Samsung Magician v4.0 Released

Version 4.0 dated February 5,2013

Support - Solid State Drives MZ-7PC064D | Samsung Memory & Storage

Title : Samsung Magician Software Ver.4.0 Release note
Applicable Model

- 470 Series: MZ-5PA064* MZ-5PA128* MZ-5PA256*

- 830 Series: MZ-7PC064* MZ-7PC128* MZ-7PC256* MZ-7PC512*

- 840 Series: MZ-7TD120* MZ-7TD250* MZ-7TD500*

- 840 PRO Series: MZ-7PD128* MZ-7PD256* MZ-7PD512*
Improvements & Modifications

- Completely redesigned User Interface and User Experience.

- Simple, graphical indication of drive health status and Total Bytes Written.

- At-a-glance system configuration information, including SATA interface and AHCI support.

- Re-designed OS Optimization featuring three common user profiles and detailed explanations of each configuration option.

- SSD Authentication feature to verify your SSD is a genuine Samsung product.(From 840 and 840 PRO Series)
Operating System

- Windows XP SP2(32bit) / Vista(32/64bit) / 7(32/64bit) / 8(32/64bit)
Note

- SSD will not be recognized if it is part of a RAID array

- Magician does not work with SSDs connected via the SCSI controller interface.

- Only MBR and GPT partition types are supported. Magician may not work with other partition types.

- Secure Erase is not supported for Windows 8 systems.

 

WhatsThePoint, Thanks for the information. Finally got rid of that annoying upgrade notice on start up that was plaguing the 3.2 version of Magician.

4.0 is a nice step up in the interface and tool set.

 

Its great to see Samsung update and maintain their Magician software. Thanks for posting this WhatsThePoint.

Updating to Magician 4.0 right now...

 

Is this software really necessary? I have an 830 and it seems to run fine without the additional software.

 

Display is a little funky. Chops letters at the bottom and top.

 

I would hope not, that would be really silly if you needed to have Magician installed before you can use your Samsung SSD. But then again, if it wasn't there, you'd be saying why isn't Samsung making an SSD Utility App like Intel does! Just sayin'.

Its not necessary, but it doesn't hurt anything either. The choice is obviously yours. I don't even know anymore.

 

I found these tips to improve your SSD's performance. No software required.
I've read that Magician software (perhaps earlier versions) can mess up solid state drives.

 

Thanks for that link, i will try out these tweeks

 

I've never bothered with any of these tweaks and my Samsung SSDs have never given me any problems that I am aware of. I've got over 2.4 TB written to my 830 alone. I'm not at all worried about prematurely wearing out my SSDs with hibernation or restore files. These SSDs can handle years of use without worry so why bother really? I use my computer day in and day out like i would any other computer with a HDD. I suppose if there are no consequences in doing these tweaks, then it wouldn't hurting anything. But like I said, I've never bothered and don't plan on starting any time soon. Sometimes I don't even bother with the tweaks the Magician software recommends, sometimes i do. Again, it hasn't affected me in any negative way so I'll just keep using my computer like I'm use to.

I believe tilleroftheearth would agree with me here and hopefully he can share more of his wonderful knowledge.

 

Thanks. (blushes).


I do agree that no tweaks are better almost 100% of the time - for almost anything the system will be used for. I have seen the v4 Magician in action and it is laughable vs. the Intel SSD Toolbox (though both offer to 'optimize' certain parts of your system to the detriment of the user experience, imo).

Physically install the SSD, do a clean install of the latest O/S (Win8x64 Pro highly recommended) and install the minimum amount of drivers/services/software you need to get your work done. This is what ensures a fast, stable and reliable system.

Not any magic tweaks that do more harm than good in the overall user experience.

 

Only special thing that I did, was to not allocate 20% of the SSD's space.

Since I only have the one SSD (an mSATA SSD) and nothing else (no resident HDD in the laptop), I don't need Intel Rapid Storage Technology drivers (Intel RST, or IRST).

I did take the time to go to Crucial Solid State Products Support page and click on "all SSD articles" at: SSD Support

That brings up 35 article titles. Quicker to just glance at the titles and read the few that might mean something to you.

