|
Uterus Poker posted:My question is: is the "proper partition alignment" thing listed there worth it or not? The proper partition alignment thing is the only thing there worth listening to (although the BS about aligning to the erase block size is pointless, you just need to make sure you get the 4kB sector alignment).
|
#
¿
Dec 14, 2011 18:26
|
|
|
|
| # ¿ Apr 26, 2024 14:36 |
|
|
Aside from response bias, there are a couple other factors as well. Some SSD brands (OCZ) appear to have much lower QC standards than any HDD manufacturer, so they let far more manufacturing defects out the door. The vastly higher performance of SSDs makes for a heavier load for other components, and can expose flaky Marvell SATA controllers. Unfortunately, it's incredibly difficult to tell the difference between a flaky drive and a flaky drive controller.
|
|
#
¿
Dec 19, 2011 00:59
|
|
|
I think the op is being a bit harsh on 120GB. 40GB is fairly tight with Windows 7, but you can do it if you aren't installing much software, and you can keep one or two reasonably large games on a 60GB drive. A lot of games don't see much benefit from an SSD anyway, so I haven't found it to be much of a loss when I have to install a game on a mechanical drive. It's nothing like the night and day difference you get from putting the OS on an SSD.
|
|
#
¿
Dec 25, 2011 06:26
|
|
|
Tab8715 posted:I don't know if this has been answered or not but why the hell is flash memory still so expensive Because it has to be manufactured using discs of pure silicon with fewer than 1-2 parts per million impurities, in factories that cost several billion dollars to build.
|
|
#
¿
Dec 27, 2011 08:00
|
|
|
Your QD=32 times aren't really any better than QD=1, which means you aren't getting NCQ, which means AHCI isn't enabled. You should do that (instructions are in the OP). If your motherboard has two drive controllers, you can probably speed up POST time by disabling the RAID controller (which is probably some crappy Marvell controller you don't want to be using anyway). Tab8715 posted:I kind of want to get a 128GB Drive, but a nearly a $1 per GB it's not worth it? I paid close to $8/GB for my first SSD and it was totally worth it.
|
|
#
¿
Dec 27, 2011 17:06
|
|
|
The SSD doesn't get to pick where to read from, but it gets to pick where to write things to. It can pick the fastest/most convenient place to write to, it can cache/buffer things, and in the QD=32 case, combine all 32 queued 4KB writes into one single 128KB write.
|
|
#
¿
Dec 27, 2011 19:43
|
|
|
The 0:00-0:26 is POST related stuff... check for BIOS settings relating to memory checks/fast boot, there's probably something you can turn off to speed that up. 0:31-0:59, try taking your CD drive out of the boot order, updating your BIOS, or shaking your fist to the heavens and cursing "drat you Gigabyte!". A good part of the 0:59-1:43 is probably one or more slow to initialize drivers, but I don't know any way to diagnose which ones without using tools that are way beyond the average user.
|
|
#
¿
Dec 27, 2011 22:12
|
|
|
tzirean posted:There's nothing in there that I can see that would affect boot time. Latest boot took exactly ten seconds longer than previous, all of that coming during the mobo splash. Advanced BIOS Features > Quick Boot should be set to enabled. tzirean posted:I have two settings that can be RAID, but both are AHCI. PCH SATA Control Mode (can be IDE, RAID(XHD) or AHCI) and Onboard SATA/IDE Ctrl Mode (can be IDE, AHCI or RAID/IDE). Integrated Peripherals > Onboard SATA/IDE Device should be set to disabled.
|
|
#
¿
Dec 28, 2011 17:22
|
|
|
Dogen posted:Since video editing is a sequential task, and raid 0 only improves sequential performance, I'd say you would see some benefit there, though I don't know how much. RAID 0 will improve random performance as well on SSDs. It doesn't help on mechanical drives because of seek times. That said, SSDs get such good random performance already you're very rarely disk bound on random accesses anyway.
|
|
#
¿
Dec 30, 2011 16:05
|
|
|
Sleipnir posted:but sequential write speed is only around 160 Mb/s... What's going on here? The benchmarks you've seen are probably for 128GB or 256GB drives. Your 64GB works internally like a RAID 0 array with 8 drives in it. The 128GB+ drives work like a RAID 0 array with 16 or more drives. This doesn't matter for reads (which are already bottlenecked by other factors with only the 8-way spread), but for writes, you only get half the throughput you would with a 128GB drive.
