Endurance Test of Samsung 850 Pro Comes To an End after 9100TB

>>61093138
i forgot
>WEAR OUT FAGS BTFO

>>61093138
hdd fags on suicide watch

yea we'll see who's crying after 623 years when all your data is gone

>>61093276
>no thousand year ssd

hahahaha get fucked faggots

HDD poorfag subhumans BTFO

File: 1497217815521.png (573KB, 710x842px) Image search: [Google]

573KB, 710x842px

>>61093138

This is NOT FAIR. I can only AFFORD a HDD because FUCKING SSD cost so DAMN MUCH.

>>61093138
Doesn't matter. SSD are more prone to data loss than HDD. Keep your normalfag SSD to yourselves.

>>61093276
But I need my great great great great great great great great great great great great great great great great great great grandson to watch my hoarded tv series and porn when I'm gone.

File: seagate ssd 60tb.jpg (53KB, 696x393px) Image search: [Google]

53KB, 696x393px

Why haven't YOU ordered your 60TB SSD yet?

>>61093419
It's a great feeling that if I ever adopt I can give my porn to my 9xgreat grandson. It can be a family heirloom, and in 600 years it'll be on antiques show as a priceless piece of history- the amazingly valuable porno collection of some asshole

>>61093138
not a surprise since the 830 pro also had ridiculous endurance (in the 5+ PB range)

Are intel MLC SSDs also this good?
Got 480GB model, always keeping 240GB space free for TRIM and wear leveling. Writing daily maybe 5GBs.

>>61093470
One day the archeologists are digging for a SSD full of porn, the Holy Grail mentioned in Bible 5.1.

>>61093443
That's the size of a 3'5? How much will it cost.

60TB man, wtf.

>>61093921
Uhm sweetie, SSDs lose data if it hasn't been powered on for a while (read: weeks)

>>61093138
What is the best way to make them last as long as possible in an every day machine. Games, internet, movies.

What are the do's and do nots of SSDs?

>>61093953
Disable superfetch, put page file on a HDD, appdata on a HDD and all browser activities on a HDD

>>61093564
Intel = Crucial

>>61093564
Do you keep that space unpartitioned? I've read that this helps for the SSD wear leveling for some reason.

>>61093953
The best way to avoid a ton of extra writes is to set your browser cache to a ram disk.

>>61093138
that's one full year of constant writes at 2gbps until the drive breaks. half of that if you use 4gbps pci e 3.0 nvme drives

considering that PCs are all turned into botnets at this point, i'd rather not have to buy a new overpriced as fuck pcie SSD after just a year at worst, or 3 years at best.

>>61093443
Because I still haven't bothered to upgrade my two 300GB spinning drives to 1TB, and my 250GB C: SSD is fine

>>61094004
>Constantly generating a 2GB/s data stream through your SSD
What are you even doing with it? Are you trying to use it as RAM or something?

>>61094078
>b-but muh optane

>>61093967
Why even buy one?

>>61093969
I'm talking about old Intel 530. it was the last produced SSD with Intel NAND chips.

>>61093138

From my work I have ordered x number of 850 Pro SSDs (can't say sorry.)

Right now in front of me on my desk are 4 dead 256GB 850 Pro. All of them died after about 1500 TB of writes. Way less than what guru3d claims (yes I know 512gb vs 256gb, but 512GB's rating is only 2 times the 256gb.)

I am satisfy with them. I bought them with the assumption that they would eventually burn out under my usage pattern. I've already replaced them with new 850 Pro.

But should I warranty the dead drives?

By the way, the Intel S37xx SSDs I have are still kicking after twice as much writes as the 850 Pro. Of course they are more than twice as expensive and are much faster...

>>61093342
iphoney logic

>>61093564
Intel's enterprise SSDs actually have a firmware lock. You can't run the drive into the dirt, as soon as it hits the designated write load (which is incredibly small compared to normal NAND wear) it'll turn into read-only permanently
Intel calls this a feature.