All the rest of it (i.e. TRIM and GC (Garbage Collection)) needs nothing to be done.

For the moment, I've also made sure none of my software does defrag of the SSD drive.

Tiller, what are your current thoughts on defrag???

 

Defrag an SSD. I find improvements by running PD 12.5 (custom SmartPlacement) about once a month.

See:
http://forum.notebookreview.com/sol...64-smartplacement-defragging-perfectdisk.html



As a matter of fact, I did this recently to a Win8x64 system (not mine) that was just clean installed about 2 months ago. Running PD (demo) for the first time; It defragged around 8K files with around 70K fragments. The responsiveness of this system improved markedly (as also noted to me by the owner the next day).

Defragging works because no matter how fast an SSD is; the max latency to access all those files fragments still make up a real and discernible amount of time in day to day use.

I don't recommend to do it continuously, obviously, but the benefits are enough to have me continue using it with even my SSD-only installs too.

(Another reason it works; PD Pro defrags free space too - not just files like every other defragger does - this helps the SSD find large blocks of 'clean' nand to write new data too - at the fastest possible speed the controller allows).

 

Contrary to that, this article with the following passage disagrees with defragging solid state drives:

"So, no, you should not defrag an SSD. And performing one will actually reduce the life of your drive. All of the SSD manufacturer’s know of this problem and they have come up with an optimization technique with the use of the TRIM command."

 

Uh, about that article.

It is wrong.

Performance goes up, the additional wear is inconsequential and HDD and SSD's are both Random Read devices (opposite of what the author stated in his story).


The nominal lifespan of SSD's hasn't been a question for a very long time now - even though that article was supposedly written in May of 2012, it reads more like 2+ years ago.


Everybody will say don't defrag an SSD. But I like to push the things I'm told to not do by the 'experts'.

 

FWIW, I don't think running a defrag every so often (once every 1 or 2 months) is going to kill the effectiveness of the number of writes over the lifetime of an SSD. Over time files generally "settle down" with modifications (you stopped playing this game or that, or that spread sheet or presentation file have served their purpose and are no longer modified), most of these files will be placed sequentially after fragmentation and end up in a nice order on the SSD that never changes. So, a defrag on these files ends up just skipping over them anyway. The same will go for system files... [EDIT] On some levels, perhaps this would benefit an SSD in some ways. If a file on disk is always in a read-only state, perhaps a defrag wouldn't hurt as it would move the data to another NAND cell, freeing up that NAND cell for some possibly new write cycles (assuming it had not been modified for a long period of time).[/EDIT]

As for any real benefit, although an SSD can access any LBA in constant time, a sequential file ends up evenly spread over multiple banks of NAND chips. Don't quote me on this, but I'm pretty sure others have posted this may lead to a small performance gain. I believe others have stated controllers can access each individual chip at the same time. So there may be a bit of a win here as a read can grab all of the data at once across those chips, rather than waiting for read after read if the LBAs end up mapped on the same bank of NAND cells. Again, I'm uncertain about this, but if someone has a white paper or some other technical article they can post, I'd be most obliged.

I do want to note, the "snappiness" of an SSD is a subjective measurement until tiller or someone shows us any kind of self made benchmark. BootTimer? Loading of a large file, etc. However, in the grand scheme of things, I don't think tiller, his acquaintance, or anyone else is going to lose the ability to write to their SSD any time soon because of a defrag 4 to 6 times a year.

Now scheduling a drive for constant defrags... we'll that's another story, and should probably be avoided.

 

My Windows Experience Index is currently rated at 7.4 and I haven't used the Magician software.
Using the Samsung optimization, would it increase that rating or would it not make any difference whatsoever?

I also omitted from disabling pagefile all together as instructed in the tweaking article.
I did however change it to 1024 max since I have 16 GB of RAM on board.
My previous setting was set astronomically high by Windows 7 at 24420 MB!

 

I wouldn't put too much into the WEI. I'm running 7.7 on my Crucial C300 - a 30 month old drive.

With 16GB, in general cases, using virtual memory is up to you. However, some apps will require tons of memory and others will check it on start up and won't run w/ out it. FWIW, I'm running 12GB of RAM w/ a 12GB page file on the SSD.

