VOGONS


Evolution of a Socket3 System to a POD @100MHz

Topic actions

Reply 60 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member

@feipoa
I have some additional questions to your 486 Comparison and the 3D Games Comparision

  • In your 3D tests you used for the 40MHz FSB CPUs a PCI speed of 27,5MHz (post #56) and 40MHz (post #71), was the PCI @40MHz stable with complete testing config (Graphics + 3Dfx + IDE + ...)?
  • In the 486 comp. all 40MHz FSB tests were tested with 20MHz PCI speed. Hurts to see like the 20% overclocked CPUs perform often slower than the originals in the DOS applications. Why did you chose the M919 for your tests (3D tests too) and not in general the MB8433 or one of your faster SiS496 boards - 66MHz support?
  • I compared the (in general low) values for the PCPbench (640x400x8 LFB) with my results with the HOT433 (same chipset, other graphics card). Seems, that the chipset has an issue with the LFB. Could you find the rootcause for that?
  • Which 486 PCI board is your personal favorite and why?

I am very interested in the backgrounds.

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 61 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++
PC-Engineer wrote:

Seems, that my diagram suggest a ranking. It is difficult to reed between the lines. The intention was to list the absolute results for direct compairison of individual values from other users.

I think the majority of the people who see this thread will skim it, look at the charts, and move on. Since the CPUs on the chart are ordered from lowest sum of results to highest sum, people could draw the wrong conclusion and rank the CPUs in this fashion. Perhaps if the CPUs weren't listed in descending order, but were more random, it would be less likely that someone could misinterpret the results. I also appreciate the value in the raw data, so I usually include normalised and non-normalised data. The normalised chart is a lot more representative of total performance for these four benchmark condition; however, it is still not a fair overall comparison because results from PCPbench are tallied twice, allowing for PCPBench to hold twice the importance of Quake or DOOM.

Interesting that PCPBench is 10% faster with the i486DX-120 compared to the cx5x86-120. My previous results indicate only a 1-2% lead for the i486DX-120 at 640x480x8.

PC-Engineer wrote:
CPU scaling via overclocking: 33MHz - 40MHz FSB (+20%) intel DX2: +19,5% intel P24T: +19,5% intel DX4: +19,4% Cyrix 5x86: +19,2% […]
Show full quote

CPU scaling via overclocking: 33MHz - 40MHz FSB (+20%)
intel DX2: +19,5%
intel P24T: +19,5%
intel DX4: +19,4%
Cyrix 5x86: +19,2%
AMD DX4: +16,6%
AMD Am5x86: +13,5%

The scaling issue with the Am5x86 appears to be due to your DOOM results. I wonder if you jotted down the value correctly? I got closer to a 20% increase in my previous tests. Your results show only a 4% inrease in DOOM from 133 MHz to 160 MHz.

PC-Engineer wrote:

I got a 2842A and and replaced the 2842VL. The Jumper of the 2842A was allways set. Do you have any indication that the jumper has a negative influence?

It has been so long since I setup this system, but I don't think I did much analysis on this.

PC-Engineer wrote:

I thought this would be a wait-state jumper, but without a performance drop (only measured in transparent mode) by toggeling. I beleave it must have another function (interrupt?). Couldn't found a hint for it until now. And no, i didn't consider the synchronize mode in the comparison. Do you have an expection?

Could be the IRQ 2/9 jumper. Some cards have two jumpers, one for IRQ, the other for 0 WS.

When using a Trio64 on this motherboard with DOOM, transparent yielded a 6.3% improvement (61.7 fps vs. 65.6 fps). With the Mach64, there was a 12.1% improvement (58.44 fps vs. 65.50 fps). I use an S3 Vision968 and must use synchronise as the system won't boot with transparent. For comparison, synchronise yields ony 58.4 fps. If the Trio64 came in 4 MB, I'd probably use that card on this system.

Are you able to measure the difference in synchronise vs. transparent with your graphics card in this board?

