VOGONS


Asrock 775Dual VSTA & Core4Dual thread

Topic actions

Reply 520 of 697, by VDNKh

User metadata
Rank Newbie
Rank
Newbie
sirotkaslo wrote on 2022-01-27, 06:24:
Hi, i'm having issues with one of the drivers needed for running win98se on 4coredual-sata2 r2.0. […]
Show full quote

Hi,
i'm having issues with one of the drivers needed for running win98se on 4coredual-sata2 r2.0.

After I install via 4in1 5.24a I get a device that can not start standard dual pci ide controller. I take it this is the sata controller? But its running in ide mode since xp install went smoothly. Does anyone have the needed driver?
thanks!

The SATA controller won't work in ACHI mode in 98 but it will work in RAID mode if you install the RAID drivers. RLoew's SATA patch might work but I haven't tried it on my R2.0. I can report that XP has the SATA driver (and all the other drivers for that matter) for that board pre-loaded in SP3, which is why it worked. I think you are better off using IDE on this board for 98 if you can.

If you want to set everything up with RAID drivers you can slipstream the driver into XP like I did here. I'm not sure how to do that in 98 but let us know if you figure it out!

henk717 also posted a custom 4in1 driver that is more compatible with 98SE, while it won't fix the issue you are having, it will install correctly and not leave a resource conflict like VIA's will.

Another note on the RAID driver for the 4CoreDual-SATA2 R2.0 board, if you have EMM386 loading it will hard freeze on the 98 splash screen if there's a device in RAID mode connected and the RAID driver is installed. This is relevant if you have some kind of SB16 emulation going on with your soundcard.

Reply 521 of 697, by sirotkaslo

User metadata
Rank Member
Rank
Member

Yeah I spent some time working on this board, but I think it will be xp only with the HD 3850. I was thinking of using it with the fx 5900xt+v2 in SLI. XP works great, but I want my sli running in 98 as I heard there are some compatibility issues.
Thank you!

Reply 522 of 697, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
sirotkaslo wrote on 2022-01-06, 14:19:

I have a weird issue with my setup. Looks like the hd 3850 is cpu limited, I get about 520fps in q3 at any resolution from 800*600 to 1200*1024.

I have the 4coredual-sata2 with beta bios, E7500, 2*2gb ddr2 ram, hd 3850 with the 13.3 drivers. If i'm going to replace the cpu, which one is better the E5800 or the QX6700?

I would try a bit older driver first (as Quake III is very old game, and newer drivers can perform poorly in it). Try for example: 9.1/9.4 or 10.2/12.4 drivers.

E7500 probably doesn't have enough clock speed to get higher FPS in this game.
It's faster than E5800 at stock clock in some tests because it has larger L2 cache.
Which will help a good bit when memory performace is sub-par (like on this board), and especially on higher CPU clocks (minimum FPS).
Regardless, always make sure CPU is working at full clock when you are testing - C1E/SpeedStep can get finicky and sometimes not work properly.

As for maximum performance you can expect :

3DMark 01SE.png
Filename
3DMark 01SE.png
File size
25.92 KiB
Views
1652 views
File license
Public domain

^All LGA 775 CPUs were VID moded to request from MB +0,1V more than stock Vcore (that's why these OC clocks were possible), and yes - I used 4CoreDual-SATA2 R2.0 as MB for above test.

PS. Personally, I would use combination of 3DMarks and SuperPI/HyperPI as stability tests on this particular MBs (program : 01SE + 03 + 06 + SuperPI/HyperPI), with actual games You will be playing on it as ultimate stability test.
I say this because Linpack/LinX/IBT (Intel Burnin Test) are VERY hard on VRMs, and this style of boards (like previously mentioned), aren't stellar on VRM quality.
If you don't have enough airflow over MOSFETs they can be easily blown up by too high temps (regardless of CPU temperature).
You can also think about Vdroop+LLC this board has, as sort of permament solution to underpowered VRM shortcomings (force lower voltage on highest powered CPUs under load, to make it draw less power, making VRM's job easier at the same time).

157143230295.png

Reply 523 of 697, by McM4r

User metadata
Rank Newbie
Rank
Newbie

Hello guys, someone tried to install a fresh w10 in m.2 nvme with adapter and pci express riser? Would be interesting

*i have 4core dual vsta with 3850 agp and qx9650

Reply 524 of 697, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

You need Clover or specific SSD which has option ROM to boot from legacy BIOS for that.