 

My own Win7 performance rating for "disc" (as in the drive, as in the SSD) is 7.8 where max is a possible 7.9

That's with a lowly 256GB Crucial M4 mSATA SSD, on a lowly ThinkPad X220. No HDD on the machine.

Try it yourself and find out!

If you use "sleep" then setting to "1024 max" is probably a bad mistake.

Which is what? ~24GB?

With 16GB of RAM, going into sleep mode you will need a file of about 16GB in the root directory ( i.e. C:\ )

Check the size of the hidden system file called pagefile.sys

On my X220, with 6GB of RAM, pagefile.sys is just under 6GB in size.

On my T530, with 12GB of RAM, pagefile.sys is just under 12GB in size.

BTW the screwy RAM totals are temporary; target is to get X220 to 8GB and T530 to 16GB in the next few weeks.

 

I do not know what SSD you have. My 830 WEI is 7.9. My Sequential R/W are 540/430 my IOPS are R/W 74000/37500 with the Samsung Magician.

I agree with you that wear is not a major issue if like all things done in moderation.

On everything else not so much. Your disagreement with linked article on sequential or random is more of ambiguity or semantics. However we want to confuse the disscussion HDD's have a physical linear structure a line that is a circle and even a thousand of them are still linear. Linear is related to sequential.

Tiller you are aware that where the OS sees the "data" and where it really is are not the same. Think what the SSD controller is and does in the first place. Your Overprovisioning argument is counter to your incorrect belief that defragging actually puts files in any kind of real physical proximity.

That said I find most of your SSD beliefs to suffer from lack of real benifit vs actual dangerous behavior. Trim/Garbage collection are the "defragmenting" for SSD's. On Overprovisioning, if someone can't on their own avoid filling up their SSD above 80% it is a great idea. My 30% used space does not need it yet/maybe ever.

 

I just looked and cannot go any higher on data transfer than 7.9 which my 830 is reading.

So regarding defragging solid state drives, it would then be advised to do so every so often?
Now is that performed through Windows 7 or some other defragger?

 

Good to hear your 830 sounds like the WEI is what it should be.

I don't think defragging does any real good I just don't think if only done rarely won't really hurt. So many writes that the finite lifespan is not a real major point. If anyone thinks in ten years their current SSD will be their primary storage device they must not care about being so far behind the curve as to be "off the cliff".

 

Hey I may still have my SSD within a decade from now!
256 has enough room - doesn't it?

I've got 12% fragmentation on my 830 with 216 GB to spare.

 

I think you are talking in circles. I know what I'm talking about when HDD's and SSD's are the topic. I never said that defragging put files in physical proximity what it does is lets the O/S put one request in for each file (instead of hundreds of requests for highly fragmented files. This is what gives more 'snappiness' after a defrag (and leaving the SSD to 'rest' for a while).

The most important thing that PerfectDisk Professional 12.5 does though that no other defragger I know does; it defrags free space. This gives maximum performance from the SSD as any new writes are going to defragged nand (here, defragged means nand pages that have been consolidated, leaving ready, clean and complete (not partial) pages ready to be written to).

You think you know what you're talking about with storage subsystems.

Keep learning.

 

Me and the rest of the world. What is trim and garbage collection? You talk about things that are already done you just have some odd ways.

Physical proximity means nothing to the SSD controler. Block x and block y are still different blocks as such they need to be "requested" so back to what does it matter.

You keep talking pages and I know in your speak it means what you say it means. But since SSD's are restricted to "block" writes it would be nice if you threw that term in every now and then.

You have questioned my knowledge before. Fine so be it. No one in any feild of expertise supports much of what you say. I can see flaws in your logic. I will hang my head with the current experts at this time.

Once again do you know what trim and garbage collection are? 90% everything you talk about doing but the accepted method.

 

You and the rest of the world will catch up to my 'odd ways' when I will already have moved onto something new.

Just like the world became enlightened when Anandtech released their most recent article about the need for extra spare area - a topic I was ridiculed for since 2009 when I claimed that SSD's are not as consistent as a properly setup and partitioned HDD.

The 'accepted method' is fine if you want to be out of date. Much like everyone once 'knew' the world was flat and the center of the universe.