>>61093943
Are you in 2013?
1 year is standard nowadays even without a discrete capacitor. With a decent capacitor or two that jumps to 2 years easy

>>61094463
730 does too
Still the reel Intel controller

I posted >>61094485

>>61094490
That's actually not entirely correct. If the SSDs are sold thru an OEM, then the OEM can ask Intel to modify the firmware's behavior to fit whatever warranty/performance criteria the OEM wants.

And they do this, ALL THE TIME. I've wasted a few weeks of my life dealing with OEM/Intel engineers on these issues.

>>61093138
IT'S OVER 9000

>>61094534
Their enterprise ssd don't get sold to OEMs they get sold to clients

>trusting a drive with zero tolerance for failure and minuscule parts and disks spinning at 7200rpm

>>61093138
Clearly fake. SSDs are trash that die after 6 months.

>>61094485
>But should I warranty the dead drives?

do you want free stuff or not

don't understand why people don't abuse warranties more, I guess it keeps the deals better for those who do I guess

>>61094534
How do you disable the write lock?
You gotta ask Intel politely?
Of course the OEM can tell Intel what they want to buy that's the point of being OEM, I'm talking the enterprise SSD that get sold

>>61093443
because it's not for sale

>>61093407
>SSD are more prone to data loss than HDD
sourgrapes.jpeg

File: 1494107702447.jpg (58KB, 547x547px) Image search: [Google]

58KB, 547x547px

>>61093443
>seagate

>At that average of 40-gigabyte daily usage, that SSD would have lasted 623 years

HDD BTFO

>>61094575
The cells used for over provisioning are used up. If you want a chance at data loss, ask Intel to disable it or something. It's not arbitrary, and enterprise sure loves it when they have a warning on failing hardware.

>>61094485
did you ever hear about the "discard" and "noatme" mount options?

>>61094078
ask microshit why my drives are constantly being read/written too, especially when it's idle.

>>61094575
1. Have very good reasons
2. Buy enough shits from the OEM to make them care
3. Void all possible warranty claims

The Intel drives I am talking about are Intel storage SSDs that you can add to your $20k server from a Tier-1 server manufacturer.

All the SSDs Intel, not OEM, branded.

File: Toshiba-Roadmap-QLC.jpg (595KB, 2000x1312px) Image search: [Google]

595KB, 2000x1312px

>>61093138
Holy fucking shit my dick. Affordable and big SSDs fucking when? I want a 16TB SSD instead of 4 loud and heavy HDDs.

>>61094663
>toshiba
>deskstar ssd extreme

>>61094631
This suggestion is the equivalent of saying 'try restarting your computer' when your app isn't working.

FWIW, these options won't save the drives under my usage pattern.

>>61094618
I never once judged the write lock good or bad I'm just saying, expect your Intel server SSD to become unusable much earlier than other SSDs that you can run into the dirt
It's nice for RAID when you can freeze and rebuild a drive as soon as there's a data loss risk but for the normal consumer, you're just gonna see a pathetic write endurance and lockdown that may or may not be necessary
If Intel band on controllers are better than everyone else they should have no issue pushing back the write lock (because even though it's understandable, it's still a fucking small write endurance, artificially limited or not)
I haven't checked the Intel write lock number but last I remember it was like 1/3 what a normal Samsung SSD would reliably last until dying

>>61094639
You're not talking about an SSD though, you're talking about Intels customer service to OEMs
If you are talking about SSD tech, what are you talking about? I can't tell from your previous posts about company prqctices

>>61094750
What happens is most companies buy their servers from a server OEM with everything configured (RAM, SSDs etc.)

The OEM provides, say, 5 years of warranty on all the parts, including the Intel enterprise SSDs.

When a customer has issues with the SSDs, first the OEM engineer will handle it, when they determined they can't fix it themselves, they escalate the issue to Intel and Intel brings in their own SSD engineers.

Then all 3 sides try to debug and solve the issue together.