Legacy compatible PCIe SSDs: Re: what SSD for XP ?

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 525 of 697, by Shagittarius

User metadata
Rank Oldbie
Rank
Oldbie

I would love to put win10 on this, only problem is the AGP drivers. Does anyone have a silver bullet AGP driver that will work with win10 for this board. I tried a lot of drivers but was unsuccessful.

Reply 526 of 697, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

AGP support was removed from modern Windows 10. Most likely, old Windows 10 builds will be majorly dropped from software support after Windows 8 support will end (2023). So overall it's pointless to pursue.

Last edited by The Serpent Rider on 2022-05-13, 15:52. Edited 3 times in total.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 527 of 697, by Shagittarius

User metadata
Rank Oldbie
Rank
Oldbie
The Serpent Rider wrote on 2022-05-13, 15:42:

AGP support was removed from modern Windows 10.

I know I was hoping that I could find a 3rd party driver made for another OS that might work in win10. I've found a thread where someone said this was done successfully but I havent been able to find the driver.

Reply 528 of 697, by McM4r

User metadata
Rank Newbie
Rank
Newbie
Shagittarius wrote on 2022-05-13, 15:43:
The Serpent Rider wrote on 2022-05-13, 15:42:

AGP support was removed from modern Windows 10.

I know I was hoping that I could find a 3rd party driver made for another OS that might work in win10. I've found a thread where someone said this was done successfully but I havent been able to find the driver.

I have hd3850 agp working with w10 just install the register workaround uploaded in previous post and then 12.4 ago hotfix if you have hd4000 series agp i just have the stock driver of windows and is working flawlessly (via chipset)

Reply 529 of 697, by schmatzler

User metadata
Rank Oldbie
Rank
Oldbie

AGP still works, you just need to use the AGP driver from an older Windows 10 release.
I've documented that here:
Re: Asrock 775Dual VSTA & Core4Dual thread

"Windows 98's natural state is locked up"

Reply 530 of 697, by Shagittarius

User metadata
Rank Oldbie
Rank
Oldbie
schmatzler wrote on 2022-05-13, 18:22:

AGP still works, you just need to use the AGP driver from an older Windows 10 release.
I've documented that here:
Re: Asrock 775Dual VSTA & Core4Dual thread

Many thanks I will give this a try. Think the 12.4 will work with a 9800pro? I dont remember where support stopped.

Edit: Looks like 10.2 for 9800 Pro 64bit.

Reply 531 of 697, by Mister No

User metadata
Rank Newbie
Rank
Newbie

Has anyone gotten the multiplier adjustment working with 4CoreDual-SATA2 and Core2 Extreme QX9650 or QX9770? I have a QX9650 and I can successfully change the multiplier only if I set the "Overclock Mode" to "[Auto]" on the "CPU Configuration" screen, which effectively sets the FSB to 266 MHz.

If I manually set the FSB to 256 MHz (or any other lower or higher value, including 266 MHz), I can adjust the multiplier, but the setting is not persistent - it reverts back to the default of 9 as soon as I save the BIOS and the computer restarts.

It's really weird, I know, but I have tried every possible combination of settings and BIOS versions I could think of and nothing works except the above-mentioned workaround with setting the "Overclock Mode" to "[Auto]". And since my CPU is unstable at anything higher than 256 MHz (like all quadcores, judging by this thread), that makes the multiplier adjustment in BIOS practically unusable for me 🙁 I have to use ThrottleStop and change the multiplier from within Windows, which I would like to avoid, if possible.

Has anyone had this problem and somehow solved it? If so, please share the solution with me.

P.S. I am using the BIOS version 2.20A and the motherboard I have is Rev 1.

Reply 532 of 697, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
McM4r wrote on 2022-05-13, 05:54:

Hello guys, someone tried to install a fresh w10 in m.2 nvme with adapter and pci express riser? Would be interesting

*i have 4core dual vsta with 3850 agp and qx9650

NVMe won't work, installing it will prevent board from booting. But I only tested my 4CoreDual-eSATA2 R2.0 with beta BIOS.

Mister No wrote on 2022-05-16, 22:40:

...

Disable C1E/SpeedStep for multiplier control, also use "Manual" instead of Auto when you can.
Vdroop on CPU from VRM may prevent CPU from working after OC, it depends on Your board capacitors state. In general for Quad Cores OC on this board, a VID mod on CPU is needed. It increases default voltage on CPU (bypassing "no CPU voltage control" in BIOS).

