SSD = MEME

>>52545255
Well I'm glad I don't turn off and leave my computer unplugged for weeks at a time then.

>>52545274
there's gonna be a time and day when you have to walk out of your house and not return for a while faggot

File: 1405665850513.jpg (40KB, 800x600px) Image search: [Google]

40KB, 800x600px

That's why smart people buy SSDs with 10 year warranties.

>>52545326
have fun sending back your SSD every week, i'm sure the shipping costs will be of no issue

>>52545255
Why would you keep your sensitive data on a SSD?

That only applies to enterprise drives

Good thing I keep data that isn't easily redownloaded on a seperate mechanical drive then.

>>52545401
no you dipshit

>>52545255
I've left my 850 evo and 950 pro off for over a month with no issues ever. Fucking retard.

>>52545326
They will replace your SSD, but the data will still be lost

>Written May 11

Here's another article from May 21

http://www.pcworld.com/article/2925173/debunked-your-ssd-wont-lose-data-if-left-unplugged-after-all.html

>>52545421
Thank you Rajeesh, I'll purchase 5TB SSD for $5500.99 and store my sensitive data on it.

>>52545421

OP Status: [BTFO]

Once again a retard gets a hold of cherry picked information and swims in his confirmation bias so much that he forgets to ensure his information is up-to-date before shitposting about it.

>It was intended to help data center and enterprise customers understand what could happen to an SSD—but only after it had reached the end of its useful life span and was then stored at abnormal temperatures. It’s not intended to be applied to an SSD in the prime of its life in either an enterprise or a consumer setting.

>>52545432
Go ahead! As long as it's not a used SSD past it's lifespan you should be good to go.

File: backblaze2.jpg (26KB, 635x414px) Image search: [Google]

26KB, 635x414px

>>52545421
>In a conversation with Kent Smith of Seagate and Alvin Cox, the Seagate engineer who wrote the presentation that set the Internet abuzz, PCWorld was told we’re all just reading it wrong.

>SEAGATE ENGINEER

HAHAHAHAHAHHAHAHA

Credibility of that article is shit.

>>52545461
>the Seagate engineer who wrote the presentation that set the Internet abuzz
Yeah, the credibility of OPs article is shit

>>52545281
I've had SSDs for years

Always turned the computer off for even up to a month at a time. No issues.

File: 1448815651085.png (568KB, 600x580px) Image search: [Google]

568KB, 600x580px

>>52545461
>>52545421
>“I wouldn’t worry about [losing data],” Cox told PCWorld. “This all pertains to end of life. As a consumer, an SSD product or even a flash product is never going to get to the point where it’s temperature-dependent on retaining the data.”

That's right, if you don't worry about losing data, keep them stored on our Seagate hard drives and on any SSD you can find

OP = FAG

Why has literally no-one had this problem? It's because it doesn't exist. Sensationalism sells and bullshit articles get views.

Why is it, that on /g/, there's so much shitposting regarding SSDs but NO ONE has ever had a problem with them. No failures, no malfunctions, no lost data.

>>52545481
>my particular experience is the norm

Wanna do the test right now? Make sure to MD5 check everything on your SSD and then reaffirm it after one month of inactivity.

Do it

>>52545481
>it didn't happen to me so that means it's not a problem

>>52545416
same

>>52545489
yeah why does nobody purchase 2TB SSDs to store data on

surely it can't be because of the price

>>52545255
It appears all SSD's have this kind of problem, on different time lapses (unpowered standard consumer ssd's data lasts much more than 7 days under regular conditions, we're taling at least a year) but it's still a good idea to know this and take precaution.

I'm not going back to hdd's because of this but I will take backups more seriously.

>>52545513
>using tape to backup

i wish i was rich

>Extremetech

Isn't this just another memesite?

>>52545493
>md5
>not sha1
gtfo

>>52545593
whats the difference and the point is to checksum all of your files and see if anything got corrupted

i dont give a fuck which method will be used, just Do it and confirm no data has been lost after one month of not being used

>>52545617
What if the damage is enough to make MD5 return the same hash of the original file.
MD5 is much prone to collisions than SHA1

>>52545255

I think it's pretty clear at this point:

The meme is that "SSD = meme", but SSDs in themselves are certainly not a meme. In fact I suspect the people pushing this meme are doing so from PCs using an SSD for their system drive, have witnessed the advantages of said drives, and want to troll as many people as possible so that they don't enjoy the same improved user experience as them.

>>52545508
It IS because of the price.
No one needs 2TB of SSD at the moment.