File: image.jpg (27KB, 480x480px) Image search: [Google]

27KB, 480x480px

>>61094618
>warning
>arbitrarily bricking your SSD at a fraction of its expected lifespan
th-thanks intel

File: 1317757009786.png (187KB, 409x338px) Image search: [Google]

187KB, 409x338px

>>61094663
give it another 30 years. they have to milk us extra hard first. the manufacturers are clearly colluding to drop better ssds at ridiculously high pricepoints one year at a time.

>>61094638
>>61094004

>>61094904
You can still read your data.

File: 20170401_204305.jpg (389KB, 1156x863px) Image search: [Google]

389KB, 1156x863px

Where I would put my SSD/HDD caching solution- IF I HAD ONE

Really I think SSD caching is where it's at. Those SSHD drives are retarded because they only put .25 cents of NAND in them. If they sold a 4tb drive with 128gb of NAND(~$45), or if a really good caching tool came out and you could do it yourself people would stop buying these insane 1 tb ssd's.
I almost fell for the PCI-E(960 evo prob) boot drive because "there is no way something that fucking fast isn't better than my tlc drive" but it really only boots windows 1 second faster and is negligible for most software unless you're working with large files which are stored on the pci-e drive. I'm not the only one either, the 256gb $127 960 evo sells like fucking crazy on amazon yet in a controller experiment I don't think most of those customers could tell the difference between a budget SSD and the 960 evo.

>>61094931
only once. After power cycle it kills itself.

>>61094931
For how long though?
I read it works read-only for like one boot cycle but that was a really old article from a while ago
Not trying to argue, I love my 730 as much as anybody would but Shirley you can see why this system could be annoying for consumers not using RAID of sensitive 24/7 data

>>61094949
>t. someone who doesn't actually know how big of an improvement an SSD is but talks like he does

File: c2bf780b57bcbf50b5d1e5b9d77bb511.png (7KB, 964x145px) Image search: [Google]

7KB, 964x145px

>>61095002
I wish you have just pointed out where I'm wrong instead of saying some stupid shit like that. A sufficiently big SSD cache should be nearly indistunguishable from all SSD storage. I don't know what the ratio is supposed to be, maybe it's 128gb to 1tb. If you can barely tell the difference and have a 500gb ssd caching a 4tb HDD that would be a way better solution for a bunch of people.
Intel's software for optane can actually do this but shortly after launch they changed something so it only works with intel SSDs and only SSDs below a certain size.
I bought into SSD in 2011, which is part of the reason why I wanted to buy into PCI-E storage(I like to stay ahead of the curve with storage).

>>61095197
>buys crap-brand SSD
>whines that it isn't good
Okay pal.

>>61095233
no anon, he's saying that a big HDD cache is just as good as using only ssds

>>61095233
When did I say my SSDs aren't good? They're great. I'm just saying that if I bought a 4tb HDD and used my 500gb SSD as a cache for it that I believe it will feel like a 4tb SSD. I bought that one in the picture in 2011 and it still actually works. I'm fairly certain I've been using them longer than you. Right now I have a 500gb SSD as my boot drive and a 128gb ssd as a boot drive in my laptop.

>>61094930
that's not collusion that's just captialism.
what's the sense in spending gross amounts on development only to have your competitor release a better product immediately after?
These manufacturers maximize profits by making these yearly baby steps, it's less efficient for them to blow their load one year and then have nothing to garner sales the next.

File: ssd.png (258KB, 900x650px) Image search: [Google]

258KB, 900x650px

>>61093138
>tfw my SSD will never need replacing

>>61095279
it won't, the cache can't read your mind, and if you're doing it manually it sucks

>>61094568
I don't know. I have one in a 24/7 server currently in his 5th year and one in my laptop (daily driver) in his 3rd. But maybe some jews keep replacing them?

>Lasts longer than you or Western civilization will
>No need to defragment
>lower noise
>lower power consumption
>no rotational velocidensity to worry about

Looks like poorfags are done for.

