I use to have a Pentium 133mhz build with 16mb ram and 250mb hard-drive.
It ran WinNT 4.0 just fine. WinNT 4.0 Should be fine to get your your Adaptec SCSI controller and ethernet cards working.
I used mine with a US Robotics 33.3 modem and Network card.
Will run Netscape web browser just fine.
Do you happen to have the EISA configuration utility for this board? That was my point of posting here. Without that I won't be able to use EISA cards.
I/O, I/O,
It's off to disk I go,
With a bit and a byte
And a read and a write,
I/O, I/O
Not sure if it applies to you, but some of the links from here
https://web.archive.org/web/19980128144503/ht … rboards/server/
also refer to a 1.00.18.BI0 and 1.00.19.BI0 with SCU v3.40 & overlay 1.30A, though unsurprisingly all the actual file links are dead. I did see this alternate site for these, though not sure if you'd want to risk it
https://fr.pcdrivers.guru/intel-altserver-ser … 89519.html#os60
wrote:Not sure if it applies to you, but some of the links from here […]
Not sure if it applies to you, but some of the links from here
https://web.archive.org/web/19980128144503/ht … rboards/server/
also refer to a 1.00.18.BI0 and 1.00.19.BI0 with SCU v3.40 & overlay 1.30A, though unsurprisingly all the actual file links are dead. I did see this alternate site for these, though not sure if you'd want to risk it
https://fr.pcdrivers.guru/intel-altserver-ser … 89519.html#os60
Interesting, and surprisingly enough - these seem to be legit. There is a lot of fluff on that page, drivers for PCI-E cards and for Windows 8/10 which this system never supported, but 3 of the files appear to be the real deal:
10018bi0.exe - I renamed this to .zip, extracted it, it contains another autoextracting archive BIOS.exe which was again renamed to .zip and extracted, and it appears to contain the files for BIOS 1.00.18.bi0, flash utilities and some pretty detailed notes on how to flash from different previous revisions of the BIOS, a bugfix list etc.
10019bi0.exe - similarly renamed to zip and extracted, and the SW.exe contained within renamed and extracted, seems indeed to contain the files for BIOS 1.00.19.bi0
s340130a.exe - contains 3 inner layers of self-extracting archives (each renamed and extracted with 7Zip), eventually you get to the files for SCU 3.40 and overlay 1.30 which are the appropriate ones for BIOS 18 and 19.
I can't try them as I don't have the board yet, but I will repack them in a non-executable format and attach them to this post for easy access without jumping through hoops on suspect servers (which may disappear tomorrow).
Thank you for the links above, they proved useful.
BTW anyone downloading this, you shouldn't trust me and test the files by yourself if you intend to use them.
Also, the only real difference between BIOS 18 and 19 is a fix for the correct display of Feb 29th on leap years beyond 2000.
I/O, I/O,
It's off to disk I go,
With a bit and a byte
And a read and a write,
I/O, I/O
If either of you guys still have this board I’m wondering if it may be one of the older ones that may work with this.
(If it’s even true) which I’m hoping one of you guys may be able to test.
Sphere478 wrote on 2021-01-17, 04:40:If either of you guys still have this board I’m wondering if it may be one of the older ones that may work with this.
(If it’s even true) which I’m hoping one of you guys may be able to test.
Not totally convinced about the veracity of that 2001 claim, but that aside, it referred to boards with their own SMP glue logic. This is a plain i430NX board, first of the "wire CPUs in parallel and let them sort it out themselves" solutions. Not "one of the older ones".
Oh okay. How do I spot the ones we are talking about?
dionb wrote on 2021-01-17, 10:54:Sphere478 wrote on 2021-01-17, 04:40:If either of you guys still have this board I’m wondering if it may be one of the older ones that may work with this.
(If it’s even true) which I’m hoping one of you guys may be able to test.
Not totally convinced about the veracity of that 2001 claim, but that aside, it referred to boards with their own SMP glue logic. This is a plain i430NX board, first of the "wire CPUs in parallel and let them sort it out themselves" solutions. Not "one of the older ones".
Sphere478 wrote on 2021-01-17, 15:55:Oh okay. How do I spot the ones we are talking about?
Turn it round: i430NX and i430HX are the only two So5/7 chipsets that do SMP via the onboard CPU logic, so pretty much anything that does have So5 but doesn't have i430NX or HX would be a good candidate.
dionb wrote on 2021-01-17, 21:08:Sphere478 wrote on 2021-01-17, 15:55:Oh okay. How do I spot the ones we are talking about?
Turn it round: i430NX and i430HX are the only two So5/7 chipsets that do SMP via the onboard CPU logic, so pretty much anything that does have So5 but doesn't have i430NX or HX would be a good candidate.
thanks! 😀
Any of those in this thread? Or anyone around here have one that you know of?
Sphere478 wrote on 2021-01-17, 21:13:[...]
Any of those in this thread? Or anyone around here have one that you know of?
Been digging around and found at least one positive target for you to hunt for:
HP NetServer 5 LS series.
They're available as LS, LS2 and LS4.
That "LS4" means quad-CPU, which is beyond APIC, so the system must use its own logic. And it does:
https://www.itprotoday.com/compute-engines/hp … server-5166-ls4
HP's proprietary Extended Express Bus, a high-speed, scaleable bus for SMP
Now, no idea if you can even get one K6 working on a beast as strange as that with single CPU with appropriate interposer, but if you can, you could see what happens with two on either end of that EEB.
dionb wrote on 2021-01-17, 21:33:Been digging around and found at least one positive target for you to hunt for: […]
Sphere478 wrote on 2021-01-17, 21:13:[...]
Any of those in this thread? Or anyone around here have one that you know of?
Been digging around and found at least one positive target for you to hunt for:
HP NetServer 5 LS series.
They're available as LS, LS2 and LS4.
That "LS4" means quad-CPU, which is beyond APIC, so the system must use its own logic. And it does:
https://www.itprotoday.com/compute-engines/hp … server-5166-ls4HP's proprietary Extended Express Bus, a high-speed, scaleable bus for SMP
Now, no idea if you can even get one K6 working on a beast as strange as that with single CPU with appropriate interposer, but if you can, you could see what happens with two on either end of that EEB.
Nice find. A quad k6 would be amazing hahaha!
I can’t even find a picture of the board. Probably some weird form factor also 🙁
Sphere478 wrote on 2021-01-17, 21:47:[...]
Nice find. A quad k6 would be amazing hahaha!
I can’t even find a picture of the board. Probably some weird form factor also 🙁
Definitely. Don't expect this sort of stuff on (Baby)AT, this is the kind of things on multiple cards in stupendously heavy server towers.
Quad K6 isn't going to work, as each CPU card on this machine can take two CPUs and runs those with APIC, so if the assumption about K6 SMP is valid, you can only run one per card. So you need an LM4 with two dual CPU cards, and then insert a single K6 on each.