PC-Engineer wrote:

In your 3D tests you used for the 40MHz FSB CPUs a PCI speed of 27,5MHz (post #56) and 40MHz (post #71), was the PCI @40MHz stable with complete testing config (Graphics + 3Dfx + IDE + ...)?

Could you paste vogons links to the post? I'm not sure how to view post #56 and #71. For the Ultimate 486 Benchmark Comparison, when the FSB was set to 40 MHz, the PCI bus was automatically set to 26.7 MHz. Unfortunately, the PC Chips m919 sets the PCI bus automatically. I didn't realise this when I started benchmarking. It is the only 486 board I've experienced which does this. I used a different motherboard for 486 CPUs in the Ultimate 686 Benchmark Comparison so that the PCI bus was kept at 40 MHz. Also, the latest benchmarks using the Voodoo on a socket 3 have the PCI bus at 40 MHz: Re: Voodoo 1 vs. Voodoo 2 on a 486 The hardware used in that thread is listed here: Re: Voodoo 1 vs. Voodoo 2 on a 486

For the Biostar MB-8433UUD and the hardware selected, I did not experience any issue with the PCI bus at 40 MHz. I have noticed that many PCI-based socket 3 systems do not function well with a PCI network card, the MB-8433UUD being the exception.

PC-Engineer wrote:

In the 486 comp. all 40MHz FSB tests were tested with 20MHz PCI speed. Hurts to see like the 20% overclocked CPUs perform often slower than the originals in the DOS applications. Why did you chose the M919 for your tests (3D tests too) and not in general the MB8433 or one of your faster SiS496 boards - 66MHz support?

I figured this out too late. There is no indication in the BIOS that the BIOS is scaling down the PCI bus when a 40 MHz FSB is set. I realised something was odd with the result and discovered the issue after hooking an oscilliscope up to the PCI CLK signal. On my list is to redo the 486 benchmarks with the MB-8433 and a more fine-tuned selection of benchmarks.

I dodn't have a SiS 496 board which can do 66 MHz reliably. Apparently the Lucky Star LS486 can do 66 MHz reliably, but I don't have one. The MB-8433UUD can do 66 MHz reliably.

PC-Engineer wrote:

I compared the (in general low) values for the PCPbench (640x400x8 LFB) with my results with the HOT433 (same chipset, other graphics card). Seems, that the chipset has an issue with the LFB. Could you find the rootcause for that?

I've had about a dozen HOT-433 boards over the years and have always run into some sort of problem with them. I still have two original versions of HOT-433 and two 4.0 boards, but don't care for them at all. The fact that both revisions of this board have a non-functional PS/2 mouse header is a bad sign from the onset.

PC-Engineer wrote:

Which 486 PCI board is your personal favorite and why?

MB-8433. It can be modded to accept 1024K. Has a working PS/2 mouse port. Can boot Windows 2000 without any issue. I could not get the HOT-433 to boot W2K using a PCI SCSI host adapter card. The system would lock-up between HDD access and CD-ROM access. There are some other nice UMC 8881-based boards with 1024K, but they don't have a PS/2 mouse header. For SiS 496 boards, I like the Chaintech-486SPM MSI MS-4144. I have the DTK PKM-0033S setup in a system with a Cyrix 5x86-133/4x; it is a SiS 496 board, but it comes with those darn leaky rechargeable batteries, so by now, most boards for sale will be ruined. I've convered the board to using a native PS/2 mouse and coin cell battery.

A more general answer is that I like any PCI 486 board which has 1024K cache capability and PS/2 capability. In many cases the KBC needs to be modified for PS/2 capability. I do not care much for boards which have PCI+VLB because of bus bridging which reduces performance.

I do not care for boards which only support single-banked L2 cache.

Plan your life wisely, you'll be dead before you know it.

