to find these days.

Dual AMD K6 computer

Oooo good link

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

Don’t worry that it doesn’t know what it is it just means that the code that the cpu is sending isn’t something the bios has in it’s database.

Here is my k6 3 on a motherboard that prettymuch runs fine on it. Just doesn’t say the name and speed right. It’s actually running at 450mhz but bios is just confused on speed and name report. It doesn’t hurt anything.

Continue with tests as long as you think voltage is proper.

There are a lot of people who will be building a dual tillamook if you can succeed including myself. The second you tell me that you got two graphs in task manager I’m ordering another one.

Without L1-cache enabled for a NT-system (2000/XP) the performance drops to a middle-range-486...That's what I mean with "not recognize"
The only possible scenario is the use of only one CPU (one socket) with SetMul under DOS/98/ME -> not interesting for me; this would be very sad for this board

gonzo wrote on 2021-01-26, 07:19:

Sphere478 wrote on 2021-01-25, 18:59:

Don’t worry that it doesn’t know what it is it just means that the code that the cpu is sending isn’t something the bios has in […]
Show full quote

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

Don’t worry that it doesn’t know what it is it just means that the code that the cpu is sending isn’t something the bios has in it’s database.

Here is my k6 3 on a motherboard that prettymuch runs fine on it. Just doesn’t say the name and speed right. It’s actually running at 450mhz but bios is just confused on speed and name report. It doesn’t hurt anything.

Continue with tests as long as you think voltage is proper.

There are a lot of people who will be building a dual tillamook if you can succeed including myself. The second you tell me that you got two graphs in task manager I’m ordering another one.

Without L1-cache enabled for a NT-system (2000/XP) the performance drops to a middle-range-486...That's what I mean with "not recognize"
The only possible scenario is the use of only one CPU (one socket) with SetMul under DOS/98/ME -> not interesting for me; this would be very sad for this board

I understand. If we can get smp working, it would be worth circling back around to the cache issue. there is probably a fix for that but if smp doesn't work then who cares 🤣 but if it does!!!!!

Sphere478 wrote on 2021-01-25, 18:05:

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

Well, somebody can give me a dual-socket-5-board for Tillamook-testing. It's a Siemens-Nixdorf S26361-D823 (or SNI D823, or PCD-5T ) from a Server Primergy 350/351.

The problem: it's an EISA-Board, and it absolutely (!) needs a so-called EISA Configuration Utility (ECU) to adjust the FSB-Frequency (actually they are two P75-CPUs mounted on it, that means a very low FSB of 50 MHz only 🙁 )
There is no chance to change the FSB using the "normal" BIOS 🙁

Can this ECU-tool can be uploaded here from somebody, or is there a link to download it? That would be nice.

I still do not have any idea, if the Tillamook-core with a voltage interposer will be recognized by this board (I still do not have it in front of me). Maybe the problem of using of two Tillamooks together will still remain, but anyway...

gonzo wrote on 2021-01-29, 10:46:

Well, somebody can give me a dual-socket-5-board for Tillamook-testing. It's a Siemens-Nixdorf S26361-D823 (or SNI D823, or PCD- […]
Show full quote

Sphere478 wrote on 2021-01-25, 18:05:

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

Well, somebody can give me a dual-socket-5-board for Tillamook-testing. It's a Siemens-Nixdorf S26361-D823 (or SNI D823, or PCD-5T ) from a Server Primergy 350/351.

The problem: it's an EISA-Board, and it absolutely (!) needs a so-called EISA Configuration Utility (ECU) to adjust the FSB-Frequency (actually they are two P75-CPUs mounted on it, that means a very low FSB of 50 MHz only 🙁 )
There is no chance to change the FSB using the "normal" BIOS 🙁

Can this ECU-tool can be uploaded here from somebody, or is there a link to download it? That would be nice.

I still do not have any idea, if the Tillamook-core with a voltage interposer will be recognized by this board (I still do not have it in front of me). Maybe the problem of using of two Tillamooks together will still remain, but anyway...

If it’s socket 5 it will definitely need a interposer upgrade adapter. Because those sockets will try and run the core at 3.5v which is no bueno. So don’t try it.

Btw, I did some voltage jumper experiments on my tyan tomcat. Got it to 2.15v

Re: Dual CPU Socket 5/7 Boards

You could always send one of your tillamooks my way 😀

Did some jumpers

Decided to do em fashionable 🤣. Try to keep em lookin pretty.

I got the 512k on motherboard cache working but the 4x doesn’t seem to be working. What do I have to set the motherboard multiplier at? Should I put more jumpers in?

More pics

Hey, that´s 768 MB of RAM! Without using Bank 0?? Or did you put FP-RAM in Bank 0?
So the BIOS handles more than 512 MB, would be interesting what the HX northbridge does under some OS. It would use just 512 MB of it I´m afraid.

