Of defrag c: -b; defrag c: /x; Defraggler and PerfectDisk10

Is believing that defragging offers real performance benefits a myth?

Unless I'm living in some magical kingdom (looks around studio - nope! ), it isn't, for me.

For an overview of the threads where I'm discussing various aspects of my Hitachi 7K500 HD upgrade...
See:
http://forum.notebookreview.com/showthread.php?t=441674

This thread shows how I tested and compared identical machines (except for RAM & HD's)
See:
http://forum.notebookreview.com/showthread.php?p=5609040#post5609040

continuing from those identically setup machines the following further tests below between Windows defrag tools, Defraggler and PerfectDisk 10 were timed.


PerfectDisk10=PD10
Toshiba MK3252GSX=Tosh
Hitachi 7K500=7K500

I have been told (davepermen, where are you? ) that PD10 is only a 'placebo' effect that is only in my mind. davepermen, I already know my mind will be going soon, but this area is certainly not where it starts!


I received my new Hitachi 7K500 and although I was itching to get it set up 'perfectly' as soon as possible, I had the opportunity to test my exact VAIO in its original configuration (4GB RAM vs. 8GB and the 'stock' Toshiba HD) side by side with my new Hitachi 7K500 install. I couldn't pass up on this opportunity - not only for me, but for everyone else who wonders what an 'upgrade' will do for their systems. And also to answer the question that is on everyone's mind; does defragging 'work'?

I know, but this is the first time I've documented it to this degree. Hope you enjoy the read and I'll be happy to answer any questions that I can.

To the numbers!

So, this first set of numbers is before I ran any defrag commands (nor were there any other non-Windows defragging utilities even installed):

1st time start Premiere Pro CS4:
Tosh: 85 seconds vs. 7K500: 50 seconds

Shut down (after 10 minutes idle):
Tosh: 14 seconds vs. 7K500 13 seconds

Boot to Desktop (initial view):
Tosh: 68 seconds vs. 7K500 46 seconds

Desktop is 'Usable' (wireless connection shows connected):
Tosh: 145 seconds vs. 7K500 90 seconds


Next, I ran the windows built in process idle tasks command along with the defrag c: -b and defrag c: /x commands.

The first one forces Windows to build a list of 'boot' files along with any outstanding tasks it may have.

The second one forces Windows to put those boot files in the fastest part of the drive - the beginning.

The last command forces Windows to compact any free space.

I did not do a built in Windows defrag - because its useless! As its based on Diskeeper technology and that is 'non' technology to me (I've tested it many times over the years and it is as good as Norton Antivirus in improving the performance of my systems).


The 'process idle tasks' command run times:
Tosh: 12 minutes vs. 7K500: 1 minute

The 'defrag c: -b' command run times:
Tosh: 120 seconds vs. 7K500: 90 seconds

The 'defrag c: /x' command run times:
Tosh: 145 seconds vs. 7K500 60 seconds


(What's interesting is the process idle tasks 'test' I timed. This 12 times improvement shocked me - but there is no way to do a 're-test' unless I re-installed everything again.

Thinking about it, this huge improvement is the much more efficient and speedy Hitachi taking advantage of small time bites when I was busy doing other things. Although the Toshiba had the exact same time advantage, it's obvious that the Hitachi could do much more with the small time slices, than what the Toshiba could).


This second set of numbers is after I ran the above built-in defrag commands.

2nd time start Premiere Pro CS4:
Tosh: 38 seconds vs. 7K500: 30 seconds

Shut down (after 10 minutes idle):
Tosh: 14 seconds vs. 7K500 13 seconds

Boot to Desktop (initial view):
Tosh: 66 seconds vs. 7K500 41 seconds

Desktop is 'Usable' (wireless connection shows connected):
Tosh: 105 seconds vs. 7K500 70 seconds


Hmmm... me thinks defragging does offer some benefits, huh?

Keep in mind that this second start of Premiere Pro above is mostly being boosted by SuperFetch (in addition to the defrag commands used). It increased the Hitachi's speed by 20 seconds, which is great (67% faster), but it increased the Toshiba's speed by 47 seconds which is a 223% speed increase. Yes, caches work wonders. But fast drives are still faster.