>>52545593
>SHA1
>Not SHA256
Enjoy your collisions faggot

i have a 64gb ssd from 3-4 years ago and i've had it powered down for periods of well over double that

according to this chart, my drive (which idles at 40c) is going to get at least 105 weeks of off-time, likely more because my room is normally 15-20 degrees celsius, not 25.

>SSD = MEME
SSD = MEME is a meme. you're a meme.

>>52545769
>SHA256
>not SHA512

>>52545811
SHA256/SHA512 doesn't really matter, as long as it's based on SHA2 and not SHA1

>>52545255
ITT: Stupid people.
You use CRC to check for data corruption you faggots.

>>52545691
This.
Now I can sell all my disgusting HDD's to people who think they are savvy consumers not falling for the 'SSD meme'

>>52545691
Probably some asshurt Seagate etc. employee.
If they don't have indigenous SSD manufacturing they're going to the dumpster or competing with secondhand repackages.

I don't think my laptop has ever been inactive for 7 days

>>52545909
two Seagate engineers said SSD is safe to store data on you fucking illiterate chink

>have ssd
>power down pc
>unplug from wall
>walk out door
>hike across the U.S. for 5 months
>get back
>data still there
>WTF /G/ LIED TO ME
>REEEEEEEEEEEEEEEEEEEEEEEEEE

I got a question for all you people who are worried about the unlikely chance of your SSD failing:

Have you ever even heard of backups?

>>52545932
The clickbait shit will travel farther than the correction.

Also, nintendo 64 cartridges use solid state storage, and I've plugged them in after years of not playing them and the data is still there.

File: 1451434445287.png (104KB, 480x480px) Image search: [Google]

104KB, 480x480px

>>52545962
HDD can fuck up as well

>trusting news articles posted on /g/
>ever

Yeah fuck you guys. You faggots spread so much disinfo that I've learned to stop clicking links on this board. You pajeet fucks.

>>52545415
Did you even read the article?
> In worst-case scenarios or high storage temps, the data on an ENTERPRISE DRIVE can start to fail within seven days.

>>52545985
so if enterprise can't handle it, the consumer will die even faster

>>52545255
>I dont have an ssd for reasons better start
convincing people they fail all the time

>>52545978
Cool. I never said they didn't.

Have you ever even heard of backups?

>>52545985
That's after it's projected lifespan fyi

>>52545255
this thread is a bait and meme in one. Quite remarkable.

>>52545461
I have been using a 3tb Seagate external as my main hard drive for the last 3 years.
Scared.png

>>52545811
>SHA512
>not SHA1024

f4gg3t

>>52545281
Not for a week straight, plus there is this magical thing called wake on LAN which you can send through the internet. If need be I could simply power on my device and then shut it down after I remote in

>not falling for the SSD RAID meme instead.

https://www.youtube.com/watch?v=eULFf6F5Ri8

File: 1440834565254.gif (804KB, 300x300px) Image search: [Google]

804KB, 300x300px

Are you all this god dam retarded? New SSDs have 4+ years of data retention if left without power at room temperature.

That article was explaining how SSDs under extreme wear and temperature ranges would quickly lose data. It has nothing to do with a consumer SSD in your laptop/desktop staying at room temperature most of the time.

>>52545255
>itt a butt hurt poorfag high school kid, how do i python? how do i dos and dox hacking with the linux operator system? rate my pc im thinking of building for $350 its a 1337 gaming/ hacking build i got the spec from reddit and linus tech tips

File: blog_q3stats_manufacturer-e1444680042365 (1).jpg (60KB, 699x720px) Image search: [Google]

60KB, 699x720px

>>52545461
Another retard posting out of date memes. Backblaze have officially declared that WD are now the worst hard drive manufacturer, so be sure to burn all the ones you bought, goy!

>>52545255
I left a shitty gen 1 OCZ SSD in my hot garage during the summer for 6 months, and it didn't lose anything.

This seems like useless scaremongering to me... it's like warning people that their computer could catch fire and burn the house down if something short circuits in the power supply and dumps line voltage on the 5v rail... sure it's theoretically possible, but it virtually never happens.

>>52546973
>Doesn't use a PC while in a sauna
Pls.

>>52547128
Data retention on ssds is insanely good even under hot temperatures. However as you write to them more they become less reliable.

For example ssds which become read only due to being used in servers 24/7 can lose data in about a month if left in hot temperatures.

Of course the average user would take over 20 years to wear out a 256gb ssd enough to cause the data retention to become less than a year.

Your ssd probably had more than 1 year of data retention left even in the hot temperatures it was subjected to.