>>61094568
t. bought the first ever ssd that came out

>>61095304
Well it would cache the most used shit of course which means you'll have to use it for a while but the additional option to force it to to cache some things would be great too. So you have your OS cached of course and all of your desktop software since it's so small, then it will shift all your other files back and fourth depending on what you use.
I've never tried a caching solution because I've heard the current software is absolutely terrible.

>>61095293
>groups of different corporations and retailers
>all working together to drive up prices of even overstocked products
>retailers decide it's time to jackup the price just for the sake of jacking up the price
>all retailers do so at the same time
how is that not collusion?

>>61095302
You're using an Evo not a Pro
It's a generation before the tested one
128 gb won't last as long as a higher capacity drive

>>61095350
because it's happening naturally, there's no colluding between them.

>>61094443
You asked how to maximize life, that's done by minimizing writes. Put your installed applications/os on ssd and all cache things on disk.

Or just put it all on ssd because 9 petabyte write endurance don't give a fuck.

>>61095277
He's wrong and the proof is easily found and long-standing

>>61094967
Put it in another system

>>61095347
>oh boy, I want to install this 40GB thing
>the cache didn't read your mind, so it's only on the HDD
>you wait 1+ hour to install it

>>61095302
>This SSD is NEVER obsolete!

>>61095375
I know, he doesn't

>>61095347
>I've never tried caching
There we go

File: 3.png (82KB, 810x353px) Image search: [Google]

82KB, 810x353px

>>61093276
>>61093443
Except flash memory is inherently volatile and leaks electrons when left unpowered. Each write to a cell decreased its longevity in regards to retaining electrons, and can be made even worst in relation to operating and power off temperatures. SSDs and flash memory in general is only good for instances where IO is necessary, e.g. a bootdrive holding OS. But for storage, it's worst than pointless. Having data on one of these X TB drives means eventually you'll have to transfer the data to magnetic/other storage if you want to archive it. The endurance of of an SSD is moot when turning if off for a couple weeks will corrupt data.

>>61095353
960 pro came before 960 evo

File: hdd.png (248KB, 900x650px) Image search: [Google]

248KB, 900x650px

>>61095353
>tfw the SSD has only 12TB written to it after 2 years of use
>tfw I still have to use an HDD if I want to data horde because I need to sell my firstborn if I want to buy another SSD

File: 1471597312770.gif (911KB, 227x175px) Image search: [Google]

911KB, 227x175px

>tfw you buy a good SSD at clearance for cheaper than budget SSDs of half the size

>>61095417
SSD power off endurance has been like a year standard for a few years now, with many claiming up to two years
A capacitor makes this practically a non-issue, whoever is cold-storing SSD drives should already know about the better options for longetivity

>>61094663
>I want a 16TB SSD
Not for archival, I hope? SSDs are not an upgrade to HDDs like manufactures want the consumer to believe. It's a different technology, and is unsuitable for archival purposes; flash storage will eventually leak electrons, and this increases to a worst case when power-on temperature is low, power-off temperature is high, and cells are worn. Magnetic storage doesn't have this problem.

>>61095407
And you have? I doubt it.
I just want to try it.

>>61095387
Wouldn't it be possible for the software to leave or even dynamically allocate part of the cache for writes and then re-write the file to the HDD in the background?
There is a bunch of cool shit you could do really. Thats why in my original post I said we're lacking a really good caching solution.

>>61095548
ssd caching for files coming from external storage exists and is widely used, but it doesn't do shit to the files on the internal HDDs

>>61095492
>>61095417
>magnetic shills

>>61094663
>128TB HDD
how the fuck

>>61095580
I'll give you an example:

>you have a desktop with an 500GB ssd
>you want to send 40GB of small files to your NAS through your LAN
>instead of waiting 10 years because HDDs have shit random read/write, the ssd cache will suck up all those files quickly and then write it to the hdds

>>61093138
>Open up my Samsung magician
>840Pro, 15.8TB data written
Feels good.
Yes, I am aware that 840 and 850 are different, but the capacity of writes should be around the same neighbourhood maybe.
Owned the SSD since... April 2013.
Feels good.