I also wanted to see what a 'warm' start of Premiere Pro would be:
Tosh: 14 seconds vs. 7K500 13 seconds

What's a 'warm' start? It's starting up a program right after you've closed it. This is basically telling you the fastest your system can ever load that program, because the files that the program uses stay in RAM until you load something else to take their place.


Next, I downloaded and installed Defraggler on both systems.


Since I don't know the program at all (and I don't have the time, nor the wish to properly 'learn' it) I just downloaded it, installed it and hit 'Defrag'.

Defrag very 1st time with Defraggler:
Tosh: 34 minutes vs. 7K500: 22 minutes

Now, the familiar tests:

This second set of numbers is after I ran Defraggler once as it reported both drives as 0 files fragmented and both drives started with 1% fragmented when Analyzed with Defraggler. 1% was 385 files fragmented for the Hitachi and 325 files fragmented for the Toshiba.

3rd time start Premiere Pro CS4:
Tosh: 38 seconds vs. 7K500: 30 seconds

Shut down (after 10 minutes idle):
Tosh: 14 seconds vs. 7K500 13 seconds

Boot to Desktop (initial view):
Tosh: 83 seconds vs. 7K500 38 seconds

Desktop is 'Usable' (wireless connection shows connected):
Tosh: 145 seconds vs. 7K500 70 seconds

Hmmm... this seems that on the slower HD, Defraggler made things worse to get to a 'usable' desktop? Yet on the Hitachi, it made it show the desktop slightly faster, but slightly slower again to hit the 'usable' state.

One thing these raw numbers don't show (even though they are accurate) is that although the wireless connection icon showed 'connected' at 145 seconds, the Toshiba system continued to thrash the HD for an additional 50 seconds before it quieted down. The Hitachi's HD light was 'off' at the 70 second mark.

So, how would you 'count' this result above? I give you both numbers and let you make your own decisions. Remember that all tests are run 3 times and averaged (except, of course, for the 1st time runs for anything) - myself? When I saw that Defraggler had undone what Win 7 had achieved, I un-installed it immediately.

Finally, I installed PD10 on each system, here are some timed results:

Analyze C: Drive with PD10:
Tosh: 80 seconds vs. 7K500: 61 seconds

1st 'Online' defrag with PD10:
Tosh: 46 minutes vs. 7K500: 23 minutes

1st 'Offline' defrag with PD10:
Tosh: 317 seconds vs. 7K500: 337 seconds

Analyze C: Drive with PD10 after 'Offline' defrag:
Tosh: 90 seconds vs. 7K500: 79 seconds


Note: the 'offline' defrag times are correct - but remember that the Toshiba was defragging a 4 GB hibernation file and the Hitachi was defragging an 8GB hibernation file. (In addition, of course, to the other system files they also needed to defragment, 'offline').

Now, as defragmenting is mainly a HD intensive process, the online defrag shows just how much faster/better the Hitachi HD really is. BTW, both drives showed 0% fragmented files before the online defrag started.

If you're not familiar with PD10, then you may wonder what it was doing for 46 and 23 minutes on the Toshiba and Hitachi drives, respectively.

What it does is not only defrag the file system, it also places files optimally so that future fragmentation is delayed as much as possible. It achieves this by placing the boot files first (at the beginning, or, fastest part of the HD), then places any files that are rarely modified (mostly read, not written to), then it places occasionally modified files after them and finally; recently modified files and lastly the directory entries on the file system. An offline defrag moves unmovable (unmovable while online) system files to their optimum position on each partition while also defragmenting them too.

Remember, PD10 is only for mechanical HD's - Don't use on an SSD on a regular basis.

Finally, PerfectDisk 10's numbers:

This set of numbers is after I ran PerfectDisk 10 as above.

4th time start Premiere Pro CS4:
Tosh: 42 seconds vs. 7K500: 37 seconds

Shut down (after 10 minutes idle):
Tosh: 14 seconds vs. 7K500 13 seconds

Boot to Desktop (initial view):
Tosh: 45 seconds vs. 7K500 39 seconds

Desktop is 'Usable' (wireless connection shows connected):
Tosh: 73 seconds vs. 7K500 66 seconds


This set of numbers is after I ran PerfectDisk 10's online defrag a second time (after the offline defrag run).

