I currently have an Samsung 830 as my OS drive, and I have changed the path of my Documents/Video's/Pictures etc to be stored on a secondary hard drive as they take up too much storage. I have run the OS Optimization software which disables the system indexing, however I would like to be able to still search for files etc in my documents. Is there any reason why I should do the steps in the link below? http://mintywhite.com/windows-7/7ma...-move-search-index-to-another-physical-drive/ Thanks
Enable AHCI, use OS with TRIM support is the only two things you need to follow with SSD. No need to disable indexing, swap, hibernation etc. The only reason to move these out of the SSD is if you need the space they occupy.
I don't disable indexing, but set it up selective in the indexing options for just certain directories, which all happen to be on other drives than my ssd. Moving the index is not needed. If only directories like \music, \movies are to be indexed the index itself is fairly static as most people don't write and erase a large portions of files in such locations. So in reality there really won't be a lot of writes to index file. Also ssd have way more life than most people thought years ago when such things started being done and have been proved to be unnecessary. Moving a lot of those things off the ssd is in reality more of a penalty, especially swap.
Exactly! Like both faugusztin and lysol said. Please stop breaking Windows by disabling stuff, which you call "optimizing". SSD has came a VERY far since the first generation. You can write intensively on them, and it will last 5 to 11 years based on your usage, and which SSD you have. Assuming 5 year as an absolute worst case, by that time you'll already would have a new bigger and faster with even more number of write SSD in your system. It's been proven time and times again, that SSD write limit is of no concern in a home computer, even if you do like programming where compiling trashes the SSD in write amounts, and I hit that compile button A LOT per day. SSD Life says I have more than 9 years left. So, make sure your SATA controller is set to AHCI, and enjoy your SSD like an HDD. Leave Windows do it's thing. Assuming you have Windows 7 or even better 8, it knows that you have an SSD and takes care of everything for you.
OK thanks for the help guys I will enable it again, it was only turned off as when I ran the the Samsung optimization software it automatically turned it off.
I really don't understand why it "needs" to be AHCI? its not needed for TRIM and the only other 2 reasons to use AHCI really don't apply to SSD's Native Command Queuing (not something that would effect an SSD) Hot swapping (never done it with any drive)
1) boards have 3 options - IDE (limits you to SATA1 speeds), RAID and AHCI - both are technically AHCI mode, but RAID mode usually enables the UI for the RAID on the board durring the POST. 2) NCQ do apply for SSD, most SSD have multiple channels, thus more commands thanks to NCQ means higher utilization. http://en.wikipedia.org/wiki/Native_Command_Queuing#NCQ_in_solid-state_drives Both are good enough reasons i think.
My understanding is modern motherboards although the BIOS says IDE actually means serial ATA and this does not limit speed to the same as IDE I have installed SSD's not using AHCI and see no performance degrade in fact on my GF's DELL 1720 laptop ATA is FASTER in benchmarks than using AHCI by more than 50MB/ps on a SATA I controller. post 1 and 3 showing someone who seen the same results http://en.community.dell.com/support-forums/laptop/f/3518/t/19300004.aspx
If possible, can we ignore 4 year old Core 2 laptops systems, which weren't SATA2 to begin with ? Well, on paper they were SATA2, but to keep battery life as high as possible many laptop manufacturers did the most retarded thing on their laptop mainboards - they set the ICH9M controller to special power saving mode, which was equal to SATA1 speeds. What is described in the article is the exception to the rule, hell, it's even possible Dell messed up their BIOS and what was labeled AHCI was ATA and vice versa. The speeds would suggest that is the case. Anyway, a small test : ASM1061 controller, IDE mode - 324.4MB/s seq read, 23.49MB/s 4k read, 25.37MB/s 4kQD32 read. ASM1061 controller, AHCI mode - 379MB/s seq read, 23.51MB/s 4k read, 211.6MB/s 4kQD32 read. Intel controller, AHCI mode - 531.8MB/s seq read, 27.91MB/s 4k read, 225.7MB/s 4kQD32 read. I don't have the Intel/IDE combo due the W7 crash when switching from AHCI to IDE. But you can already see the differences. Sure, sequential speed doesn't seem to be limited by SATA1 speeds anymore in IDE mode, i apologize for that. But take a look at 4k read speeds, and then at 4kQD32 speeds with IDE and then look at both AHCI results. That is Mr. NCQ doing it's job in AHCI mode with both controllers.