>>61095632
SSD, not HDD.

>>61095469
>many claiming up to two years
This is undoubtedly a best case/average use case. It's not a static number, it depends on both operating and power-off temperature as well as cell wear.
>whoever is cold-storing SSD drives should already know about the better options for longetivity
But the consumer has no idea because it's not let know to them by the manufacturer. What happens when fill up some large SSD - do you transfer to an HDD/other? In which case you may as well have used an HDD in the first place. But many won't even bother, and if using an external SSD enclosure, will just put it off to the side to rot. I just don't understand the point of large capacity SSDs being touted or marketed for general use storage where IO is a non-issue.

Comment on that site says about 25% of SSD's failing after 3 years. Not due to writes but due to the hardware dying. Anyone got more info on this? I have an 840 Pro that is 3 years old.

>>61095666
the controller will probably shit itself way longer before the cells, satan

I'm not getting an SSD just to replace it after just 5 years.
This is worse than I expected.

>>61095701
It's pathetic you don't have one already.

Manufacturers could sell us 10TB SSD's for $200 RIGHT NOW. But then they would not have a road map to milk us dry for the next few decades.

They have been colluding since forever.

File: dance.gif (460KB, 257x200px) Image search: [Google]

460KB, 257x200px

>9.6 TBW since July 2016
>850 EVO 500 GB
>It will last at least 150 TBW according to Samsung
>that means at least 15 years of use
>tfw it's actually 60 times that

>>61095701
>I'm not getting an SSD just to replace it after just 6 centuries

>>61095701
>after just 5 years
Dude, as I just mentioned, I got my SSD back in the April 2013, I've only written 15.8TB and while I am sensible with it, I'm also not scared of using the SSD.
OP's quote specifically says that the weakest SSD's lasted for 187 - 280TB data written.
Roughing it out to 4TB/year on the writes, that's 46.75 - 70 years worth of service - on the weaker SSD's.


You DO NOT have to replace an SSD in your lifetime unless you've done something seriously wrong with it - and most, if not all by this point, automatically deny (at least Windows) the ability to defragment them due to the damage that causes.

>>61095772
Blame Apple for their NAND hoarding

>>61095799
>automatically deny (at least Windows) the ability to defragment them due to the damage that causes

It causes no damage
https://www.hanselman.com/blog/TheRealAndCompleteStoryDoesWindowsDefragmentYourSSD.aspx

>>61095836
Well, wasn't aware of that. Thank you very much for putting me right on the matter.
At least I was half right in that it doesn't do a usual defragment like it does for HDD's, but I wasn't aware it did a once a month 'housekeeping' style defragment. Pretty cool to know, actually.

File: ssd.png (288KB, 901x663px) Image search: [Google]

288KB, 901x663px

Mine is still running fine.

>>61095799
>Windows
Not sure about Apple, but other *nix systems at least generally use file systems that don't need defragmenting in normal use.

File: ssd.png (251KB, 899x650px) Image search: [Google]

251KB, 899x650px

>curious how many TBs have I written in a year
>expecting 20
>only filled 16 times over
H-how? What sorcery is this?
This is my only drive I use for everything.

>tfw you have a 850 pro
Feels good man

my old SSD is an 840 pro (still in use), new one is an 960 pro.
I expect both to work until they're obsolete for capacity reasons.

>>61095364
>it's only natural to jack up the price of outdated tech which you have a massive overstock of
>at the same time as all the other retailers doing so
lol...

>>61096468
The market dictates the price, moron. If nobody bought them, they'd drop the price.