|
|
#
¿
Dec 30, 2011 20:29
|
|
|
icantfindaname posted:I'm pretty sure the difference in data is because the capacity on the box is measured with regular decimal metric prefixes, where 1GB equals 1000MB, and on your computer it's measured in binary prefixes, wehre 1GB is 1024MB. So you aren't losing any capacity. Correct me if that's wrong though. Well, there are 128 binary gigabytes of flash memory in the drive. The other 9GiB is spare area used for wear leveling, to extend the lifespan of the drive. That said, I'm betting it's not a coincidence that the spare area set aside happens to be pretty much exactly the same as the difference between binary and decimal gigabytes.
|
|
#
¿
Jan 1, 2012 02:53
|
|
|
I'd recommend the Intel 320 120GB for $200. Doesn't do quite as ridiculously well in benchmarks, but makes up for it with reliability and a good warranty.
|
|
#
¿
Jan 2, 2012 01:17
|
|
|
It is. But TRIM is not nearly as big of a deal as people make it out to be. Yes, write performance will degrade over time as the drive goes from being pristine to well used. Depending on the drive, it can sometimes be a fairly significant degradation. But for most users, even in the worst, most degraded state, the SSD is still at least an order of magnitude faster than any mechanical drive out for the considerable majority of disk operations. And for some small portion of your usage (likely well under <10%) it might end up being a bit slower than the fastest HDDs available. And that's not even considering the garbage collection that a lot of drives use to keep up performance without TRIM anyway.
|
|
#
¿
Jan 2, 2012 06:36
|
|
|
whiskas posted:My understanding is that SSD's never lose the ability to read data, so even if the drive malfunctions the data is still recoverable, whereas if a traditional hard drive malfunctions either the data is lost or you have to pay $$$ to recover it. The "never" part has already been pretty well addressed, but I wanted to point out that the "if the drive malfunctions" part is completely wrong as well. SSDs are supposed to still be able to read data if they use up all of the available write cycles without malfunctioning. If the drive actually malfunctions, your data is far less accessible than it would be with a traditional hard drive, not more. You would have to pay $$$$ to recover any data at all (from a non-sandforce drive), and probably $$$$$$ if you want any reasonable chance of recovering a files larger than 4KB or from a Sandforce drive.
|
|
#
¿
Jan 6, 2012 01:19
|
|
|
My first gen OCZ Agility did something similar. I found I was able to read from it just fine, but any attempt to write to the drive would lock up. I imaged the drive, and then at OCZ's request ran a secure erase on the drive. That fixed it up and it's worked fine ever since. It must have been a firmware issue, so giving it a clean slate cleared up the problem. Edit: quote:EDIT: So Clonezilla is spitting out a lot of I/O errors so there is no prospect of performing an image backup. It does seem that copying the data off is possible at least. You can try taking an image with ddrescue.
|
|
#
¿
Jan 7, 2012 05:13
|
|
|
BlackMK4 posted:CF card + ide/sata converter? Not really a very good idea generally. Windows will refuse to install to most CF cards, if that's what you're going to use, and they tend to be pretty slow with firmware that's not optimized for good random access.
|
|
#
¿
Jan 9, 2012 03:16
|
|
|
Your winsxs folder isn't actually as big as windows explorer says it is. Most of the files in it are just hard links (or even multiple hard links) to files that exist elsewhere on the file system, so even if you deleted them from winsxs, it wouldn't actually free up any disk space. stealth edit: that command is a legit way to free up real disk space, though.
|
|
#
¿
Jan 12, 2012 16:19
|
|
|
Treytor posted:What's the general consensus of enabling or disabling write-back cache in Intel RST with SSD raid-0 setups? If you enable it, you should consistently get SSD level write performance to your cached drive(s), which is pretty good and could easily be a noticeable, significant performance boost. The downside is that now your SSD becomes a necessary part of a drive array; if the SSD dies or whatever, hello file system corruption. Essentially, you're creating a RAID 0, albeit with much higher chances of recovering data if the SSD does fail. I don't think there's any sort of consensus, it's up to you decide if you are willing to trade a higher risk of data loss, and inconvenience should you want to move a drive to another system, in exchange for significantly better write performance. dud root posted:The winslx command didnt work for me- I used a x64 Win7+SP1 integrated install, could that be why? Said something about service pack files not found Yeah, if SP1 is integrated then the files that command deletes were never there in the first place.