>>52545255

Good thing I don't keep any important data on my SSD, only OS and programs which can be quickly reinstalled/redownloaded

>technology board
>supposedly full of people who like good technology
>instead it's full of people trolling and memeing literally 24/7 to make one of the best technologies to come in recent years bad

shouldn't you be busy masturbating to your favorite linux distro or some shit?

>>52547486
op is a troll you dipshit. A typical SSD has around 4 years of data retention while being powered off.

>>52547444
>use
This literally solves the data retention problem.

OP is a MEME

File: saveme.png (2MB, 1920x1080px) Image search: [Google]

2MB, 1920x1080px

Friendly reminder that Flash memory is waterproof.

http://www.digitaltrends.com/photography/panasonic-camera-lost-in-shipwreck-recovered-with-working-memory-card/

>>52546117
>not building a 10 drive machine with multiple raid levels and automatic tape back-ups
It's like you want to lose data, senpai.

>>52549434
Do people really keep more than 2TB of data? I have a 2tb ssd and I only have like 800gb of data on it. I feel like I went overboard.

>>52545420
>what are backups
You do backups, don't you Anon?

>if a drive is stored at 25C or operated at 40C, expected data retention for a client drive is 105 weeks, or nearly two years
Which pretty much all consumer drives are.

>high storage temps [with enterprise drives] [...] data loss in as little as seven days
Basically, the moral of the story is, if you run a datacenter, keep drives cold, and for data that won't be used for many weeks, you're probably better off with tape or a hard disk.

2/10 bait OP.

>>52545461
>>52547042
Blackblaze is a meme anyways, they recover refurb and consumer drives not designed for enterprise workloads, subject them to enterprise workloads in custom built enclosures with poor temp and vibration control, and then rate their failure rates.

When Seagate had the most failures, they were failing because of the bad enclosures and temps, when a couple years earlier they were "harvesting" seagate drives (as in paying a $10 premium to people to go to costco, buy an external consumer HDD, and then ship it to them via prepaid label).

Blackblaze a shit.

>>52545974
N64 cartridges use battery-backed SRAM, not flash.

>>52545255
Fake and gay.
I swapped disks for a month because of reasons and left my ssd unplugged for all that time.
No data loss.

>le data retention meme
USB sticks are much shittier SSDs at their core. I have sticks that haven't been used in 4 years and still retain all their data perfectly (compressed archives so corruption would definitely be detected). And these are cheap chinese sticks using the shittiest possible 2x nm TLC cells.

>>52550108
>I can't read

File: 1442699216685.jpg (59KB, 500x493px) Image search: [Google]

59KB, 500x493px

>>52549662
>I have a 2tb ssd

>>52545401
Enterprise drives are more reliable.
They just go through more wear and tear and hotter situations hence the week.

>>52550269
Not him but what's wrong with a high capacity ssd? I know they're pretty expensive right now but at least he didn't buy an ipad pro.

>>52545255
>May 11, 2015
>over a week of not using the ssd at all
Even the article you linked said it wasn't a big deal

>>52549057
All electronics are waterproof as long as you don't have a current flowing through them while submerged.

>>52550464
Water doesn't even conduct

>>52550343
>Enterprise SSDs, however, have entirely different characteristics. An enterprise drive stored at 25C and operated at 40C has a retention rate of just 20 weeks. In worst-case scenarios or high storage temps, the data on an enterprise drive can start to fail within seven days. 3D NAND, which uses an older manufacturing process, might rate better in such metrics, but JEDEC doesn’t include that information.

>A recent presentation by Alvin Cox of JEDEC demonstrates just how large the gulf is between enterprise and client drives. Temperatures while active and in storage are both listed. If a drive is stored at 25C or operated 40C, expected data retention for a client drive is 105 weeks, or nearly two years. Let the storage temperature creep up to 30C, or 86F, and the drive should still hold data for an entire year.

>>52550474
Pure water, sure.
But most water is not pure and the impurities in it do conduct.

>>52546161
>Not SHAHAHA/0
Disgusting.

so according to the article, storing a consumer grade ssd at a temperature of 25c when powered off, data retention is good for two years straight when powered off. creep it up to 30c and its solid for a year.

while enterprise drives, when powered off, at 25c, data is retained for 20 days. push the temps higher, to 40c, and it will only last seven days.

i'm confused though, the article talks about "operated" like "stored at 25c or operated at 40c." does that mean that if the drive is operated at 40c it will only keep data for a year even though its always on?