157143230295.png

Reply 533 of 697, by Mister No

User metadata
Rank Newbie
Rank
Newbie
agent_x007 wrote on 2022-05-17, 05:39:

Disable C1E/SpeedStep for multiplier control, also use "Manual" instead of Auto when you can.
Vdroop on CPU from VRM may prevent CPU from working after OC, it depends on Your board capacitors state. In general for Quad Cores OC on this board, a VID mod on CPU is needed. It increases default voltage on CPU (bypassing "no CPU voltage control" in BIOS).

Thank you very much for your input, I've read this thread thoroughly from the start and your contributions are definitely very helpful and informative.

I can use "[Manual]", but as I said, in that case the value I set for the multiplier (under "[Ratio CMOS Setting]") is not preserved after the BIOS is saved and computer is restarted. If I use "[Auto]", the value I set under the "[Ratio CMOS Setting]" is preserved. I tested all multiplier values between 6 and 14. If I use "[Manual]", the multiplier reverts back to the default of 9 as soon as I save the BIOS and computer restarts. If I use "[Auto]", the computer boots normally after the BIOS is saved and multiplier stays at the value I set previously under the "Ratio CMOS Setting". Really weird, huh?

The only problem with "[Auto]" is that my CPU is unstable at anything higher than 256 MHz and "[Auto]" is 266 MHz for my CPU.

C1E/SpeedStep is already disabled, otherwise I wouldn't be able to change the values under "[Ratio CMOS Setting]".

I am aware of the weak VRMs on this board, but that is definitely not the root cause of my problem. My CPU is stable at 3.2 GHz with default vcore of 1.25V, I tested it with a 24-hour run of IntelBurnTest. I used ThrottleStop to change the multiplier from within Windows. I have a beefy top-blow cooler with a 135 mm fan that blows directly onto the MOSFETs and vdroop on this motherboard is so huge that vcore under load goes sometimes as low as 1.125V (more than 0.1V vdroop!). My QX9650 consumes no more than 75W, I checked with a power meter and all capacitors on the motherboard are in a top notch condition. There are no overheating issues.

Besides, If my CPU is stable at 3.2 GHZ with 1.25V default vcore, than 2.56 GHz (256x10 set in "[Manual]") should be a piece of cake, but it's not. It's still not working - immediately after restart the multiplier is reverted to 9. The same happens If I try any other multiplier value between 6 and 14. Higher than 14 causes no boot, because more than 3.6 GHz is not possible with the default vcore.

P.S. How are you changing the multiplier of your QX9770? Also using ThrottleStop from within Windows like me or you have no problems to do it from within BIOS?

Reply 534 of 697, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie

I changed my CPU multiplier in BIOS. Over x14 was possible, but CPU-z couldn't tell what CPU I was using anymore.
Just to confirm : Does CPU-z see SSE4.1 as instruction that is supported on your CPU ?

Can you check your CMOS battery with multimeter ?

157143230295.png

Reply 535 of 697, by Mister No

User metadata
Rank Newbie
Rank
Newbie
agent_x007 wrote on 2022-05-17, 14:27:

I changed my CPU multiplier in BIOS. Over x14 was possible, but CPU-z couldn't tell what CPU I was using anymore.
Just to confirm : Does CPU-z see SSE4.1 as instruction that is supported on your CPU ?

Can you check your CMOS battery with multimeter ?

Yes, CPU-Z shows that SSE4.1 is supported, my CPU is correctly recognized.

I already checked the CMOS battery and it's fine (multimeter shows 2.95V). Nevertheless, just to completely rule it out as a culprit, I replaced the battery with a known working one from my other computer with the same result - multiplier control doesn't work in "[Manual]" 🙁

I don't know what else I could possibly test, I've tried other BIOS versions as well, but the problem is always there. Could it lie on the revision of the board? I know it's a long shot, but I see that you have R2.0 and I have Revision 1. That's the only meaningful difference I see between our systems. I also have HIS 3850 AGP, but that's the topic for another thread 😀

Reply 537 of 697, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++
agent_x007 wrote:

NVMe won't work, installing it will prevent board from booting.

Probably not an issue for drives I've listed, including NVME.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 539 of 697, by The Serpent Rider

User metadata
Rank l33t++
Rank
l33t++

Most of those drives listed are PCIe 2.0. Although yes, none of them are using NVME protocol. Samsung 950 Pro won't post, but Plextor M6e will.

I must be some kind of standard: the anonymous gangbanger of the 21st century.