Re: Dual CPU Socket 5/7 Boards more info

majestyk wrote on 2021-02-03, 07:39:

Hey, that´s 768 MB of RAM! Without using Bank 0?? Or did you put FP-RAM in Bank 0?
So the BIOS handles more than 512 MB, would be interesting what the HX northbridge does under some OS. It would use just 512 MB of it I´m afraid.

it seems these chipsets, the 430tx and 430hx say they can only do 256mb, and 512mb but in reality it seems with the right ram modules you can actually do 1.5x that. It works best with bank 0 unpopulated on my tyan tomcat if I put 8 x 128mb simm sticks in it it won’t post. But works fine on 768 with bank 0 open.

Memtest sees all the ram🤷‍♂️

I’ve run it a while now with 768, seems stable on the hx but it only seems to work with bank 0 open.

edit: ended up that 768 while it did pass mem test was not stable and only a glitch max ram on 430hx is indeed 512mb though you can get 768 to detect it simply does not work other than in memtest which isn't detecting the problem. i've beat the issue to death and there appears to simply be no way to make it work. save your self the trouble and stick to 512mb

I’ve done the 4x and l2 mod (blue and yellow on diagram) I’m reading the thread, seems maybe I should do the dual voltage mod also in case I put it in a board that detects it as single voltgage

Can I just do all the mods and be done with them? Or will they hurt on some boards?

Also, my board has multis for 1.5-3.5x how do I get to 4x? Use the 2x setting?

Btw, I tried doing smp with a tillamook in primary socket and a mmx233 in second socket and it only detected the tillamook. Dono if this means the tillamook won’t work or if that means that it won’t try to smp with mismatched processors it’s a tyan tomcat s1564. It’s thought that older socket 5 boards with powerleaps, non smp chipsets using proprietary smp solutions can probably smp amd, cyrix, and tillamook. There is a thread from someone who saw one of these in the wild but little details were given in the thread as they weren’t the owner of the computer

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

Sphere478 wrote on 2021-01-25, 18:05:

Are you running dual tillamook and it is showing two graphs under nt/2000/xp? […]
Show full quote

gonzo wrote on 2021-01-25, 11:54:

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS :-( -> Speedsys returns a CPU-result […]
Show full quote

BAD NEWS: the L1-cache of the Tillamook is automatically DISABLED on booting by the BIOS 🙁
-> Speedsys returns a CPU-result of about the level of a 486DX2-50 (at 233 MHz Tillamook-speed).
-> CheckCPU says "L1 Cache disabled", too.
-> the L1 can be enabled with SetMul -> unfortunately it works with DOS/9x/ME only 🙁

So, is there a possibility to enable the L1-Cache under 2000/XP, because of the using of both CPU-sockets?
Or, much better, is it possible to enable the L1-cache per some "wire-trick" as shown here for L2-cache, before entering any NT-system?
If not possible, the dual-Tillamook-project dies very soon, I'm afraid... 🙁

Are you running dual tillamook and it is showing two graphs under nt/2000/xp?

The best chance for dual tillamook is probably a early socket 5 motherboard with a voltage interposer like powerleap k6 III one

But of course interposers are hard to find these days.

Dual AMD K6 computer

majestyk wrote on 2021-01-25, 12:36:

This could be due to the issue of cache-integrity. It was outlined here quite convincing: […]
Show full quote

This could be due to the issue of cache-integrity.
It was outlined here quite convincing:

http://www.vcfed.org/forum/showthread.php?737 … mook-s&p=608177

I´m sure "Eudimorphodon" meat L1 cache when speaking of L2 cache, since he writes about the integrity of the "on die caches" which must be L1. (PI MMX lack a L2 cache.)

Oooo good link

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

can you give some clarification, was the motherboard showing in post 1 or 2 cpu detected? and what motherboard are you using?

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

Seems to me you have two issues:
1. Your motherboard doesn't have the correct microcode for the Tillamook, but it should be fixable in a similar manner to the LGA771 Xeon mods
2. You might need some more pins to be bridged to re-enable SMP, or it might also be fixed by modding your BIOS with the correct microcodes.

TechieDude wrote on 2021-03-02, 12:45:

Seems to me you have two issues: 1. Your motherboard doesn't have the correct microcode for the Tillamook, but it should be fix […]
Show full quote

gonzo wrote on 2021-01-25, 18:30:

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket. No luck: in the POST there is shown "i(unknown)", […]
Show full quote

BAD NEWS II: today I got another (2-nd) P-MMX-Tillamook, for the 2-nd socket.
No luck: in the POST there is shown "i(unknown)", and that's all. The same result using one or two Tillamook(s).
For instance, I used another two Pentiums P54C (without MMX) -> the POST results in "Intel(P54C)x2".
As I already wrote, the system is still not tested with any OS (I was simply researching the hardware-opportunities in advance)

So, for me, this mainboard does NOT:
1. recognise the Tillamook-core correctly, and
2. work fine with two Tillamook-cores together (as assumed in other threads, the Tillamook-core is maybe not SMP-able by itself)

So, for my point of view, I will no longer try to adopt two Tillamook-CPUs onto this board. Sorry - simply no chance... 🙁

Seems to me you have two issues:
1. Your motherboard doesn't have the correct microcode for the Tillamook, but it should be fixable in a similar manner to the LGA771 Xeon mods
2. You might need some more pins to be bridged to re-enable SMP, or it might also be fixed by modding your BIOS with the correct microcodes.

a jan patch should clear up the identification issue.