AHCI and SSDs Typically, many hardware review sites, as well as SSD manufacturers are recommending that AHCI mode is used with SSD drives. However, we have run our own tests here in the forum, and we believe this is misleading, only where SSD drives are used independently (ie not in a RAID array). AHCI mode as previously explained enables NCQ (native command queuing) which is really not required for SSDs as they do not need optimising in this way as there is no physical movement of heads or platters. In many cases, it can actually hinder SSD performance, and even reduce the lifetime of your SSD. SSDs and the TRIM command The TRIM command is dependent on the SSD itself, and the operating system supporting it. It is not dependent on the host controller, and AHCI is not a requirement. TRIM is natively supported in Windows 7, as well as Linux since kernel 2.6.33. http://forum-en.msi.com/faq/article/ide-sata-and-ahci-all-you-need-to-know
Wikipedia is not a trusted source. If you check out anything to monitor it's really out dated and simply wrong, and you don't need to be an expert to know that it's wrong. I don't care that it comes from MSI. They have trusted engineer as sources, if they quote Wikipedia.. then this is a huge problem. NCQ doesn't degrade SSD life, as you are using it for reading. So this is B.S. Also, you exclude computer with HDD's inside who needs NCQ for a nice performance increase. And you also block the usage of eSATA ports due to lack of hot-swappable. The only downside of using AHCI mode, is that if your system uses a BIOS, and not UEFI full supported motherboard, it MAY take a bit more time to pass through the POST process before it boots the OS. In Windows 7, in order to get TRIM to work, you need MSAHCI drivers loaded, as it contains support for TRIM. If you use Vista or earlier, then you use the SSD built-in garbage collection and TRIM support. The reason why you want the OS managing TRIM, is because the OS knows what you are doing, the SSD doesn't... only guesses. So, it doesn't do a good job. Also you can see on the internet people reporting faster SSD speed with their controller set to AHCI mode than IDE. Doesn't mean your Dell system showed the reversed that this is the law. Dell screwed up, or due to technology limitation at the time (eg: SSD's where new, TRIM was not supported by SSDs 4 years ago, so they are all going theoretical) you have this strange reverse effect. Like people like me with SATA-3 controllers from Marvel on their motherboard, but ends up SLOWER than Intel SATA-2 controller built-in. It happens, it sucks, but this is life. Don't jump on first gen stuff, especially when the technology isn't there to confirm anything.
like the MSI faq says they got the conclusion from peoples results on the forums not wikipedia I personally don't think their is much in it either way (nothing you would ever notice in real world use) their are loads for and against all over the net, personally use AHCI just because its the latest but a few days ago I installed an SSD via AHCI in a mates pc.. benched it then had an issue with another drive so turned AHCI off benched it again and got almost the same result.
Because MSI knows the internal architecture and SSD firmware better than SSD manufacturers, right ? There is a reason why they recommend AHCI, and first and foremost it is NCQ and link power save. In single threaded applications, there can be a marginal increase in performance (read single percentages) due smaller overhead of the protocol. But once your system needs to do 2 tasks, AHCI takes the lead due NCQ. No, SSD doesn't have moving parts. Instead, they have delays due cost of write operation, they have multiple channels where each can do a separate piece of work, thus increasing performance when more than one operation is required from SSD. Without NCQ, what SSD or HDD does is - computers sends an ATA command to the drive, drive performs the command and sends back the result. Until that result comes back, the ATA interface is blocked. With NCQ, SSD or HDD can receive multiple commands - SSD usually accept 32 commands in the queue. This means the computer can send more commands to the drive at same time. And in case of the SSD made in this decade that means improved multithreaded performance. And finally, benched with what ? HD Tune ? HD Tach ? ATTO ? All of those are single threaded benchmarks, which don't even touch NCQ. Please take a look at latest pcper review with the IOMeter values : http://www.pcper.com/reviews/Storag...inx-zeroes-competition/IOMeter-v20060727-IOps With IDE mode, your results are every time that of the column with 1 thread. And as the description says :
Let me guess, the difference between "almost same" results was in 4k-64thrd, right ? Well, we are back to my explanation again - every other test is single threaded, which in many cases is not what happens in real world. And this is the point of NCQ with SSD... And don't tell me it is pointless.
why are you ranting? I didn't say it was pointless anywhere? as far as i can see your original explanation said without AHCI an SSD would be limited to SATA I speeds LOL
This is actually true. But it depends on your SATA controller. My old computer if I didn't enable AHCI, I got locked down to SATA-1 speed. Like I kid you not, you run any benchmarks you want, it will not pass SATA-1 speeds. I recall even talking about it on this forum, while for some people, they didn't see any difference. It really depends. But as NCQ and eSATA was a feature that people would want, to simplify everything, we (and the rest of the internet) says to set the controller to AHCI mode to get the full performance of their SATA drive.
I am not ranting, i am explaining why when you look at single threaded benchmark you don't see the whole picture and by mistake you (and many others) think IDE mode has same results as AHCI, while cripling the multithreaded performance. And i apologized for the fact, that with the latest controllers this is not true anymore. But i am in same position as GoodBytes, i clearly remember that if i ran my SATA controller in IDE mode then my speeds were limited. It was probably "feature" of ICH9 and older (P45 and older).
You said earlier to ignore old systems and totally dismissed what I said then you use old systems to backup your point LOL The system I benched the SSD with ACHI and without on was a AMD board 4 years old I have it here Gigabyte ma770-ud3 Yes there might have been slight differences in the numbers but certainly no reduction to SATA 1 speeds and certainly nothing a user without a bench mark program would notice. anyway this is total thread hijack @OP sorry for stealing your thread.