Reply 62 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member

I found the root cause for the worse values of the Am5x86 in Doom (and Quake). If i set JP26 from ND to D then Doom jumps from 57,5 to 64,3 fps (Quake from 17,2 to 17,5). So the performance increase (for the four Benchmarks) from 133MHz to 160MHz jumps from +13,5 to +18,5, which is a reasonable scaling now.
If i keep JP26 in ND position then, apart from PCP bench Mode 100, the synchronize mode is clear faster than the transparent mode. This effect is only observable with this CPU and only @160MHz. Additional i got two (not reproducable) crashes while the short benchmark session with JP26 in D position and transparent mode, one crash took place while leaving the editor in Norton. With the ET4000W32 i got similar benchmark values and no crashes (while this short usage), but this card din't post with JP26 in ND and transparent mode. Toggeling this jumper in synchr. mode showed no influence to the benchmark values. Maybe toggeling JP26 could be a solution for your problems with your S3 968 in transparent mode.

I updated my previous posts and the diagrams with the new findings.

Additionally i did benchmarks for the comparison with trnasparent and synchronize mode:

The attachment Synchr. vs Transp.JPG is no longer available

Doom: ----------------- POD +7,4% / Am5x86 +4,8%
Quake: ----------------- POD +2,3% / Am5x86 +0%
PCPBench SVGA: ------- POD +7,6% / Am5x86 +7,9%

feipoa wrote:

I've had about a dozen HOT-433 boards over the years and have always run into some sort of problem with them. I still have two original versions of HOT-433 and two 4.0 boards, but don't care for them at all. The fact that both revisions of this board have a non-functional PS/2 mouse header is a bad sign from the onset.

Understand, but i didn't try the PS2 port until now on my HOT433 and i have no need to install Windows 2000 😀 But appart from this two issues this is a decent board with a (online) support page at the supplier. The 486SPM is very decent too, but my board needs help to support PS2 - i read your tutorial.
I was wondering that you got your MB-8433 running stable with 66MHz and the cache hack because of EMC and signal quality of your wiring - well done!
In general i cannot understand why many user favorite the M919.

Thanks a lot for the explanations!

PS: i had bad experiances with single banked cache too, especially the chinese ISSI chips seems to be alergical against single banked mode (also with 33MHz)

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 63 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I've made the cache hack jumperable now, so I can more easily switch between 1024K and 512K double-banked modes. For 66 MHz FSB, I went back to using 256K because memory read was more reliable with only 1 WS. For 1024K, I had to use 2 WS for the level of stability I desire. What I would like to do is fine a way to do your dirty TAG modification so that I can use 256K in write-back mode and cache all the 64 MB of memory. But I don't have the UMC 8881 pinouts.

It is not just W2K that the HOT-433 has trouble with. Every now and then I would see the same symptoms on boot with NT4. I'm guessing PCI bus mastering wasn't properly implemented for host controller cards. How long have you been using the HOT-433? I thought it was a good board in the beginning too, but not after years of testing.

Single banked cache requires more wait states than double-banked for 40 MHz. Some older motherboard manuals identify this in their "auto" mode for memory in the BIOS.

Looks like your percent increase for transparent mode falls in like with my results of between 6-12%. Unfortunately, I have a hand written note in my manual that I must set the VESA clock delay (JP26) to "Delay" rather than "No Delay". And based on your tests, you mentioned that "Delay" and "Transparent" caused some irregular crashes. I could always test it again though.

Plan your life wisely, you'll be dead before you know it.

Reply 64 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member

I keep my eyes open for the needed pins of UMC8881.

The HOT433 was my first 486 retro rebuildt. I needed a PCI board because i wanted to use it in my portable housing (coming soon with 486 DX50 ISA) without possibility of long cards. As i got the Highscreen Colani housing i was able to build a VLB system and the priorities changed.

My first config was the HOT433 + Adaptec 2940U (only for HDD - CD-ROM on IDE because the SCSI versions were too long) + MGA-2064W. I only used DOS 6.22 and Win3.1 and played with it around nearly a half year. I think the HDD usage of Win 95 and Win NT is much higher than in DOS, so my usage behavior is not representative. Shuttle is aware of the problems with SCSI. In their FAQ they recommend to use PCI SCSI controller only in PCI slot 1 (near ISA slot).