This only applies as a minimum requirement for NAND that has reached the point where it is officially worn out, which for MLC, is typically 3000 cycles.

The three problems with this are:
a) Typical NAND performs better than this ... this is a minimum requirement and I bet everything other than Samsungs dodgy 19nm TLC performs better than this.
b) Almost nobody actually gets anywhere near wearing their NAND out. And fresh NAND is a lot more resistant to bit rot. At least an order of magnitude better.
c) The conditions required to induce rapid degradation are very convoluted. You need to operate the NAND at very low temperatures (below typical PC temps), to induce as much damage to the NAND as possible, then store the NAND unpowered at very high temperatures (hot enough to burn your skin).

I suggest people generally do not need to pay any attention to the media whores. While the relationship between storage and operating temperature, and unpowered retention is good to know, it needs to be taken in the proper perspective.

Why would you take your enterprise SSD containing important data to the Arctic and leave it off for a week straight?

>>52550673
this

you only face this issue once the ssd write endurance is worn and stored in high temperatures.

most consumer grade ssd's have a write endurance of 70tb. some are 140tb, like 500gb drives since they have two 150gb ssd's in time and the controller writes back and forth between the two to balance out the writes.

so it takes a lot of writes to hit the limits.

limit your writes and don't store the drive in an oven.

>>52550757
two 250gb inside and*

>>52545281
Still no problem. The PSU still provides power on +5VSB when the system is off.

(Unless you're one of those people who unplug literally every electrical device in their house when they leave for a while in case it somehow starts a fire)

>>52545255
I have a 512 MB flash drive from about 2007 that I've lost, found, and powered on for shits in giggles in a cycle of about maybe once a year, and it hasn't lost anything.

File: 1418613632287.png (88KB, 431x433px) Image search: [Google]

88KB, 431x433px

>>52550992
*shits and giggles

>inb4diamonddozen

File: tumblr_nvpu846kZD1s0hg4lo1_500_1_.gif (2MB, 500x375px) Image search: [Google]

2MB, 500x375px

>>52549947
>limit your writes and don't store the drive in an oven.
Oh my god, don't tell me how to live my life MOM!

>>52547454
>However as you write to them more they become less reliable.

What if you dump your OS onto an SSD and everything else on a different drive? Technically the OS is always active, and always being updated and changed. Would that potentially create problems?

>>52545326
>people live like pic
WHY

Reminder Seagate is spreading FUD to discredit SSD's and promote HDD's.

File: 1452988222835.jpg (110KB, 793x960px) Image search: [Google]

110KB, 793x960px

>>52545255

OP is a retarded faggot.

0/10, fuck off, sage and hide!

What a load of shit. My desktop was completely unplugged from the wall for a month and there was no data loss.

>>52550474
it can induce corrosion though.

>>52549434
Where would I find a good tape drive? Everything is in the thousands and I am not looking to drop more than 100 on one

>>52550757
>most consumer grade ssd's have a write endurance of 70tb.
No, that's not how it works. Most modern SSDs have 1-5K write cycles depending on how cheap they are.

Assuming you get an okay quality SSD then you can expect at least 3K write cycles. So:

64GB = ~192TB of writes

256GB = ~768TB of writes

1TB =~3,000TB of writes

Still, you'd have to write ~64GB of data per day for 8 years straight to get anywhere close to killing your 64GB SSD. Make that 1TB of fucking data per day for 8 years straight to get anywhere close to killing your 1TB SSD.

Average joe can barely write 2GB of data to their SSD browsing the web each day so even a tiny 64GB SSD will outlive them.

File: specs.png (155KB, 1124x720px) Image search: [Google]

155KB, 1124x720px

>>52551711
no it actually is. from manufactured specs. you can talk all you want from people running their own benchmarks to see, like techreport did, but i'm going off of manufacturers specs.

example, samsung states a 75tb write endurance for my 340gb 850 evo. crucial states 70tb for my m500 240gb, and crucial states 80tb for my mx200 250gb.

>>52545496
Didn't happen to me either, and I know an entire school that uses them and has no issues. Are you sure it's not you calling your personal experience the norm?

>>52551094
Not at all. On average a SSD will be subjected to an average of 1GB of writes per day by the OS at worst. SSDs have wear leveling tech so those writes are subject to the overall 3-5K write cycles SSDs have.

So just from the OS updating and shit alone, you'd waste like 6 write cycles on a 64GB SSD. Which is really really insignificant.

You are a faggot by yourself. But spreading false facts. Makes you a retard too.

