This Strange 486 Motherboard Boots Amazingly Fast!
Vložit
- čas přidán 3. 06. 2024
- Tired of those 'plain' 486 motherboard with VIA, SiS, ALi or UMC chipset and Award/AMI BIOS? How about the Octek 486 with one amazing feature: lightning fast boot time! Plus a VL card that combines graphics, HDD controller, FDD controller, serial/parallel controller and joystick controller!
Timestamps:
00:00 Introduction
01:35 Boot!
05:24 The VL-COMBO card
07:06 How were these board sold? - Věda a technologie
Holy moly now that IS a fast 486 post...
Super speedy , very interesting for embedded applications
Ooh, fancy! My 486 is only a lowly AMD DX-40.
Maybe 80 is like an abbreviated I/O address for the IDE controller or drive? It'd be unusually high since I only know non-RAID IDE controllers to end their I/O space at 3F7h and not, say 3F8 or 8h, but maybe its not unheard of.
80 is most likely the BIOS interrupt 13h disk number. Floppy disk start at 00 and fixed disk start at 80.
It's all thanks to the MR BIOS on it (i recognize that boot sequence), if you had the same board with the AWARD BIOS, it would be as slow as other Mainboards
Also, it would have booted up faster, the only slow down you really had, was the HDD spin up which wasn't fast enough for the BIOS! lol
Yes, I LOVE those MR BIOSes. They are not only fast but also very powerful. I wish we had more motherboards with MR BIOS on them.
Impressive. Does it skip any form of POST?
i beclieve it hid some simple POST process behind that spinup prompt, yet still the time is not enough for a 'normal' BIOS to finish the POST process, meaning Octek may had to cut some corners.
There is a lot of hardware like this at a higher price, and a thing about that 'spinning up' that really comes to no POST, and such. This hardware using being more reliable and such can go without a lot of common POST stuff. There is also a lot about DOS and this does not have much in autoexec.bat and himem.sys in terms of OS. But yeah, for a BIOS it is quick, and also a lot of the slowness isn't complexity it's timers so ordinary users can just see things. This was over 20 years ago for me, and I had this kind of hardware. There is a lot more than just complexity in the 486 architecture at play here. I can also literally see it on the screen because of the standard at the time.
Are you suggesting that the board is just waiting instead of POSTing in the background? That is cool but also a problem to me. Since when some part eventually break down, this no-POST booting will make troubleshooting harder. And you did spot that I disabled the CONFIG.SYS by renaming it, just to skip the HIMEM.SYS part of the boot process. I know that a lot of slowness is added intentionally (and will explain it further in a future video). For example, Trident cards display a VBIOS message for extended amount of time just to make sure that the user's CRT light up enough to display that a Trident card is installed. And because Octek builds both the motherboard and graphics card, there is less need to make yourself known but more to streamline the experience for the user. (I don't quite understand your last sentence,. though)
@@andyhu9542 It has been a long time since I've used hardware like this. There might be a motherboard jumper to do more testing. There are ISA and PCI diagnostic boards than can reveal more issues about the system than the POST anyway.
Try this with an IDE to CF card adapter, I want to see how fast it is without having to wait for the hard disk to spin up!
That's a bit of a problem... my adapter doesn't work with 486 motherboards.
@@andyhu9542 Oh, that's a pity. But if the Bios supports LBA, maybe you can use a SATA to IDE adapter and use a SSD?
I wonder if the board checks the IDE status register bit 6 or if the spinup is just hardcoded. If it checks that, then a CF card might truly give a stupidly fast boot on this board ;-)
Well, until you show me a Promise VLB IDE Cached controller (which had a dedicated 80286 for that job) with at least 8Mb of RAM actually loading Windows 95 faster than a pentium 133 on a DX50 or DX/2 80 (yes not DX4 FSB Must be 40Mhz or 50Mhz in to see that controller fly)
.. Until then you won't know what a Fast 486 is. Of course you must also have at least the DOS drivers or make sure you get the version of the controller that comes with the Bios extensions chip. For Windows, Netware or OS/2 you still need drivers though but if the chip is present it will almost not be important performance wise (we'll won't be important on DOS that is elsewhere the advantages are important)
This will give you the fastest IDE 486 EVER. Trust me, I was building servers for "high availability /performance" Netware FileServers for accounting databases on multiuser environments.
This kicks some serious butt for the age. Some people wouldn't believe this board with ram maxed out did to 486 performance what it did.
Only when the K6 and Pentium Pro/2 came out which had a lot more cpu ram to use as cache did performance began to shift to newer computers vs that old server. And even then I bet if we put some CF cards adapted to IDE on it the thing will finally reach its max performance. Old hdds couldn't actually allow this controller to fully shine on raw sequencial, only random shared access was king, but with CFs I'm betting it can hit the theoretical limit per channel of 100Mb/s non cached, for cached it always got close to the ram max throughput anyhow.
I think you need to turn off auto-focus and just fix it at a spot (like the motherboard). Other than that, the video's fine.
I forgot to do that when filming. It dodn't become apparent until post-editing.