>>61096250
*suddenly dies*

>>61096487
*grapes sourly*

File: 1471679350910.jpg (30KB, 450x450px) Image search: [Google]

30KB, 450x450px

>>61093276

>>61096477
no.
>i have product thats cheap as fuck to produce/store
>instead of selling at relevant prices
>i'm going to hoard it until someone makes a massive publicity stunt about it (ie start a fire in the factory producing it)
>at which point we all jack up the price and then sell it at this new inflated pricepoint
>in the meantime we'll keep the price high so nobody wears down our stock for when something happens, we can make the most profit off people who simply can't wait till we slowly drop the price back down a little bit
storing ram is easy as it takes up very little space and barely degrades at all, which is why/how these retailers have millions of sticks in stock.

have you not noticed how DDR4 has been out for over a year now, yet DDR3 prices haven't gone down at all

File: 14468746878.jpg (7KB, 193x115px) Image search: [Google]

7KB, 193x115px

>>61096038
>477GB
>only 16GB free
You're currently fucking it hard.

You need to let at least 10% free on a SSD for it to work properly.
Less than 10% free increases the wear a ton.

>>61096640
not him, but why does that happen?

>>61096683
so there's a log of what you do offline kept until you eventually connect to the internet again

same reason why HDDs performance drop to usb 1.0 speeds when you fill it past 85%

>>61096640
I'm not so sure this is such a huge issue right now. Most mfgrs have ample over provisioning to help mitigate this.

>>61096716
log likelihood

why cant just setup system on platter. get all data up to date as possible. then copy to ssd. use ssd to boot only, and any new writes go to platter with the exception of updates to games/op sys.
is this not possible?

>>61096744
why do you want to over-complicate a simple thing? install OS and games on ssd, throw other files on HDD

if you try to do that thing you'll be prone to a lot of tweaks on paths and subsequent bugs

>>61096716
Full retard.

>>61096811
games are like 100gbs now, fuck that shit. games go on HDD

>>61097201
Full retard.

>>61096744
install OS to a SSD and set it up to do weekly backups to a HDD.

>>61097201
If you can afford so many 100+GB games then you can afford a new, larger SSD.

>>61097201
show me 5 non AAA games that are so big

>>61097259
Fuck that noise, I'd like to see 5 AAA games that are 100 gbs.

>>61097299
search for dx12 games
pretty much everything on that list

>>61097322
Still waiting for an answer.

>>61097367
>being this dense on purpose
anon this is not how you troll people
unless you want to be thought of as an idiot

>Not buying HDD over SSD
>Wanting 1 Second Start Times for Games and Apps
>Unrealible, despite all the BS charts
>Muh Speed

My PC I bought last year, an i7-6700k, 16GB DDR3, 2TB HDD RX 470 Windows 10 Home. Starts up in 8 seconds. Games load in around 5 to 10 seconds. Get into Apps on average 3 or 4 seconds.

Why buy an SSD? Its a joke.

My current PC I bought recently. Ryzen 7 1700x / 32GB DDR4 / 2TB HDD / RX 580 Windows 10 Home.
Starts up in 7 seconds. Games load in 4 to 9 seconds. Get into apps in 3 or 4 seconds.

Both PCs run everything "game" related no problems. Fast, cool smooth. Do 3D modeling / rendering on both. Done fast, cool with the ryzen cpu being a tad faster.

Yeah. Fuck SSDs its a meme and a waste of cash, like led lights and liquid cooling.

File: 1475829842040.png (249KB, 900x649px) Image search: [Google]

249KB, 900x649px

40TB already?

>>61093407
>citationfuckingneeded.jpg

Never had an ssd die, never hear about them dying. Have had many many hard drives die on me.

File: 1462588198804.png (5KB, 205x166px) Image search: [Google]

5KB, 205x166px

>>61093138

I wish they didn't make this public, if anything this will motivate manufacturers to decrease the quality of their product

>>61097417
Thanks for conceding your point.

>>61094990
Not all SSDs even go into read only mode, and all that do fail after a power cycle.

If your OS drive goes into read only mode you wouldn't even be able to mount any sort of media to copy files off.