I'll take a chance to be proved wrong (if I truly am), but in the meantime, I am reaping the rewards of my educated guesses on the best way to squeeze the last drop of performance from storage subsystems.

Something, btw, that I have been doing since the early 80's...

take care.

 

O&O Software says:

"The defragmentation of SSDs leads to no improvement in their performance, and may even negatively influence their life expectancy.
As a result, such defragmentation must be avoided through the automatic detection of SSDs by the operating system or the defragmentaion software.
On the other hand, the automatic and regular execution of the TRIM command increases SSD performance and significantly extends its life span."

View the PDF file.

In addition, Samsung's reply to defragmenting SSDs:

"SSD’s do not need defragmentation because they have no moving parts and can access any location on the drive equally fast. Defragmenting an SSD can acutally impair performance. Make sure your OS recognizes the drive as an SSD and enable TRIM to keep performance of the drive consistent. Moreover, Samsung SSD’s are wear-leveling over the entire space allowing the drive to work globally, so defragmentation is not required."

"Techniques like TRIM and garbage collection will help maintain your solid state drive’s performance over time."

 

I was wrong about page writes. Looks like it can be done. But I like Super Bee's find and post.

 

Okay, one step at a time. At least I wasn't burned for being a witch this time.

 

Just my opinion on this:

Physical proximity doesn't meaningfully affect the time it takes to access a page in the physical nand. In fact, having all the data on one nand package will actually be slower due to having less bandwidth than 2 or more chips (think raid0, but for nand). In this sense, fragmentation will actually be good for the ssd.

However, placing all the fragmented parts of a file into as few pages as possible will free up other pages for writes (essentially what garbage collection does), which may lead to a performance improvement. Also, note that the physical locations of a file are abstracted from the OS when using an ssd. So what the OS perceives as fragmented will not reflect how the controller perceives it. I.e. when you defrag, pages that are logically next to each other may still be on different nand packages.

So make sure that you guys are talking about the same thing (physical vs logical fragmentation).

I'm on my phone right now, but I can post links to back up what I am saying later.

 

Hi.

Does anybody get a crash with SM 4.0 after you select an os optimization, i use maximum reliability, it makes changes to the notebook and asks you too reboot, but after the re-boot the SM 4.0 autostarts and then crashes, i have removed and re-installed twice.

John.

 

You're confusing a couple things. Page file is memory space used on disk, and the only time it's needed is when either there is insufficient memory to run a program or when a particular program is specifically designed to use it. It has nothing to do with sleep mode. The file to which it seems like you're referring is hiberfile.sys, which is the area of your hard drive or SSD where active memory data is stored when the system hibernates. Sleep mode is different than hibernate, as data stays in RAM (which remains powered) and therefore does not require any additional storage space.

 

Agreed that: "Sleep mode is different than hibernate"

The guy put a max limit on his pagefile.sys file of 1GB.

Maybe it doesn't make any difference for sleep, but short of actually trying it myself, I also wouldn't be making a statement like your above comments.

Meaning that there is a clear relationship to memory and the file size of pagefile.sys.

Perhaps it only matters if one tries to hibernate the machine?

FWIW on my one remaining WinXP/32 machine, I have 8GB of actual RAM but the pagefile.sys in the C:\ root is ~3.2GB. Which makes sense given 32 bits is limited to upper limit of less than 4GB.

 

I thought so. Hinted at this earlier in the thread - http://forum.notebookreview.com/sol...ung-magician-v4-0-released-2.html#post9075330 . Thanks for the confirmation.

 

Speaking as someone who has also reduced pagefile to 1GB and who uses sleep mode often, I've had no issues whatsoever. I also got rid of the massive hiberfil.sys file (my system has 32GB RAM) by disabling hibernate. Saves lots of SSD space that way!

 

Here is another source about over provisioning... Interesting read.
Understanding SSD over-provisioning | EDN

 

V4 seems very unreliable to me, half the time i start the app, no ssd condition information will show, and i will have to close the app and re-start it, also when you select a ssd profile, i selected maximum reliability, it tells you to restart your notebook, but when i boots into windows V4 starts and then crashes straight away, i re-installed the app a few time but it`s just the same

John.

