wrote:One thing Ive noticed is that when someone builds a high end hot rod 486 its usually based on a PCI board which are not so common.
The reason is that some of them (UMC chipset boards) can use a 50, 60 or 66 mhz bus. That can make a huge difference in performance over a 33 or 40mhz bus that most other 486 boards max out at. Sadly, the UMC chipset used in VL bus boards is not the same.
wrote:now this may sound stupid but how does one go about checking WB? cachechk? I have it set to WB in BIOS and it seems to be posting just fine. haven't had a chance to put it in a proper case and load an OS yet.
Check it with speedsys. You can even see which cache is affected when you turn on WB.
wrote:wrote:One thing Ive noticed is that when someone builds a high end hot rod 486 its usually based on a PCI board which are not so common.
The reason is that some of them (UMC chipset boards) can use a 50, 60 or 66 mhz bus. That can make a huge difference in performance over a 33 or 40mhz bus that most other 486 boards max out at. Sadly, the UMC chipset used in VL bus boards is not the same.
What sliderider forgets to mention (beside undocumented, potentially unstable features) on some boards is that PCI has the following benefits for people building their 486 systems:
- PCI graphic cards are much more common and cheaper to buy than VLB
- You can install a 3DFX voodoo
- You can use S-ATA, ATA-133 and SCSI PCI controllers which opens up alot more storage options.
- 486 PCI cards often offer PS/2 mouse connector & coin battery. Some VLB boards do as well, but it isn't as common.
But don't be afraid to go for a high end VLB 486. My VLB system is as of this writing still on top of the socket 3 systems in Phil's VGA Benchmark database.
well....It may be a minute before I can test the board because its really ticking me off. had some time after work so I started messing with it, which may actually be my problem half the time. Usually the only time I have to mess with my old PC's is late at night and when I do and I hit a problem I want to solve it so I end up being up till 4AM and more and more frustrated. anyways I think the other half of the problem is I don't really have a proper case to put the board in...but anyways
The VLB slots work. after reseating the cards and hitting DEL on boot as suggested I get video with my Speedster but not the S3 so the S3 may be dead which is a possibility since I found it on a shelf at goodwill.
the frustrating part is the system is very unstable. I tried 2 different ISA ide/fdd controller and a SCSI card with a floppy controller on it and I cant get it read anything. the fdd controller on the scsi card starts to load ms-dos from floppy but then resets. The only controller I had luck with was a generic VLB ide/fdd controller. but about 75% of the time I get all kind of corruptions and errors as it boots DOS. I can confirm my fdd drive and DOS disk is good. sometimes I just get odd symbols and it locks up. sometimes I get it telling me about corrupted files. once I got a "divide overload" error. every once in awhile It will boot into DOS from the disk but its none to stable. I tried to run speedsys a few times to see what it says but its always the same thing. Its reports my 486 is running at 22mhz (this contradicts the post which states 66mhz) and I HAVE double checked my jumpers then when speedsys gets to completing the little cpu speed bar at the bottom the program crashes.
I really wish I didn't sell my extra AT cases a few years back....
so.....blah!
wrote:My VLB system is as of this writing still on top of the socket 3 systems in Phil's VGA Benchmark database.
Of course it is, POD is a remarkable CPU and runs only at 33 MHz bus not stressing the motherboard and other components much. The next hot-rod VLB system (486DX4-100) holds the 15th place in 486 class and belongs to Kixs.
wrote:wrote:My VLB system is as of this writing still on top of the socket 3 systems in Phil's VGA Benchmark database.
Of course it is, POD is a remarkable CPU and runs only at 33 MHz bus not stressing the motherboard and other components much. The next hot-rod VLB system (486DX4-100) holds the 15th place in 486 class and belongs to Kixs.
You're the first one to say that tbh. The general consensus about the POD is that it is slower than AMD 5x86 X5-133 and Cyrix 5x86's, which also Feipoa confirms in his 486 benchmark on a PCI motherboard. It seems that in a VLB board the POD really takes off giving better performance compared to a X5-133 or POD in a PCI board.
Even Wikipedia says the same thing:
The AMD 5x86 and Cyrix Cx5x86 processors ran at higher clock frequencies, scored higher on many benchmarks and were considerably cheaper
hmmm? so in a PCI 486 system the Cyrix or AMD scores better then the POD but in a VLB system the POD is generally superior?
wrote:hmmm? so in a PCI 486 system the Cyrix or AMD scores better then the POD but in a VLB system the POD is generally superior?
I don't have any data/benches other than my own to confirm this. Cyrix and AMD scores better in ALU performance compared the the POD on PCI systems, but ofc when it comes to FPU there is no competition in favor of the POD.
wrote:You're the first one to say that tbh. The general consensus about the POD is that it is slower than AMD 5x86 X5-133 and Cyrix 5x […]
wrote:wrote:My VLB system is as of this writing still on top of the socket 3 systems in Phil's VGA Benchmark database.
Of course it is, POD is a remarkable CPU and runs only at 33 MHz bus not stressing the motherboard and other components much. The next hot-rod VLB system (486DX4-100) holds the 15th place in 486 class and belongs to Kixs.
You're the first one to say that tbh. The general consensus about the POD is that it is slower than AMD 5x86 X5-133 and Cyrix 5x86's, which also Feipoa confirms in his 486 benchmark on a PCI motherboard. It seems that in a VLB board the POD really takes off giving better performance compared to a X5-133 or POD in a PCI board.
Even Wikipedia says the same thing:
The AMD 5x86 and Cyrix Cx5x86 processors ran at higher clock frequencies, scored higher on many benchmarks and were considerably cheaper
Wow! That's interesting. I thought it had a far better Pentium-class FPU to begin with. I must say, I'm a bit confused because I tought POD and Cyrix 5x86 were two ultimate 486 CPUs, the first having Peitium-class FPU (look at that Quake score!) and the second having ultimate ALU.
I wonder what makes POD shine in a VLB system... I wish we had more POD benchmark results. btw, with that giant cache and professional videocard your system might be the best 486 on this forum 😀
The bus and the overall clock speed of the POD hold it back. The POD is also communicating with memory across a bus that wasn't designed for a Pentium.