Good job I backup

File: 1496864018626.jpg (53KB, 768x432px) Image search: [Google]

53KB, 768x432px

>>61093407
I fix PCs. Never seen a SSD die in my 3+ years experience. Get at least 1-2 bad HDDs per week.

>>61093138
And how long do HDDs last?

Seriously, I'm glad to see the life length SSDs have, but is that somehow better compared to HDDs?

Unless HDDs last only 20 years before breaking down, I feel that we can do as we did. Boot off of SSD and let HDD take the bulk of the data.

File: 1484973422727.jpg (10KB, 319x370px) Image search: [Google]

10KB, 319x370px

>>61097469

You can bet your fucking ass they will go the path of Intel soon, meaning the SSD will go into read-only mode as soon as the factory write cycle number is reached.

Intel seems to be really good in making monies....

File: .png (46KB, 458x481px) Image search: [Google]

46KB, 458x481px

>>61097463
>AHCI Mode: Deactivated
Why?
>238GB (11GB Free)
To a drive maker, 1GB = 1,000,000,000 (1,000^3)
To Windows 1GiB= 1,073,741,824 (1,024^3)
You can use a ratio of 93% to convert: 93% of 256GB = 238GiB
You have 0% overprovisioning, which wears down the drive more since you write to it a ton. Might just want to upgrade to a .5TB/1TB SSD at this point. And keep around 10% OP for better endurance.

>>61099058
Also, that chart doesn't account for the GB/GiB difference.

People are still scared of SSDs dying? It's not 2010 anymore.

>>61096151
>AHCI not in use
It's like you don't even want the speed benefits

File: 989271.jpg (31KB, 584x730px) Image search: [Google]

31KB, 584x730px

>>61093138

>mfw I didn't fall for /g/'s "He fell for the SSD meme" meme

>>61099990
How many levels of meme are you on right now?

>>61093138
I never had an issue with the PRO line.

EVO's are TRASH

>>61093138
What's happening to MLC drives? Today I can't find newer drives that aren't TLC!

>>61100542
>no queued trim
>needing kernel blacklisting to avoid issues
PROS are also TRASH

>>61100779
What are your opinions on Toshiba/OCZ PCIe/m2 drives. You said 2 things I don't understand and that makes me think you're smart. I use mushkin SSD's personally because I've always had good performance and no failures. But I feel inclined to get PCIE storage.

File: 1445721360402.jpg (13KB, 200x255px) Image search: [Google]

13KB, 200x255px

>>61094004
>>61094638
fuck, really?
I was looking into running Win 7 on a 960 Evo

>>61094605
>seagate
>bad in 2017
keep perpetuating this meme, it'll just allow me to buy Seagate drives for 20% less than WD because their reputation is worse

>3tb Seagate HDD costs the same as a 1tb WD Black

>>61100802
TRIM was originally a non-queuable command and you couldn't mix it with R/W ops. Then they introduced queued TRIM.
The problem was, that some controllers had buggy implementations, yet they still reported its support to the OS and when the OS issued a queued TRIM command... the drive freezes or suffers data loss.
Linux kernel devs had to blacklist some drives (including Samsungs) because of that. Some manufacturers (Crucial) issued FW updates to avoid problems.
https://blog.algolia.com/when-solid-state-drives-are-not-that-solid/

>>61096583
That still doesn't make it collusion.

>>61100973
you're right i'm stoopud

>>61093943
>(read: weeks)

Don't you need a temperature of 40C+ constantly for this?

>>61093138
Huh, I remember SSDs being shit in 2012. Looks like I'll get one next time instead of the ol HDD only.

Granted, 5 years for a measly 40GB write a day is atrocious, the 60x figure is quite reasonable indeed.

>>61094638
Is M$ using your drive to cache their own data sort of like a rootkit + bittorrent?

What if they used customers' own computers to store updates and distribute them to other customers?

Is this why Winblows is so insanely bloated in comparison to OSX and Linux?

brb, gotta get the tinfoil from the kitchen...

(but seriously, fuck M$)