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 65 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++
PC-Engineer wrote:

Shuttle is aware of the problems with SCSI. In their FAQ they recommend to use PCI SCSI controller only in PCI slot 1 (near ISA slot).

PCI slot 1 is normally where I place the graphics card so it can have some room to breath if I add a fan to the graphics heatsink.

Plan your life wisely, you'll be dead before you know it.

Reply 66 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Unfortunately my the system would not post with NO DELAY and TRANSPARENT enabled.

Plan your life wisely, you'll be dead before you know it.

Reply 67 of 83, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote:
PC-Engineer wrote:

Shuttle is aware of the problems with SCSI. In their FAQ they recommend to use PCI SCSI controller only in PCI slot 1 (near ISA slot).

PCI slot 1 is normally where I place the graphics card so it can have some room to breath if I add a fan to the graphics heatsink.

Hi,
I've been working with a Shuttle HOT-433 v1-3 without PS/2 port since a few weeks.
I'm using a HighPoint 370A IDE controller to avoid PIO transfers.
I faced a few problems, when installing Windows 2000 (with Smart Boot Manager and cdrom connected to internal IDE port).
1) Windows 2000 failed when recognizing keyboard controller with 433AIP16 BIOS.
Solution: BIOS upgrade from 433AIP16 to 433AUS2C.
2) F6 disks worked in first step, but failed when copying the files.
Solution: Setting printer port back to ECP. Without, Windows assigns floppy DMA channel to paralell port (in SPP mode)!
3) Windows 2000 hard disk access has been halted and probably resetted for several seconds. Several entries in eventlog. Same with other IDE and SCSI controllers. No change when placing in other PCI slot.
Solution: Assign IRQ 12 to controller.
4) No way to install a PCI network card without removing the PCI mass storage controller.
Solution: Use ISA NIC.
5) When copying SP4 to the HDD and doing a verify over network, 4-byte-transfer errors occured by random but with system: address: 0x*3F? 0x7F? 0xBF? 0xFF? the integrients were replaced by first DWORD from following 16-byte-block (=cacheline): for example 0x*3F0-0x*3F3 by 0x*400-0x*403 or 0x*3F8-0x*3FB by 0x*400-0x*403 and so on... No difference with L2 256K 2banked, 256K 1banked, 512K 1banked, 1024K 2banked. No difference with 32M, 64M or 128M RAM (cacheable area respected).
Solution: Switching L2 from WB to WT.
6) System does not read a boot sector at 40 MHz FSB. Even with slowest settings, even with both caches disabled, even from floppy.
Solution: 33 MHz.

Do you have any suggestions or similar problems?

I'm sad, I cannot make my HOT-433 faster.
I don't even know what happens when I try to add a PS/2 mouse port because of the IRQ 12 issue.

Last edited by Disruptor on 2019-12-16, 22:49. Edited 1 time in total.

Reply 68 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Your problems with the HOT-433 sound very similar to issues I've experience with this board over the years. My conclusion every time I think to give this board a second chance was 'don't bother'. This board is just riddled with problems.

Did you try FPM or EDO RAM? Of all the PCI 486 PCI boards I've tested, I recall this one liking EDO memory.

I am curious if the Biostar MB-8433 BIOS would work properly on this board to solve some of the problems.

I don't think you will get a PS/2 mouse working on this board. I suspect that the BF suffix on the UMC chipset is needed, whereas this board is AF. Note that the keyboard controller for the UMC 8881/8886 chipset is in the chipset.

Plan your life wisely, you'll be dead before you know it.

Reply 69 of 83, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie

DRAM is FPM.

My HOT-433 has 8881F and 8886AF and IO 8663AF.
When I try to start modifications, I'll have to use an 8742 and to populate the socket 7406 (on the rear, between ISA slot 2 and 3) and to do some modifications on the wiring. There is a way to override the internal KBC. And I will have to try BIOS 433AUS26 that has PS/2 Mouse support - I have noticed that BIOS version 443AUS26 does NOT support keyboards on internal KBC!

