czr

Use this http://www.steelbytes.com/?mid=20 If it's still showing slow speed, it may be some issues with the actual drive(s) (i.e. bad sectors or slow sectors) Use this http://hddscan.com/ on each drive separately to diagnose sector problems Cheers Cezar

OK stop piss-farting around and wasting your time, I've found that the easiest/cheapest way to do is, is to use a separate SATA controller and do the "controller shuffle" 1. Install PCI SATA card 2. Install Drivers for SATA card 3. Shut down machine and put drive on SATA card, set Motherbaord controller to RAID/AHCI) 4. Install Drivers from Motherboard controller 5. Shutdown and put drive back onto motherboard controller 6. Presto! all will work fine without the need or wasted time of a re-install that generally doesn't work most of the time and all for the cost of a cheap PCI SATA card, and many less gray hairs Cheers Cezar

Have a few flash drive that have two partitions on them, a small read only partition (appear as CD-Rom under Windows Disk Management Console) and a larger standard partition Anyone know how I can delete and consolidate these into just one partition and hence 1 drive letter? Cheers Cezar

666 I'm going to have to totally disagree with your assessment of the Intel X25-M, I sold my Gen 1 in preparation for the Gen 2 and using a standard Seagate 7200.12 is like pulling teeth compared to the SSD. I'm getting a head ache just thinking about how annoyingly slow (responsiveness) my system is now in comparison. Granted that Seagate have now brain deaded NCQ on the 7200.12 and new firmwared 7200.11 (f#cking toss bags! way to ruin a perfectly good product! I'd seriously like to punch the fwit in the teeth responsible for that monumental cock up!) my FTP and DC server is back to it's pre working NCQ days (damn useless!)

Installed my X25-M and I have to say for me it is a huge improvement over the WD Velociraptor. My main application I use is Outlook and the difference it makes in doing searches in Outlook is tremendous, easily 5x-10x quicker, so am very happy, as I no longer have to walk away form the computer and do something else when doing "search field and message body" searches Basically I'm very happy, but saying that I don't think I would have got it if it wasn't a freebee, it's still too $$$ for desktop environments

My experience with Seagate slim 7200.9, 7200.10 and 7200.11 has been that they are slower at seeks and also quieter, from that I've deduced that the acoustic management of the drives is tuned towards quiet operation instead of fast and noisy seeks. Unfortunately I'm not aware of any tools that are available for these Seagate drives that can alter the acoustic management profile

Are you 100% sure it is "Chipset" based RAID and not PCI chip based RAID What chipset is it?

The following were carried out without NCQ enabled i.e. the SATA controller was set to IDE mode 64KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 42MB/s (Average) 128KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 47MB/s (Average) 256KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 51MB/s (Average) 512KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 55MB/s (Average) 1MB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 66MB/s (Average) Just for the hell of it I also carried out the test at 16MB block size and the transfer was 110MB/s which is not surprising considering how much time the head is actually spending reading compared to seeking. You can see that the drive behaves quite differently when NCQ is off compared to NCQ on, in some situations it performs better and in other it performs worse, as I've mentioned previously my own real world (read "file serving application") experienced has shown that for my application NCQ "on" is of a benefit, in you situations that may be different. Also I'd like to make a note that (256KB block size) with NCQ off you can actually hear the drive madly seeking away when there are 9 instances going where as with NCQ on there are very few drive seeks. Also as the block size get bigger there seem to be less "drive seek" which makes sense as there are less seeks in any give second, as the block size get even very big i.e. 16MB, the graphs from each individual test starts to approximate the look of the graphs of the NCQ (256KB block size) tests, which again is not surprising once you understand what NCQ does and how very large block size closely approximates NCQ I'll be posting results for the VelociRaptor soon with NCQ off

By the way, just so that there isn't any confusion, if I don't state any other parameters in my later posts, it is because the parameters have not changed to my initial post, i.e. AHCI mode , Drivers, Mainboard chipset etc I have not tested with NCQ disabled as it's a bit of a pain for me to do (i.e. have to change BIOS setting from AHCI to Pure IDE mode) and then get Windows to work happily in that mode)

