91
Windows 98 (Jun 1998) / Re: WINDOWS 95/98/SE/ME ram limitation patch (2012)
« Last post by dawful on December 24, 2023, 01:15:06 AM »No. It is pretty wasteful, to use Win9x on a multi-core machine. But it is still faster, then an older machine (depending).
About 2012, I had seen mention of someone accomplishing it. I had long left Win98 behind. So, I never followed it up. It was just a screen screenshot, of a dual CPU meter. Could have been completely bogus. A couple of years ago, when I started messing this it again, it seems like no such think could have ever been. If it had been done, maybe it was really buggy. Just a proof, not something actually usable.
I know there have been source leaks, but most usable stuff has been NT era. Without Win9x leaks, the NT code would be really hard to make use of (for Win9x). And it would still be a really long road. There are some tweaked files, built from SDK example code, stolen from Win Betas, and binary hacked. But I've not seen proof, of multi-core/SMP support.
I've always wondered if such a thing could be handled at the BIOS level. Let the hardware split up the load, so you aren't spending cycles deciding how to.
However, someone could write a Daw, VST/VSTi host, or application wrapper to run on a core of choice.
-> https://rloewelectronics.com/Programs/DEMOCORE.ZIP I'd prefer that anyway. That is how I ran Net/FreeBSD. Each core was like it's own computer.
About 2012, I had seen mention of someone accomplishing it. I had long left Win98 behind. So, I never followed it up. It was just a screen screenshot, of a dual CPU meter. Could have been completely bogus. A couple of years ago, when I started messing this it again, it seems like no such think could have ever been. If it had been done, maybe it was really buggy. Just a proof, not something actually usable.
I know there have been source leaks, but most usable stuff has been NT era. Without Win9x leaks, the NT code would be really hard to make use of (for Win9x). And it would still be a really long road. There are some tweaked files, built from SDK example code, stolen from Win Betas, and binary hacked. But I've not seen proof, of multi-core/SMP support.
I've always wondered if such a thing could be handled at the BIOS level. Let the hardware split up the load, so you aren't spending cycles deciding how to.
However, someone could write a Daw, VST/VSTi host, or application wrapper to run on a core of choice.
-> https://rloewelectronics.com/Programs/DEMOCORE.ZIP I'd prefer that anyway. That is how I ran Net/FreeBSD. Each core was like it's own computer.