5th time start Premiere Pro CS4:
Tosh: 40 seconds vs. 7K500: 32 seconds

Shut down (after 10 minutes idle):
Tosh: 14 seconds vs. 7K500 13 seconds

Boot to Desktop (initial view):
Tosh: 45 seconds vs. 7K500 37 seconds

Desktop is 'Usable' (wireless connection shows connected):
Tosh: 66 seconds vs. 7K500 62 seconds

Now, this is more like it! Does running PD repeatedly keep decreasing our times? No, but because this is the first time (and especially because it was the first time for an 'offline' defrag) the second PD run was needed to properly 'optimize' the system for initial use.

Some additional tests I ran after all the PD10 optimizations - 2 online passes, plus one offline pass:

1st time start After Effects CS4:
Tosh: 42 seconds vs. 7K500: 37 seconds

1st time start 3DsMax 2010:
Tosh: 73 seconds vs. 7K500: 50 seconds

'Warm' start up Premiere Pro CS4:
Tosh: 12 seconds vs. 7K500: 12 seconds

Microsoft Security Essentials Quick Scan:
Tosh: 2:19 min:sec vs. 7K500: 1:56 min:sec


Windows Updates (4, including the malicious software removal tool):

Tosh: 18 seconds download, 3:25 min:sec to finish installing 'online', 6 minutes total time from start of Windows Update to rebooting to finish installing them and back to a 'usable' desktop (and back to work).

7K500: 17 seconds download, 1:57 min:sec to finish installing 'online', 4 minutes total time from start of Windows Update to rebooting to finish installing them and back to a 'usable' desktop (and, back to work).


What do all the above tell me? Well, first PD10 is not a placebo effect. It can make an older HD perform better than a non-optimized faster HD.

It effectively masks the shortcomings of slower drives with weaker IOPS and/or longer access time 'scores' by making them less dependent on their weaknesses.

(Simply by placing all the directory entries together - even at the end of the data on each partition - the HD is doing much less work than it would have to otherwise).

A good example of this is the 'warm' startup times for Premiere Pro after a PD optimization run; both drives are now at 12 seconds (they were 1 second apart before and also a second longer each) - this is because the HD has to go to only a single part of the HD (where the directory entries are) to verify that it has all those files in memory, instead of jumping all around the drive to check as when the directory entries are not placed all together (the 'default' with Windows Defrag, and also how Defraggler handles that too, from what I can tell).

The best of all possible worlds, it seems, is to have the fastest HD possible and use PD occasionally too (about once or twice a month seems right) so it seems I'm set, for now.

Well, at least until the G3 Intel SSD shows up, huh?

I hope this post makes people understand the difference a properly optimized system may make.

Not only to existing, older HD's, but also to new HD's with the latest and greatest technology on offer too.

If you noticed that after PD10 the HD's are performing very similarly, you may wonder if the Hitachi was worth the trouble to upgrade?

Let me copy a part of the 'real world' thread (link at top) to let you know just how fast the Hitachi is in 'real life':

This is for Phil (mod):

43GB File folder of pictures, music, program files, various install files, ISO's, documents, etc. copied from one partition to another:
Tosh: 63 minutes; 7K500: 44 minutes. Over 40% improvement.

Sorry for the long post, but Thanks for reading (if you got this far).

 

Did you try using MyDefrag, ex-JKDefrag.exe?

 

I haven't used either versions of that product, but from what I've read, I don't think I would want to.

The closest product that I can remember that offered similar (not to this level though) of options was the original Norton's SpeedDisk. With the Norton product, there never was an 'optimum' strategy - just gave you a ton of options so you 'thought' you had/were using the best one.

I'm not saying that the MyDefrag software is the same, just that I simply do not have the time anymore to reliably and equally test (to my satisfaction) this type of product which offers almost unlimited options.