64KB Block Size WD WD1500HLFS-01G6U0 Firmware : 04.04V01 Size : 150GB 1 instance of the program at 0% Position ~ 123MB/s (to determine baseline performance) 9 instances of the program ~ 14MB/s (Average) 128KB Block Size WD WD1500HLFS-01G6U0 Firmware : 04.04V01 Size : 150GB 1 instance of the program at 0% Position ~ 125MB/s (to determine baseline performance) 9 instances of the program ~ 19MB/s (Average) 512KB Block Size WD WD1500HLFS-01G6U0 Firmware : 04.04V01 Size : 150GB 1 instance of the program at 0% Position ~ 125MB/s (to determine baseline performance) 9 instances of the program ~ 51MB/s (Average) 1MB Block Size WD WD1500HLFS-01G6U0 Firmware : 04.04V01 Size : 150GB 1 instance of the program at 0% Position ~ 125MB/s (to determine baseline performance) 9 instances of the program ~ 74MB/s (Average) Look like the Seagate still performs better then the WD, so I can only conclude that the WD NCQ implementation is still "more" broken then the Seagate, I'd expect the WD to be faster if they had equal NCQ implementations, as it is much faster at "seeks" I suspect that if I was to carry out this test on other model (read slower) drives they would still be slow (i.e. WD 1TB 7200rpm, etc), then what I now have to consider still somewhat broken NCQ implementation (Seagate), at least it's not as bad as in the past where NCQ performance sucked across the board. However I must say that I'm rather disappointed that block size makes such a difference to transfer. Does anyone know how to possibly set the block size that Windows use by default, it seems that this could potentially have a pretty big impact on drive performance.

64KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 19MB/s (Average) 128KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 127MB/s (to determine baseline performance) 9 instances of the program ~ 29MB/s (Average) 512KB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 128MB/s (to determine baseline performance) 9 instances of the program ~ 60MB/s (Average) 1MB Block Size Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 129MB/s (to determine baseline performance) 9 instances of the program ~ 88MB/s (Average) Pretty crap results at 64KB and 128KB, interesting that 256KB seems to be optimal* might run some different block size transfer on the VelociRaptor * and it was pure coincidence I picked that size, I wasn't trying to show bias, I've always used 256KB for every test I've ever carried out, so it was just for "consistency" that I kept using that size. When I first started testing this NCQ "thing" out years ago, I did test at different block sizes at the time to determine if block size made a difference (I think it was in the Seagate 7200.8 or 7200.9 days), it didn't at the time.

Just did some more testing, this time with the 3 platter 1TB Seagate, and the new WD VelociRaptor Please note that all performance numbers are from MS Performance Monitor (Windows XP) Seagate ST31000333AS Firmware : SD15 Size : 1000GB 1 instance of the program at 0% Position ~ 125MB/s (to determine baseline performance) 9 instances of the program ~ 97MB/s (Average) WD WD1500HLFS-01G6U0 Firmware : 04.04V01 Size : 150GB 1 instance of the program at 0% Position ~ 126MB/s (to determine baseline performance) 9 instances of the program ~ 33MB/s (Average) Kind of disappointed that the new Raptors haven't improved in this area, was hoping that they had, it's still not going to stop me from using it as my Boot/OS/App drive, the 1TB Seagate is still my Storage drive and I've added the Seagate 640GB as my temp/scratch drive

CTRL+F is your friend! Please look at post #39

I've never tested any of the current Samsung, so I can't comment, nor have I ever commented on Samsung F1 drives, so I'm unsure where you believe you have seen me mention the F1(s)

So far all 7200.11 drives I have tested display signs of proper NCQ implementation, so basically all are quick, however some are a bit quicker then others ST3320613AS (Faster Transfer : dude to higher density platters. Slower Seek : I'm assuming "quiet seek" is enable on this unit) ST3500320AS (The "Benchmark" drive) ST3640323AS (Faster Transfer : dude to higher density platters, my current OS/Boot Drive, as I'm waiting for the new 150GB Raptor - if it ever comes out- ) ST3750330AS (Pretty much the same as the 500GB) ST31000340AS (Pretty much the same as the 500GB, maybe slightly slower seeks)