For defragging XP . . .Thanks Bunches,NoelPS: I have a copy, but Sam has suggested I think Ult Defrag, but not sure how they might differ.

Ultimate Defrag if FSX is your primary application of interest AND is located on the same drive as all your other stuff.PerfectDisk if FSX is not your primary application *OR* FSX is located on its own disk.UD will let you bring whatever directory (or series of directories) you want to the front of the disk drive, giving it the best possible performance. If you have a single hard drive with all your stuff on it, UD will let you move FSX to the front of the pack to make it as fast as possible. UD gives you better granular control over the locations of your files. In special cases like that, UD is in fact the ultimate. PD is good at general defragging. If you already have a dedicated drive for FSX, PD will order things by filename (leading to good, orderly drives) AND give you more benefits with its offline and system defragging. For general use, PD is as close to perfect as you will find.UD for a FSX centered but single hard drive situation.PD for a non-FSX centered computer OR a dedicated flightsim drive. -Greg

Hi Noel,I won't quarrel with Greg's comparative assessment of UD and PD.My internal drive setup is as follows:2 Drives: WD Caviar SE16 250GB (WD2500KS) SATAII + Hitachi Deskstar T7K500 ATA133 UDMA-6 (320 GB)I use the Deskstar ( E: ) for first backups, photos, music files and the odd program or three.The WD SATAII has 2 partitions:1. A 20GB Primary ( C: ) for XP and those programs that refuse to install anywhere else.2. The second partition ( D: ) is reserved for everything else, including FS9 and FSX.I was introduced to PerfectDisk late 2005 and have stuck with it since. I am currently using the latest release, PerfectDisk 2008. I have no idea how it does its job. My experience indicates that it is 100% reliable and it is quick and very easy to use. If my installations of FS9 and FSX are typical benchmark references then I would say that PD performs the task admirably of organizing data on the drive platters. Following initial loads I see very little HD activity and what I do see tends to be very brief. Yes, these can be moments when you may detect slight hesitations in the sim, but I am doubtful whether defraggers like Ultimate defrag will improve this behaviour significantly on modern drives. I imagine the theory is that if files are organized on a named basis on the fastest part of the drive then seek times for individual files will be reduced. This might be helpful when large chunks of scenery data in a particular area are being requested by the sim as the drive head has less distance to travel to find individual files. But, as I said, I really don't know just how significant this effect is in practice.What would be interesting is to see some objective evidence of actual performance gains and benefits each defragger has to offer the end user. Mike

GreggI have a single HD and use my computer,Vista Home Edit, primarily for FSX. I have Ultimate Defrag. Would you please tell me your recommendations for use of Ultimate Defragger.i.e. settings in Auto, Consolidated and the other three choices.Thank you.

I really like PerfectDisk, it's the closest thing to the Norton SU from many, many moons ago. I used to like DiskKeeper but not any more compared to the features in PerfectDisk.Etienne

fbevard:You can use UltimateDefrag to move the MSFS directories to the front of the hard drive, where access time is the fastest. For this purpose, you have to go into the Tools... Options... menu off the top menu bar, and select your C: drive. Then in the area that allows you to choose High Performance options, choose Automatic at 30%, but also select Custom and select your MSFS directory. This will bring your MSFS to the front of the drive first, then the next 30% of your most recently used files.In the Archive section, you can also select automatic at 40% to move your least used items to the INNER tracks of your hard drive, which are the slowest. When it comes time to do the defragment, I usually choose the FOlder/Filename option, and check off the "respect high performance" and "respect archive" boxes to be sure my choices are used. Let her defrag, it may take time.After doing that, I usually only do a "defragment files only", because the massive bulk of my files are already in their optimized locations. If I install a large scenery package, or adjust lots of textures, I'll do the process again - but for day-to-day use or if I only added an airplane, I don't worry about going through the time and process to reorder the drive again.Good luck!-Greg