PerfectDisk, in comparison offers the select few degree of options that really make a difference:

'Defrag Only' = Which matches almost every other defrag product I've tried and which I never use in PD.

'Consolidate Free Space' = Which offers a quick way to shrink a partition in Vista or Win 7 to its smallest possible size. Or, offers a very quick defrag option for the Temp partition when I know I'll be needing the biggest and most contiguous, free space possible when I'll be editing my RAW images (and PS's Scratch Disk thrives on available contiguous free space).

'SmartPlacement' = Is the default for a reason. This is simply the best 'just run and forget' option to really see an improvement in how your computer works.

Combine the above with the StealthPatrol technology and 'optimizing' the disk is really as simple as set and forget.

What sets PD apart though is the optimization and defragmentation of the system files during an 'offline' defrag. The 'optimization' part? It places the system files, pagefile and hibernation file where MS themselves recommend for optimum O/S performance. I know, I sound like a commercial, but I would not be repeating this if PD didn't offer real and significant gains in my computing experience.

No other product I've tried has that part ('offline') of the defrag process down pat, and combined with a 'SmartPlacement' online optimization, no other product offers such a real and noticeable performance increase for mechanical HD's - whether they are old tech and slow, or new and relatively fast, PD makes them even more efficient and effective.

A good example of the StealthPatrol technology in PD was last week when I met with a client I had seen two weeks ago when he had just downloaded the trial of PD and I had set it up for him. We didn't have time on his much older notebook to run an offline and online passes, so I set StealthPatrol and did just an 'offline' pass and had to leave for my next appointment (an analysis showed fragmentation was almost 28% on his 90% filled, 30GB HD).

When I saw him last week, I remembered that we hadn't run PD, 'online', and asked if he had - he said no; but he had only used the notebook a couple of days in the week since I had seen him. I started PD, thinking we needed to run the online pass and was pleasantly surprised to see less than 0.5% fragmentation! When asked if he noticed his system seemed 'slower' than usual (thinking StealthPatrol 'interfered' with his normal use of the computer, while it tried to defrag the old HD), he said no - actually it booted up noticeably faster and was quieter too, (less head movements, I guess).

StealthPatrol not only defrags with minimal interference to normal computer use, but it also does so with almost 100% effectiveness too, in minimal 'idle' time.

Yeah, I was more than impressed, but my client ordered PD for the rest of his office computers too - he really was impressed.

 

I have sworn by PD10 for a long time, I was first using it at version 8. Without a doubt it makes the HDD work the most efficiently possible. The added advantage you do not list is that it keeps the drive running like new longer.

As the other defrags alow the files to spread out further and further increasing access times exponentially PD10 keeps them in the best shape possible. I've used PD on installs and 2 years later the system still ran as good as new.

I've been told as well about the placebo effect. I've also benched the speed differences and proved that wrong as well. For those that don't use PD10 that is fine, but I can do an install and keep it for a long long time running at peak using PD10 while I watch others doing reinstallations attempting to regain performance.

I have the 500 GB 7200.4 Seagate and am not thrilled with it. I eliminated the stalls and click issues by using QuietHDD. With HDTune my read access is 17.5 ms. I may just get the Hitachi as 10% faster is 10% and I am not confident in these Seagate drives either.

 

you don't live in magical kingdom land. you life in the slum, and there people need drugs to get trough. drugs like different defragging programs.

magical kingdom land has ssds

merry christmas to you tiller of course man of the big posts

 

Magical Kingdom Land is still expensive and not quite ready for prime time. Generation 3 or 4 and trim etc should bring order to MKL and price of admitance will be hopefully more reasonable rather than the kings ransome as it is now.........

 

Tell that to a person who bought the 80GB version for $900 cdn after tax+shipping.

I'm eyeing my 2nd SSD sometime this year, hopefully a Postville-refresh.

 

TANWare,

Thanks for your comments and Happy Holidays to you!

The links at the top of post #1 do indicate that it keeps the drive running like new longer, but I also hint at it in post #1 above:

Quoting myself:
"What it does is not only defrag the file system, it also places files optimally so that future fragmentation is delayed as much as possible."