I have here a GA486IM too which has a 8881F and 8886F (without A) - perhaps I'll give this board a chance.

Reply 70 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Someone on this forum has already tried the using a DIP-40 KBC and 7406 inverter, but the ps/2 mouse didn't work. When I measured this board a few years ago I determined that it is not properly wired for using a PS/2 mouse with the DIP KBC, so as you say, you'll need to do some wiring modifications.

I didn't notice that AUS26 didn't support the chipset KBC - I'll have to check that again.

Have you tried EDO?

When do you think you'll get around to the PS/2 wiring?

Plan your life wisely, you'll be dead before you know it.

Reply 71 of 83, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote:
Someone on this forum has already tried the using a DIP-40 KBC and 7406 inverter, but the ps/2 mouse didn't work. When I measur […]
Show full quote

Someone on this forum has already tried the using a DIP-40 KBC and 7406 inverter, but the ps/2 mouse didn't work. When I measured this board a few years ago I determined that it is not properly wired for using a PS/2 mouse with the DIP KBC, so as you say, you'll need to do some wiring modifications.

I didn't notice that AUS26 didn't support the chipset KBC - I'll have to check that again.

Have you tried EDO?

When do you think you'll get around to the PS/2 wiring?

Be sure to have a boot disk that can flash another BIOS before doing this 😉
I'll try EDO somewhen later this week.
I'll try to work on the wiring issue after xmas.

Reply 72 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

OK. Probably best to start a new thread.
I usually flash BIOSes in an external programmer.

Plan your life wisely, you'll be dead before you know it.

Reply 73 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member

I finally managed to ran the Winstone 95 benchmark in Win 3.11 and install Windows 95 (with Winstone 96 benchmark) and discovered several issues with the system. The conclusion in advance, all findings have their root cause in the VLB-System.

The first Setting:

  • 1st VLB Card (Slot U1): Graphics Card - Miro Crystal 20SD (earlier version)
  • 2nd VLB Card (Slot U2): SCSI Controller – Adaptec AHA-2842A (later version)
  • JP26 (VESA Clock Selection): No Delay
  • VL Mode in BIOS: transparent

Testing with Windows 3.11 and Winstone 95
The First problem I discovered was freezing of graphics on Windows startup in some resolutions (800x600x8 and 640x480x24), so that I assumed a driver issue, but it wasn’t. After changing the 2842A against the older version 2842VL, the problems while startup disappear. So I installed WS95 and ran the benchmark, but it crashed in different positions with an error message. Toggeling the JP26, changed something (later crashes, but system sporadically didn’t post) but doesn’t made the system stable.
Next I switched the VLB Ports of graphics card and SCSI controller, which causes that the system wasn’t able to boot up when it was warm, but I was able to complete the WS95 first time.
An other trigger was the VL Mode in the BIOS. Setting it from transparent to synchronize mode lead to running Winstone 95 stable but didn’t solve the Posting problem.
Setting the FSB from 40MHz to 33MHz solved all Problems with Winstone 95, Posting and Windows startup.
So I changed the AHA-2842VL against an ISA version, the AHA-1542CP, and the system kept stable with 40MHz like with 33MHz FSB. The result from WS95 decreased from 141 to 131 overall score.
The next change was the graphics card to a Spea V7 Mirage P64 VL (Trio64) which healed the problems with posting and Windows startup, but didn’t solve the sporadic crashes with Winstone.
Ok, at this point I gave up for next days and kept the Mirage P64, the AHA-2842VL and the VL mode in synchronize mode. I was going to install Windows 95 and do a comparison between the Am5x86 and the POD with Winstone 96.

The second Setting:

  • 1st VLB Card (Slot U1): Graphics Card – Spea V7 Mirage P64/VL (later version)
  • 2nd VLB Card (Slot U2): SCSI Controller – Adaptec AHA-2842VL (earlier version)
  • JP26 (VESA Clock Selection): No Delay
  • VL Mode in BIOS: synchronize

Testing with Windows 95 and Winstone 96
With the second setting I had no Problems with installing Windows 95 and running Winstone 96 (800x600x8bpp) with the Pentium Overdrive @100MHz. I installed the system full equipped with both soundcards and Ethernet. Additional to the Windows 95 basic drivers i installed DirectX 5.0 and the GUS drivers.
I Addition to the POD I compared the performance of the High-End 486 CPUs Am5x86 @160MHz, iDX4 @100MHz and iDX4 @120MHz. With the change to the Am5x86 I had to switch the SCSI controller to the 2842A and set the J5 on it to get a successful post with this CPU and L1:WB. For the Am5x86 I had to switch the JP26 to Delay.
Under Windows95 and WS96 the Am5x86 takes the lead by 7% distance to the POD followed by the iDX4 with additional 5% difference. I am very impressed about the performance of the 486 architecture in Win95 environment in comparison to the Pentium – in the same infrastructure. Ok, the POD drives the L1 Cache only in WT mode (in this system/comparison), but this would give maybe additional 5% to the Pentium. In general the 20% higher clock (of overclocked iDX4 and POD) leads to appr. 16% higher application performance. If I compare my values with values of regular Pentium Systems, tested by the "PC Magazine" in December of 1995, my system with all three overclocked CPUs, is full on eye level with midrange to high level Pentium 90 – Pentium 100 (Triton) Systems with 16-32MB RAM.
The smaller FSB width of Soccket3 against Socket5 (32bit instead of 64bit) and the lower FSB clock (40MHz vs. 60/66MHz) is well compensated by the larger L1 cache of the Overdrive (2x 16kB vs. 2x 8kB) and the maxed out L2 cache (1024kB vs. 256kB) of my SV2GX4. The 486CPUs compensate their malus with a higher internal clock, very well too.

The attachment Winbench96.JPG is no longer available
The attachment Winstone 96.JPG is no longer available

Conclusion
Driving the FSB on VLB systems over 33MHz (more than specified) with more than one VLB Card can be tricky and needs a lot of different spare parts for testing. Especially the VL special mode “transparent” (from ASUS SV2GX4) takes the bus into a critical zone for high level operations in Windows.
All in all my system fits well into the years 1994 / 1995 / 1996, has enough FPU power for upcoming games like Quake and is easy downgradable by changing the CPU to slow 486 level of 1991 / 1992 / 1993.
AND: it gave/gives me a lot of fun, experiences and memories with tuning, benchmarking and comparing.

The final Setting:

  • 1st VLB Card (Slot U1): Graphics Card – Spea V7 Mirage P64/VL (later version)
  • 2nd VLB Card (Slot U2): SCSI Controller – Adaptec AHA-2842A (later version)
  • JP26 (VESA Clock Selection): Delay
  • VL Mode in BIOS: synchronize

Finally some Numbers from Winbench96 (800x600x8bpp) with my POD @100MHz and VL in synchronize mode.

Graphics WinMark:-----14,3
Disk WinMark:--------1090
CDROM:------------454

PS: In 800x600x16bpp and 1024x768x8bpp the value for WS96 and GWM96 decreses by 2-3%

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 74 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member
Disruptor wrote on 2019-12-16, 15:01:
2) F6 disks worked in first step, but failed when copying the files. Solution: Setting printer port back to ECP. Without, Window […]
Show full quote

2) F6 disks worked in first step, but failed when copying the files.
Solution: Setting printer port back to ECP. Without, Windows assigns floppy DMA channel to paralell port (in SPP mode)!

3) Windows 2000 hard disk access has been halted and probably resetted for several seconds. Several entries in eventlog. Same with other IDE and SCSI controllers. No change when placing in other PCI slot.
Solution: Assign IRQ 12 to controller.

