• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

SB750 mode comparison (IDE vs RAID vs AHCI-MS/AMD)

Poisonsnak

New Member
Joined
Feb 13, 2005
Messages
362 (0.08/day)
Likes
18
Location
Saskatoon, SK
Processor FX-8150
Motherboard Gigabyte 990FXA-UD3
Cooling Thermalright Ultra-120 Extreme lapped & shimmed
Memory Patriot Viper Extreme 8GB DDR3-1866 9-11-9-27 1.65V
Video Card(s) MSI 6950 2GB Twin Frozr II @ 6970 875/1375
Storage Crucial M4 0309 64GB
Display(s) Dell U2410
Case Corsair 650D
Audio Device(s) onboard
Power Supply Corsair AX750
Software Windows [8] Developer Preview
#1
After reading W1zzard's Agility 2 review (http://www.techpowerup.com/reviews/OCZ/Agility_2_120_GB/) I noticed some of the comments were wanting a comparison of storage controllers. That got me thinking - I only have one (SB750) but it does have 4 possible modes - IDE, RAID, AHCI (MS driver), AHCI (AMD driver). Which mode is fastest?

Well I spent about an hour benchmarking so I figured I might as well post the results in case anybody else was interested. Between each benchmark run I used HDDErase to give them all a level playing field. The drive was benched as a 'spare' drive (not the OS drive) as well. For drive settings write caching is enabled, write-cache buffer flushing is turned off, and NCQ is turned on in RAID mode (via RAIDXpert). All AMD drivers came from Catalyst 10.7. I only ran each test once so I don't really know how scientific the results are but they seemed pretty consistent (especially the sequential write speed).

I ran CDM and HDTach in each mode but the HDTach results were not very interesting (pretty much 230 - 250MB/sec read and 190-210MB/sec write with 260MB/sec burst in all cases).

The AHCI w/ AMD driver results are kind of disappointing but everything else is pretty much as I expected. Since my drive does not support TRIM I'll probably use RAID mode but if I had a TRIM-capable drive I would probably use AHCI w/ MS driver.
 

Attachments

newtekie1

Semi-Retired Folder
Joined
Nov 22, 2005
Messages
24,411 (5.50/day)
Likes
10,541
Location
Indiana, USA
Processor Intel Core i7 8700K@4.8GHz(Quick and dirty)
Motherboard AsRock Z370 Taichi
Cooling Corsair H110i GTX
Memory 32GB Corsair DDR4-3000
Video Card(s) PNY XLR8 GTX1060 6GB
Storage 480GB Crucial MX200 + 2TB Seagate Solid State Hybrid Drive with 128GB OCZ Synapse SSD Cache
Display(s) QNIX QX2710 1440p@120Hz
Case Fractal Design Define S
Audio Device(s) Onboard is good enough for me
Power Supply Corsair HX850
Software Windows 10 Pro x64
#2
Interesting, thanks for the info.
 

Poisonsnak

New Member
Joined
Feb 13, 2005
Messages
362 (0.08/day)
Likes
18
Location
Saskatoon, SK
Processor FX-8150
Motherboard Gigabyte 990FXA-UD3
Cooling Thermalright Ultra-120 Extreme lapped & shimmed
Memory Patriot Viper Extreme 8GB DDR3-1866 9-11-9-27 1.65V
Video Card(s) MSI 6950 2GB Twin Frozr II @ 6970 875/1375
Storage Crucial M4 0309 64GB
Display(s) Dell U2410
Case Corsair 650D
Audio Device(s) onboard
Power Supply Corsair AX750
Software Windows [8] Developer Preview
#3
appendix a: AHCI/MS scores poorly on HD Tune random?

Before I put my system back to normal I thought I might as well try running HD Tune as well, I remember W1zzard saying in the aforementioned review that he isn't a big fan of benchmarks run with a filesystem in place (possibility of misalignment or fragmentation).

In AHCI mode with the MS driver the drive scored strangely poorly on the random tests - the benchmark took about 5 minutes to complete instead of the normal 5 or 10 seconds and as you can see the scale is all screwed up thanks to some 100,000 ms response times. I re-ran the benchmark, restarted, re-ran it again, and it came up the same each time. Switching back to the AMD driver (or RAID or IDE mode) instantly put things back to normal.

Anyway, SB750 users, take a look and draw your own conclusions. The only thing I can add is I've read a little bit on thegreenbutton and other media/av forums and the MS AHCI driver seems to be linked to video stuttering (although to be honest the more I read over there the more confused I get) - maybe this is related?

My methodology for each driver/mode was: secure erase, seq read test, random read test, random write test, sequential write test (in that order).
 

Attachments

Last edited: