Asus Rampage III Extreme ZioMod Bios (Latest: 1601_S17)

[linxeye]

Just curious, you measure the performance on small block writes with your RAID0 of WD 600 GB or with your SSD ?

[sinders]

Someone over at station-drivers suggested disabling the IAStorDataMgrSvc windows service should solve the memory leak problem.

[ZioGTS]

Just curious, you measure the performance on small block writes with your RAID0 of WD 600 GB or with your SSD ?

Strange question... I posted my results on this thread multiple times. Of course I'm talking about my SSD (my system drive). My HDD RAID is just for data storage.

[Fernando]

Someone over at station-drivers suggested disabling the IAStorDataMgrSvc windows service should solve the memory leak problem.

This "someone" was me.

Regards
Fernando

[linxeye]

Just curious, you measure the performance on small block writes with your RAID0 of WD 600 GB or with your SSD ?

Strange question... I posted my results on this thread multiple times. Of course I'm talking about my SSD (my system drive). My HDD RAID is just for data storage.

Yeah I expected that answer. And sorry if I didn't scroll through all the screenshots. I'm actually surprised that there are any significant performance improvements over RST releases with drives setup as AHCI. But good to know!

[ZioGTS]

I have both the SSD and the two drives Raid on the Intel controller, so I must set it up in bios as Raid instead of ahci.

When I installed 11.2 package, the system choose iastor.sys as driver, while installing 11.5 I found iastorA.sys and iastorF.sys have been selected. I don't know if this is somehow related to performance being better with 11.5, but I guess there could be a link.

Samsung Magician keeps saying my system is not working in ahci mode with both driver versions, and the same happened when I was using the Marvell controller and drivers, but I'm pretty sure it's just due to the crappy check this software is doing. It detects raid mode, so ahci can't be active... And we know this is wrong. Raid drivers contain all the functions of ahci drivers, plus the raid- specific parts.

[apehode]

Hello, I registered just to ask these questions.
I recently bought a Samsung 830 128 GB SSD, and noticed that I got a lot slower speeds than people posting benchmarks on the internet have been able to achieve.
After a bit of searching I find out that that the problem lies with the marvel controller on the Rampage III Extreme board.

My question to you, ZioGTS is, have you been able to get better speeds with your modified bios? I'm currently using the latest official firmware, and I am thinking of installing yours if there is indeed an increase in speed/stability.

How do I install the drivers you recommend in the first post? There is no exe for the (Marvell 91xx driver v1.2.0.1027) so I am a bit confused.

Thank you for your time.

[ZioGTS]

Unfortunately the performance increase is marginal. I ended up putting my single Samsung 830 on the Intel, and accepting the fact that we won't experience real Sata 3 performance with our motherboard. Intel beats Marvell on everything except sequential read, and I assure you, that's not enough to choose Marvell anyway... :-(

[apehode]

Wow, I was expecting to come back and just edit my post to say "Thanks for this," or something similar, but you beat me to it.
I ended up installing your firmware, and both the drivers. I then ran As SSD, and saw that the change was as you say "marginal".

I actually had my ssd connected to the intel ports because I'd heard that it might look worse in the benchmarking tools, but that it beats the marvell in everyday use.
Guess I'll just reattach it to one of the ol' sata 2 ports then.
Has my machine gained any stability increases or such from installing your bios?

[ZioGTS]

I would say yes to your last question: ZioMod integrates the latest Intel OROM too, and it's definitely better than the original one in terms of performance and stability, IMHO.

P.S.: I installed the latest Intel RST v11.5.2.1001 (posted on Station Drivers), and the memory leak bug seems gone for good. Absolute performance is slightly better than v11.2.0.1006, slightly worse then 11.5.0.1207, but definitely more stable all-round.