File: readImage.jpg (38KB, 620x445px) Image search: [Google]

38KB, 620x445px

>>52545255
>falling for the SSD meme
>ever

>>52551898
That literally makes no sense. V-nand is good for at least 6K write cycles. A 120GB consuner grade V-nand SSD should be able to easily write at least ~700TB of data without fail. Enterprise V-nand SSDs are suppose to have 20-30K write cycles.

The TBW may be a meme from Samsung.

>>52545255
I haven't powered my old SSD up in about a year.
Oh boy.

>not knowing the difference between DRAM and flash

Is this my /g/?

>>52552140
its samsungs specs. its what they rate them for as a minimum, its why i go off of manufacturer specs and not pseudo specs people state. a manufacturer can reduces quality control and setting a low tbw allows them to do just that without getting flack.

>>52552154
The data is safe. Even moderately worn out SSDs have at least two years of data retention if left around room temperature.

File: specspro.png (59KB, 1080x512px) Image search: [Google]

59KB, 1080x512px

>>52552240
tbw specs for the 850 pro.

like seagate lists berracudas with a power on hour waiting of 100 days total, based off of 5 day a week usage of less than 8 hours a day, setting lower specs allows them to later on to cut cost or whatever reason, they can easily do so and not get flack.

flack like how kingston received from their bait and switch with their one line of ssd's. they just didn't bother to update the specs to reflect it.

>>52552384
rating*

>>52552384
That's very worrisome. Guess I won't be going to samsung for a SSD upgrade. Shame, v-nand was suppose to be the shit.

>>52552384
That's retarded as fuck. Samsung is saying that after 300 write cycles, their 1tb ssd can fail. That's fucking garbage.

>>52545255
>ITT OP has never experienced the glory or reliability of an ssd.
Based on your knowledge a USB flash drive is a meme to, they both store data by storing charges. Imagine if we all had to haul a hdd every where.

Op drops his laptop, hdd breaks and will cost hundreds to recover.

Also a friendly reminder to back your shit up.

>>52545255
>Reminder SSD will kill your data if you don't power it after 7 days of inactivity.

Reminder that SSDs powered a mars rover that has been working for over 12 years continuously, out of its planned mission time of 90 days.

File: corsair.png (60KB, 615x483px) Image search: [Google]

60KB, 615x483px

>>52552411
>>52552449
but its not just samsng.

pic related, corsair top of the line neutron xt 480gb and crucial's high performance mx200 line.

they all list lower write endurance's.

>>52552613
http://techreport.com/review/27062/the-ssd-endurance-experiment-only-two-remain-after-1-5pb

MLC NAND can usually handle 3K or more write cycles and V-NAND can usually handle 6K or more write cycles without fail most of the time.

It has been tested over and over again. TBW isn't even half the 3K write cycles MLC NAND can have.

Either manufactures are getting shitier quality control or TBW is just an arbitrary measurement manufacturers have made up for whatever reason.

>>52545255
OP, that shit has been debunked, and it was regarding Enterprise SSD's.

>>52552613
>"So, after over 7PB written the 850 Pro has finally given up the ghost. It's been an amazing run over the past 9 months, and the drive has vastly outlasted its puny 150TB warranty - by a factor of more than 46 times."

http://packet.company/blog/2015/11/25/the-end-has-come

The TBW metric is obviously flawed if a samsung 1tb 850 pro has surpassed it by 46 times.

I've personally torture tested a 1GB SD card and it failed after about 4,000 write cycles. SSDs use the same or better technology than in old ancient SD cards.

>>52550954
I do that with m power bars...

>>52545349
>ve fun sending back your SSD every week, i'm sure the shipping costs will be of no issue
Good warranties have free shipping to and from.

Also losing data does not mean your drive is suddenly dead.

The article is stating data loss in 1 week at unreasonably high storage temperatures. Last I checked, only a moron like you would store a drive in an oven.

>there are people on this board right now who have never owned an SSD and call it a meme
le earth is le flat guys right?

File: yellow.jpg (145KB, 615x500px) Image search: [Google]

145KB, 615x500px

>>52547518
>Microsoft or Apple cuck so mad at Linux users he'll find any opportunity to take a shot at them no matter how off-topic
Never change.

>>52545281
You don't know my life, faggot.

Each SSD has their own individual value for data retention after power loss.
Intel's enterprise branded as well as their enterprise-tier SSDs (the 730) are rated to last a month with its capacitors.

>>52545326

850pro reporting in.

>dataloss on a ssd lol
>OP is an idiot.

>>52553651

thank god i have the 850 pro.

poor intel fags.