|
|
#
¿
Jan 13, 2012 00:52
|
|
|
The SAS SSD is problably SLC. SLC is lower density, but has much better write endurance; MLC drives generally aren't appropriate for write intensive workloads.
|
|
#
¿
Jan 18, 2012 18:37
|
|
|
The Whiz Kid posted:It'll work, but it can be finicky. Well, it'll probably work. There are a couple small bits of the filesystem that cannot be moved while the filesystem is in use, and almost no defragging tool can handle moving it offline, either.
|
|
#
¿
Jan 22, 2012 20:44
|
|
|
It is supposed to work well enough. But either way, zeroing the drives won't help; you'd have to do a secure erase.
|
|
#
¿
Jan 27, 2012 02:25
|
|
|
It looks like that A-Data is a rebadged Crucial C300. It will perform significantly worse than the Force 3 in synthetic benchmarks, but in actual use you probably couldn't ever tell the difference between the two side by side (In Anandtech's benchmarks, it adds up to a 70 second difference in the total disk time to simulate 2 weeks worth of heavy usages).
|
|
#
¿
Jan 28, 2012 02:56
|
|
|
mobby_6kl posted:The next tier up is about $40-50 more for the same capacity, but is it worth it, in terms of real world performance? Not at all. They basically all achieve "you're no longer disk-bound" performance levels. In some workflows, you might be able to tell the difference between the faster and slower drives in a side by side comparison, but you wouldn't be able to tell otherwise.
|
|
#
¿
Jan 30, 2012 22:55
|
|
|
You can shrink your partition through Disk Management.
|
|
#
¿
Feb 1, 2012 06:08
|
|
|
You can try MyDefrag; there are a few things it will move that Windows won't. If your partition was larger than 2TB and it's not anymore, you might try Ghost again; it may not like large partitions rather than just large drives.
|
|
#
¿
Feb 2, 2012 03:42
|
|
|
Bitch Stewie posted:is there any need/benefit to not partition it so the full capacity is available? As long as TRIM is working, there's effectively no difference between unpartitioned space, and free partitioned space.
|
|
#
¿
Feb 6, 2012 04:57
|
|
|
Windows turns off Defrag and superfetch itself if you run the WEI (and won't readyboost from it either). Turning off the indexing service is retarded bullshit from the early days of the stuttering JMicron controller. Don't do it.
|
|
#
¿
Feb 7, 2012 16:37
|
|
|
You do realize that every OS in the last 15-20 years has that built in and doesn't automatically, right? CDM doesn't perform like that normally because it explicitly disables the OS cache.
|
|
#
¿
Feb 11, 2012 23:31
|
|
|
They are not even remotely fragile. You can wear them out running your production database server off of them, and not much short of that. No consumer should ever need to worry about wearing out their SSD.