 

Just installed Magician 4.0 on my Windows 8 system. Says that it cannot secure erase the SSD w/in Windows and wants me to create a bootable USB or CD/DVD. I created the bootable USB on a SD card in a USB card reader. I booted it up, but the display is all corrupt and nothing is readable.

I redid the bootable USB option with a flash drive and have the same issue. Haven't tried a bootable CD/DVD since I don't have an optical drive.


Has anyone else experienced this?

 

I'm having both these problems on a freshly installed Windows 7 x64. Can any one provide some help?

 

i had the graphical glitches and i reinstalled magician twice i think, so far the ui has been stable after the reinstall.
i also tried maximum reliability and it doesnt seem to affect the stability of the program.

 

FWIW -

When firing up Magician 4.0, moments ago, I was prompted to update to v4.1, of which I decided to pass.

After a very quick passing glance at the Sammy support website, I didn't see 4.1 listed, but a little Goggling did find a .pdf reference to it:

Samsung_Magician_4_1_Release_Notes_1_1.pdf

Page 3 shows the change log, none of which seemed particularly compelling to my application, but perhaps others may find it useful.

Edit: C & P the changes -

Magician 4.1 features a number of improvements over version 4.0:
- Multi-language support for the following list of languages is added.
 Spanish
 Deutsch
 Japanese
 Chinese (Simplified)
 French
 Italian
 Portuguese
 Russian
- Fixed the crash issue related to resolution less than 1024x768.
- Application Resize to 1024X728.
- OS Optimization: Support for Power option (Samsung high
performance) provided.
- Help File updated for all Languages accordingly
- Retaining Scheduler records from 3.2, 4.0 to 4.1

 

Installed V4.1, but now the drive information does not show when the display resolution is set to 1920x1080 and the text scaling is set to 125%!!!. Changing the text scaling to 100% allows for the drive information to be displayed properly. This was not a problem in V4.01. Unfortunately, SM constantly asks you for updating even if you say "NO" in the same session. I expect that this will have to be fixed in version 4.1.1

 

No wonder I couldn't see my disk info!
Reverted back but yes the constant update prompts can be annoying.

Sent from a Galaxy far far away

 

Still on 4.01, still see no need to upgrade, but yes, the upgrade prompts are irritating! I believe the upgrade prompt occurs twice: Once pre-launch, and once immediately post launch, but no more during the same session. My app is set to minimize to the sys tray, though. Still a PITA, but I live with it.

Tried Googling for possible hacks, to no avail.

Otherwise I like this s/w.

 

I don't even allow the SM to run - cancelling it each time I log on is annoying enough.

(But I'm on the latest version).


...

 

Well,

i usually don't run Magician continuously, but do pop it open once in awhile to check things like health, data written, firmware availability, etc.

And now that I've left it running for part of the day, I stand corrected! It does indeed regularly nag for the upgrade to 4.1!

I still like that, through Magician, I could do dynamic OP, via a windows GUI, after the initial SSD install. That actually was an (additional) selling point in my selection of a Sammy 840 Pro.

 

'Dynamic OP' is built into Windows 7 and 8 (and even Vista).

Not a selling point imo - the right time to do it is when it is new/unused at the 'clean install' stage. Otherwise, the 'unallocated' nand starts off 'dirty' and will take awhile to get to a 'fresh' and/or 'ready now' state to prevent unnecessary WA during use or simply during the TRIM/GC routines.

 

Fair enough, and I have no reason to doubt your knowledge or accuracy, because I did notice the drive seemed to take a day or two to get to full speed. Thank you.

This was my first SSD and I was unfamiliar with the technology. Being able to simply drop in my new SSD, load an OS, and then select Samsung's recommended OP for their drive, as well as selecting one of the Samsung OS optimization profiles, was appealing to me. The ability to quickly and easily check and upgrade firmware through a windows GUI was also a distinct positive.

There was an awful lot of disparate, conflicting, and just FUD info out there, and this allowed me to get running quickly and easily using Samsung's own optimizations without any stress or mistakes. So I felt there was value there for me.

Of course, this meant I put myself in Samsung's hands, but hey, it was their drive!