As for the Seagate you have, well, I had my share of failures with them (tried, and returned, 4 of them) and ultimately settled on the Scorpio Blue 500GB for about 9 months until the Hitachi became available. Some have seen no difference going from the Seagate 7200.4's to the Hitachi 7K500's, but to me, in my systems - it is not only night and day - it is heaven and earth, with the Seagate definitely not on the 'heavenly' side!

Benchmarks are the last thing I rely on when assessing a new piece of H/W or S/W. My benchmarks are the actual use I need my systems to perform in. For my 7K500, I benchmark really bad (I don't know why and Phil (mod) couldn't help me get a 'better' bench score either), but it doesn't matter as the 43GB timed copy (bottom of post #1) attests to.

I don't need to know I have the best (benchmark-wise), I just need to know I am performing better than what I had.

Cheers!

 

Good idea, but not really an issue for modern computers since that's typically handled automatically. And if I may suggest, you address one subject at a time. You wouldn't want to reveal ALL your secrets ALL at once would you?

 

Merry Christmas to you too davepermen!

Sometimes, living in the 'slum' is alright! As long as the 'reality' is better and cheaper than magical kingdom land... and the drugs (PD10) is a one time shot!

But with performance boosts like PD10 offers, yeah - it is addictive!

Hope you weren't too naughty in MKL... you got the second G2 80GB, right?

Cheers!

 

Krane, sorry, I've read the above posts and your reply a couple of times now, but I do not understand where you're coming from?

What is a good idea?

And, any information I have/use that I pass on to this forum and it's members I probably learned from this same forum too (in the distant past), so - no secrets on my part.

But, I'll try to stay focussed, succinct and on topic within my posts; even though life is rarely that compartmentalized itself, nor are computers either.

Everything is interrelated.

 

Your Defrag thread.

Secrets is a reference of speech. It means don't give out too much information too soon.

Again, a suggestion to consider your audience: From novice to avid enthusiast!

Even so, it can still be divided into parts. After all, you wouldn't want to overwhelm half your audience would you?

 

Thanks Krane!

 

Yah, I know you hint at it, but you have to spell it out for people. It isn't a placebo effect when you first optimize the system with PD and boot to desktop in 65 seconds and two years later you still boot to desktop in about 65 seconds with the same install. now you may experience slightly larger delays as the AV or other bootup software gets bloat as you upgrade but the base speed is still just there.........

 

oh, it's ready for prime time since about a year for me. and i would NEVER give them back.

it's not that expensive. it's high end. like gpu's, like cpus, like having much ram. and it's in the same price class. but the gain is so much bigger, it's actually CHEAP for what you get.

 

yeah, i got the second one. and in some days/weeks, i sure get the first one, too. my dad got interested hearing i get "one too much"

well, your reality has to suck. in my reality, disks are a lovely memory of the past, together with floppies and such and the same slowness is a fading memory...

i would be interested in a fresh installed os on a hdd, and the same on a second one. one with PD10, one without. and then, after a week of usage, a final defrag on both, and then turning defrag off.

because my experience with pd is (same with jk, but even much stronger) an initial short gain, followed by a quick drop below non-pd-drugged disks.

similar to how most drugs work, actually an initial emotional super-flash, followed by a big depression.

that's why i ditched pd. because it was very short-termish, requiring re-drugging much too often. i spent more time on defragmenting than the gains i got out of them that's when you know you're addicted, and the addiction is bad for you

but now, in magical kingdom land... no need for such drugs.. heaven is around all day long *singing a long, dancing towards the sunshine*




info: yes, it's very late here
and merry christmas

 

I have to agree on the 'cheap for what you get' part. No matter how much I've paid for 'real' performance gains, I never thought of the cost a second time afterwards!

Again, with StealthPatrol on PD10, you never have to really touch PD manually again (unless you want/need an offline defrag performed). So the drug induced magic continues...

I'm sure my reality doesn't suck because I've tested and confirmed with the SSD's I can get - for my use, the performance is the same or worse (and that is not with the Hitachi 7K500 I'm currently using - that was with the Scorpio Blue 5400 RPM drive too.

Wishing everyone (had) a Very Merry Christmas!