|
|
#
¿
Feb 13, 2012 07:40
|
|
|
Bitch Stewie posted:So what's the recommended way to do a benchmark that will simulate worst case scenario for random writes? Don't you have better things to do with your time? Unless you have an actual hard requirement you need to test, just don't worry about it. It's better than obsessing over benchmarks and posting stupid crap like this: AvatarSteve posted:but my 4k is nowhere near that good. Seriously, 80MB of random 4KB writes will take your drive one second, his drive 2/3rds of a second. There's a fair chance you wouldn't be able to distinguish the difference between the two, even side by side. But it doesn't matter, because those same writes would take even an above average desktop hard drive 3-4 minutes. Software goes to great lengths to ensure that kind of activity will never, ever happen, because it would be totally unusable for anyone with a mechanical hard drive. Edit: Inovius posted:Suggestions on the best way to clone one SSD to another so the alignments are good. I'm using an 80GB Gen2 Intel drive currently and just picked up a Crucial 128GB M4 and want to swap it in as my main drive without having to do a full re-install. Clonezilla will work well for that. It's a pain if you're trying to move from a larger drive to a smaller drive, but for the other way around it's easy and will maintain alignment. Zhentar fucked around with this message at 20:02 on Feb 14, 2012 |
|
#
¿
Feb 14, 2012 19:40
|
|
|
Mozi posted:Also, the Intel SSD toolbox is telling me that DIPM isn't supported by my storage driver - should I try to fix this, or is it inconsequential? I'm not often away from AC power but when I am could always use more battery life. From what I've seen, DIPM will save you about half a watt. If you've got an older laptop or a high-powered desktop replacement, this probably won't make a visible difference. On a more efficient, low power laptop, this will be a bigger difference; on my X220, an extra half a watt would be a 10-20% reduction in battery life for light usage. Bitch Stewie posted:Yes and no. This is being used for a database application so a poo poo ton of random writes. It's not mission critical, but speed is important. Anand describes some of his worst-possible-case testing procedure here: http://www.anandtech.com/show/5147/the-ocz-octane-review-512gb/6
|
|
#
¿
Feb 14, 2012 20:01
|
|
|
dietcokefiend posted:The Anand article linked looks entirely at burst speeds and consumer workload patterns... which might only apply for a few hours in your setup. I linked to the article I meant to, although with only a little bit of searching so it may not have been the best example. I was only linking to that particular page of the article though, and only for the procedures described (Doing a sequential write pass across the whole drive, then sustained random 4k writes for an extended period).
|
|
#
¿
Feb 14, 2012 21:55
|
|
|
redeyes posted:But its only been on for 192 hours?! How in hell did you get that much data written in 192 hours? That's only 175MB/s, easy to hit.
|
|
#
¿
Feb 15, 2012 21:07
|
|
|
It depends on what you're doing... if she's trying to play intensive flash games, like Farmville or whatever, then yeah, a better CPU is going to make a much bigger difference. But for most usage, I think getting a good drive in there is way more important.
|
|
#
¿
Feb 16, 2012 16:52
|
|
|
Sporadic posted:Also, is it true that SSD performance takes a hit if you have it almost full up? I'm going to try to pare down what's currently on my hard drive but it's still going to be a tight fit. Yes, it is, but it's not a very big difference, and at that point, your netbook CPU will likely be a limiting factor either way.
|
|
#
¿
Feb 17, 2012 19:19
|
|
|
Shadowhand00 posted:Does anyone else buy SSDs for their work laptops so that they no longer have to experience lag during work? Or do I have a sickness? With my old workstation, I was seriously tempted, but I wasn't sure IT would let me get away with it. My new workstation has two hard drives, and it's been mostly tolerable since I spread things across both drives and short stroked the OS drive (500GB drive partitioned to 100GB).
|
|
#
¿
Feb 22, 2012 21:24
|
|
|
Alereon posted:These drives should have lower random write performance and perform poorly without Trim support and/or when filled with exclusively incompressible data, It's not really clear, but SandForce is doing something weird about caching with the overprovisioned space. We'll have to wait for benchmarks, but it's possible it could actually hurt random reads more than random writes.
|
|
#
¿
Feb 26, 2012 21:26
|
|
|
redeyes posted:Yeah probably a good bet. On the other hand, sometimes sandforce controllers can be recovered by secure erasing. I've seen in happen twice. I think it may have to do with failing flash but can't be sure. I've had a failing drive fixed by a secure erase, personally. Basically, a significant portion of SSD failures stem from firmware issues; doing a secure erase is equivalent to a format & reinstall of your OS. Whatever bad state the firmware was struggling to deal with gets wiped clean and it starts working fine again.
|
|
#
¿
Mar 13, 2012 05:37
|
|
|
You can tell he's running AHCI because the 64thrd score is substantially higher. It looks like he's on 3gbps, though. But regardless, those speeds are more than fast enough for pretty much any need.
|
|
#
¿
Apr 7, 2012 04:18
|
|
|
|
| # ¿ Apr 26, 2024 14:36 |
|
|
It was stated that they're not in it for the long-term. Some people interpreted that as dropping out then and there, but I think there are still at least a couple years left before the market reaches the maturity that Intel was referring to.
|
|
#
¿
Apr 16, 2012 18:36
|
|