I tried defragging my FSX drive with Perfect Disk and O&O defrag, and found that O&O did a better job.just my 2 cents

Hi CaptainT18 "I tried defragging my FSX drive with Perfect Disk and O&O defrag, and found that O&O did a better job."Can you explain to us how you know that to be the case and have you any evidence to prove that an O&O defrag led to better performance?Just thought I'd ask? ;)I suspect most of the superior defraggers to a pretty good job and each will have it's own loyal devotees. To me the important thing is their reliability no matter how often they are used. It should be 100% period.Mike

Got it.Thanks Greggfbevard

This is a very simplistic assessment. The claim that anything runs faster when moved to the front of the hard disk has been discussed in hundreds of threads. Having large files in a contiguous state is far far more important than where the files actually are on a HD. The access time difference between "fast" and "slow" portions is measured in NANOSECONDS.I sincerely doubt that you will see any benefit from moving FSX to the fast end of your HD. Everything you modify something (AI, new aircraft, textures, scenery, etc.) it will be appended to the (so-called) slow end of your HD. In order to really keep FSX together at the top end, you will have to defrag constantly. What's more, everytime you start FSX, it builds scenery caches which are new file, thus placed at the end of your HD, the so-called "slow" part. Those are the files FSX really needs constantly when running. So again, there is little to be gained other than maybe a few microseconds during start up. The advantage of Perfect Disk is "Smart Placement", which puts those files you don't access often in one place and those you constantly modify at the other end of the spectrum. This means you don't need to defrag all the time (better for HD, less use), and for FSX it means that all the files you modify often are in one place, which includes scenery cache. In other words, FSX might run faster (again, microseconds if not nano) with this placement than with other options. Microstutters are sometimes reported as a result of defragmented disks. That is nonsense. Microstutters are always caused by FSB/CPU bottlenecks. Get a faster PC for that. That said, defraging is totally overrated. HD slow down with more than 20% fragmentation, which takes about 2 years to achieve on a normal PC (unless you serve files, or constantly compute or download like BOINC, or torrents).Any file placement option on a HD fragmented less than 10% should be indistinguishable in terms of game performance. There are a million other things that influence your game speed and fluidity. I challenge anyone to bring scientific proof that FSX runs faster with one method or the other. I haven't seen anything but anecdotal evidence and subjective impressions.

I used to use perfect disk 8 professional and it worked fine, but then I saw that NickN, who's Simviation's most knowledgeable FSX user, very strongly recommended O&O. After I bought Acceleration, I did a fresh install..... actually, I just described myself two possibilities to why I got better FPSHowever, I also have x plane 9, my first install of it with Perfect Disk 8: 22 FPS average around KBOSO&O Defrag: 27 FPS average, same settings, same weather settings, same time of day, same plane etc.

>fbevard:>>You can use UltimateDefrag to move the MSFS directories to the>front of the hard drive, where access time is the fastest. Greg, do you have any numbers as to how access time is affected by location? I guess it would be a function of what element of the equation is influenced by non-spin rate related factors, such as actuators and other *slack* in the system, as well as spin rate. What do you know?>-Greg

