ASRock 939Dual-SATA2 with X2, No setting for 1T or 2T

[ModelTech]

I have built a new system with the ASRock 939Dual-SATA2 (Bios Version 1.2) and a X2-4400+. I am using OCZ 2GB (2 x 1GB) DDR 400 (PC 3200) Dual Channel Platinum Memory with 2-3-2-5 timings.

The MA option to set the Command Timings is totally missing. No MA option at all. I've read a couple of other forums with people with the same problem.

Any insights or suggestions? Does anyone have a X2 processor working with the MA options available in BIOS?

 

[Westmassguy]

That option only appears in the OCW Beta 1 BIOS for now.
It will more than likely appear in a future stable BIOS directly from ASRock.
BTW, try setting your TRAS to 10 instead of 5, Asyn Latency to 6, Read Preamble to 5.

 

[Fam Money]

It's in the 1.20 as well......

 

[ModelTech]

The problem seems to only be with X2 processors. If you have a single core processor the MA option is available. If you have a X2 processor then the MA option doesn't even show up.

 

[Westmassguy]

Ahhh, that would explain why I don't see it then.
I'm also running with all four DIMM slots full, so 1T is impossible for me anyway.

 

[ModelTech]

Does anyone out there have an X2 processor running on this motherboard and have the MA (1T or 2T) memory option available? If so, what version of Bios.

 

[BlackPear1]

I'm running X2 3800+ on this motherboard and I've never seen the MA option in bios. I've used 1.10, 1.20 (current, for stability), 1.21 and OCBW1. Maybe they eliminated it for X2 processors because this board apparently will not run them stable at 1T. By setting 1T with A64 Tweaker and testing I confirmed this problem in my setup and there are many similar reports in forums. I hope that Asrock will fix this, because it's a nice board otherwise.

 

[ModelTech]

ARock just sent me a new Bios 1.23. I'll put try it tonight and let you know if it worked.

 

[ModelTech]

Well the updated bios seems to be running well. The MA setting is available and I could set my memory to 1T while using and X2 processor. My bandwidth in memtest98 went from 1953 (2T) to 2263 (1T). I also ran 3DMark03. My previous score was 11274 and now it is 11756, a 4.2% increase when using 1T. Im going to have memtest98 run its test for a couple of hours to see if it has any errors. I'll post my results.
All I did to get the new bios was email ASrock support and tell them of the problem. I got the updated bios 2 days later. Pretty good if you ask me.

 

[BlackPear1]

I'll be awaiting those memtest results with interest, particularly in that troublesome (at 1T) test #8. Congratulations on getting anything from Asrock. I've emailed support twice and never gotten more than an automated receipt.

 

[Lord Banshee]

you can e-mail that 1.23 bios , chris.ruesga@gmail.com

I'll give them a good test

 

[ModelTech]

Memtest locked up. Thats not good. And whenever I tried to start CodUO the system would reboot. I set it back to 2T and it seems to work fine. I believe I have good enough memory to handle 1T, I paid enough for it. So its probably the Bios/Motherboard.

Lord Banshee, Let me know how it worked out for you.

 

[Westmassguy]

I'll be happy to test it as well!
Westmassboy@comcast.net

 

[BlackPear1]

We look forward to your report with the results of your tests. Sorry the bios didn't work for you ModelTech.

 

[Fam Money]

Does it look like they changed anything in the BIOS? Maybe added more vcore and vdimm options.

 

[formulav8]

The new 'official' 1.30 bios HAS the MA feature available to us Dual-Core owners (finally). I just updated to the bios and its now there to set to 1T or 2T. Apart from that, I see NO other additions to the bios. Nothing voltage related was added either.



Jason

 

[Westmassguy]

I've tested the 1.23 and 1.3 BIOS.
Both have MA/1T/2T available now in memory options, but it's still unstable with my X2.
Also no increase in Base Clock/HTT, 270 MHz is still the highest I can go.
No increase in Vcore adjustments above stock 1.35-1.4, in fact they broke the Vcore settings all together.
The worst part, both versions disabled my Vcore mod.
The Vcore and Multiplier options are there, and can be adjusted, but after I save and reboot, no change.
Machine reboots with 1.55 volts no matter what its set to in BIOS.

I'm back to BIOS 1.20 for now.

 

[BlackPear1]

Unfortunately I'm back to 1.20 too. I'm beginning to doubt that I'll be able to use 1T on this board. I don't know whether to use this limitation as an excuse to get 2 more DIMMs or just change boards. Even after experiencing a number of problems I'm getting attached to it.

 

[Westmassguy]

Same here BlackPearl1
I have become attached as well.
Just picked up some 2 X 512 Patriot PC3200+XBL to replace my 4 X 256 Hynix.
This board loves TCCD.
I'm now running 274 MHz Base Clock/HTT.
Pretty sweet.
Still no 1T for me though.

 

[Fam Money]

We just need to get it in our heads that 1T is not as big a deal as the benchmark programs make it out to be. Though I understand the frustration of spending top dollar on a component and not being able to run it a stated settings.

With dual cores does each CPU have a memory controller? If they do like I assume they do, it may be harder to get 1T out of them. Just a thought, though I'm probably wrong and you guys are going to tell about a gang of boards that run mem at 1T with dual cores.

 

[BlackPear1]

Fam Money, you're probably right. There is an interesting thread comparing 1T & 2T performance on A64 and the difference is not as substantial in most games and other apps as it is in benches.

Westmassguy - that's a really nice o/c you have there. My hardware and/or skills are not up to yours. I can do many benches at 2.7, but can't call my system totally stable at that speed and I get an immediate reboot at 2.75. I still have options to explore and I'll work on it.

 

[Westmassguy]

Thanks.
My OCZ DDR Booster came in handy for this speed, unfortunately it just crapped out on me.
Back to 270 till I get a replacement.

 

[formulav8]

Well, IMO, for 1x simple setting having a 5% performance increase/decrease is quite a impact. If it was 3 or 4 settings and it made a 5% increase/decrease I wouldn't consider it that bad. But to me, for one single setting making a average performance increase/decrease of 5% or more is quite substantual.

Also, as a anandtech review showed a few days ago, the hit can be 10% or more.

For one setting to make that kind of impact is important, at least to me it is.


Also, I have had NO problems running at 1T. I can run it at stock 200mhz 2-2-2-10-1T. And with the fsb at 260mhz on the 133mhz mem divider I can run the mem at 22210-1T timings.

I have 2x 512MB sticks of BH5 ram and put it on the high vCore setting whether it needed it or not. I really haven't checked if it needed the voltage or not to be honest.

Anyways, this is just my opinion on the matter. And the previous poster on the 1.30 bios not having working vCore settings looks to be correct. I don't know how Asrock could have let that past inspection.

I honestly do NOT think I will purchase another Asrock board. Although I have had no issues, I am not pleased with their bios makeup and the very limited control over the board. Ex: No vCore above 1.4v, a crappy dimm limit of 2.7v, and hardly no control over the memory timings and such. These features/controls are standards nowadays.

I really just do not like their bios design I guess. I guess I am just so used to Awards bios on my mobo's.


Jason

 

[Fam Money]

Well at least I can agree with you on 1 thing formulav8; I more than likely won't ever buy an ASRock mobo again for the reasons you stated.