5) When copying SP4 to the HDD and doing a verify over network, 4-byte-transfer errors occured by random but with system: address: 0x*3F? 0x7F? 0xBF? 0xFF? the integrients were replaced by first DWORD from following 16-byte-block (=cacheline): for example 0x*3F0-0x*3F3 by 0x*400-0x*403 or 0x*3F8-0x*3FB by 0x*400-0x*403 and so on... No difference with L2 256K 2banked, 256K 1banked, 512K 1banked, 1024K 2banked. No difference with 32M, 64M or 128M RAM (cacheable area respected).
Solution: Switching L2 from WB to WT.

Do you have any suggestions or similar problems?

All described Problems sound like a conflict with DMA and Write-Back Mode of L1 cache. All actions, you describe, are triggert by a DMA access by a component (Floppy, HDD, Ethernet?). Maybe you are able to to find the correct regigisters for the chipset and update and fix the BIOS. I ran Win3.11 with a Cyrix 5x86 and a POD83 on this board, without overclocking and without any suspicious behavior.

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 75 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Yup, this is what I concluded when I setup my system:

JP26 (VESA Clock Selection): Delay
VL Mode in BIOS: synchronize

To me, test results on a stable system are far more important than faster results on a non-stable system.

Can you remind me if L1 works in WB mode for the POD at 100 MHz when using a VLB SCSI and VLB graphic card?

If you found the ASUS SV2GX4 fun, you'll find the HOT-433 just as fun and quirky. Never found a happy ending for the HOT-433 though, unlike the SV2GX4.

Plan your life wisely, you'll be dead before you know it.

Reply 76 of 83, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie
PC-Engineer wrote on 2019-12-27, 22:57:

All described Problems sound like a conflict with DMA and Write-Back Mode of L1 cache. All actions, you describe, are triggert by a DMA access by a component (Floppy, HDD, Ethernet?). Maybe you are able to to find the correct regigisters for the chipset and update and fix the BIOS. I ran Win3.11 with a Cyrix 5x86 and a POD83 on this board, without overclocking and without any suspicious behavior.

I'm still examining this problem. However, I'll open another thread for it in a few days.
Spoiler: I've tried a BIOS from another board. WB is working with it, but I could not get W2K booting it.

I've considered fixing BIOS. Can you introduce me? Which tools do I need?

Reply 77 of 83, by PC-Engineer

User metadata
Rank Member
Rank
Member

For editing AWARD BIOS you can use MODBIN. Would be helpflull when you tell us which registers in the UMC chipset were responsible for correct L1:WB support.

@Feipoa
Do you know a SIS471 mainboard with correct function of POD with L1:WB? Then i could compare the chipset registers in the BIOS, maybe i am able to find the "little" difference.

Epox 7KXA Slot A / Athlon 950MHz / Voodoo 5 5500 / PowerVR / 512 MB / AWE32 / SCSI - Windows 98SE

Reply 78 of 83, by feipoa

User metadata
Rank l33t++
Rank
l33t++

No I don't, sorry. I thought SV2GX4 had the greatest potential for L1:WB?

Plan your life wisely, you'll be dead before you know it.

Reply 79 of 83, by Disruptor

User metadata
Rank Oldbie
Rank
Oldbie
PC-Engineer wrote on 2019-12-28, 20:21:

For editing AWARD BIOS you can use MODBIN. Would be helpflull when you tell us which registers in the UMC chipset were responsible for correct L1:WB support.

@Feipoa
Do you know a SIS471 mainboard with correct function of POD with L1:WB? Then i could compare the chipset registers in the BIOS, maybe i am able to find the "little" difference.

And for AMI?
I need to examine the PCI IRQ routing table too.

Do you have a problem with L1 WB?
I'm sorry, I meant, I'm struggling with L2 WB.

With the AWARD BIOS I operate 16K L1 WB & 1024K L2 2-1-1-1 @ WB w/o problems at 40 MHz FSB. (AMD 5x86 @ 160)
DRAM is read 1 WS, write 0 WS.
The problem with Windows 2000 is an IRQ problem due to a foreign BIOS.

I'll continue my UMC 8886/8881 project in an own topic: My 486 UMC8886/8881 Project (Version 2.0)

Last edited by Disruptor on 2019-12-29, 12:56. Edited 1 time in total.