Numbers, no. Just seat-of-the-pants feel for it. I never timed it out, but I can say that it wasn't a placebo effect. (i know, it doesn't count without benchmarks, but what can I say!) In the past, I ran FS2004 and then FSX on my primary drive. The loading of the simulation/loading of flight seemed lethargic. There was a thread or two (here or elsewhere, my memory lacks) which recommended using a tool that can place the sim up front. After moving the files up front, I noted an immediate improvement in the speed of the game loading, as well as the speed of loading into a flight. There were *NO* FPS differences whatsoever - only load times.With the typical simulation directory holding tens of thousands of small files (84,042 files in my FSX dir alone), the 10-15 millisecond penalties of seeking files are magnified. Unlike fetching the occasional config file or application library, the sim has a great many scenery and texture files it needs to find and read before launching - obviously it doesn't need EACH of those files every time you fly, but a lot will be used. One can improve loading times by ensuring the MSFS directory and all its contents are located in a contiguous block on the drive, but you can further improve it by moving that block to the fastest portion of the drive, up front. You're right - it's not only the spin-rate that is involved. By limiting the distance that the read/write heads have to travel, you also improve seek times. The outer edges of a drive hold more data than the inner tracks (think circumference - the outside is bigger than the inside), so you end up with a much smaller physical distance between the first and last simulation file if it is on the outer edge vs. the middle or inner areas. That leads to less travel for the R/W heads, and faster file-to-file performance. In most other applications, it wouldn't make a lick of difference - but with MSFS and its thousands of files, the seeking can add up. As mentioned below in other posts, the differences may not be *stunning*. I did notice a tangible improvement in moving the sim to the front of my hard drive, but everyones experience will be different. I highly doubt that you would get any FPS or smoothness improvements from it - only load times. Currently, I run FSX on a fully separate hard drive (10k raptor) and that has further improved the load times... plus it keeps the sim nice and separate from the rest of my stuff. A single pass from a defragger on my sim drive, ordering by folder and filenames keeps everything nice and organized, and all works well. Ultimately though, it's a no harm, no foul adjustment. Keeping a drive neatly defragged is generally considered to be good practice, and further leveraging that process to take advantage of theoretical and practical differences in the data location on the disk can only help.-Greg

it's highly unlike for defrag to add 5 fps under any situation. to see a difference actually caused by a defragmentation state, you would have to measure the sim performance in exactly the same condition over at least 20 minutes, and compare a graph, not a single fps figure. there is no synchronous connection between fps and hard disk access. last year i had put fsx on a pc which is also running a big database accessed by 100 users 24 h a day. the hd of that server is defragged once a week. even though the database was measurably slower by the end of the week, I could see no differnce in fsx performance whatsoever (just starting the default flight and graphing fps for 4 minutes). again, defrag is important, once you get to 10% fragmentation state of above. but chosing one program over the other has more to do with believing marketing hype than scientific evidence.

I have an opportunity to see if I can tell a difference, perhaps without the defrag tool. In my recent rebuild I have a new install of XP/SP2 and ALL my basic programs including FS9, office, all of it, loaded onto one clean 15K SCSI drive. I am going to clone a new SATA II drive, but the thing has very pokey seek times: 15ms for the SATA, and 6 ms for the 15k SCSI. I knew this going into it. Read rates, on the other hand, are actually better with the SATA. So, I will check to see if there is appreciable difference, related to seek time, since these two drives will be very different seek time wise, and so if there is anything striking in this regard I would think it should show up, even tho I won't be placing FS9 folder on the outer sectors. It's a valid hypothesis you have, and I hope this will give me some clue how much this might play into the issue. The metric we don't have is just how much improvement in actual seek time you get when you move the folder to the outer sectors: is it from 16 to 14ms with this SATA, or more like 16 to 8ms? Pretty small intervals, and I'm guessing the issue mainly has to do with just how much CPU-steal is involved in the disk access question. I think it's minimal for several reasons, on today's racey hardware, but on the other hand, with FSX, the other cores can be used for just this purpose, loading textures. Kinda suggests there IS some burden to the entire sim when files are accessed.As a side note, I set up a test system to see if one could reduce the amount of paging to a swap file by reducing VM to zero. I can attest, using perfmon.exe, that this maneuver truly reduces the continues low level disk access to the disk where pagefile.sys exists: I put the pagefile on a drive without an OS, and without FS9. Prior to reducing the pf size, that disk was being accessed almost continuously from kilobytes to megabytes of activity. This stops when you reduce the pf to zero or near zero. Overall effect on the sim? Well, I'm not doing it now. Not sure it has much effect. Right now with this racehorse I'm using it's not much of an issue!