>http://techreport.com/news/29583/mushkin-previews-a-500-4tb-ssd-at-ces
>Mushkin previews a $500 4TB SSD at CES
Can't wait to build a NAS with all solid state drives. Who gives a fuck about vaporware helium tech when in just a few more years SSD will be even cheaper/reliable.
>How do you increase drive capacity further when your controller tops out at 2TB? By using JBOD. To construct a 4TB drive, Mushkin is basically putting two Reactor 2TB SSDs on a single PCB. Those drives are then joined into a JBOD with a separate controller.
What could possibly go wrong?
>>53104760
Was going to post just this. I'll wait to see the price of the 2TB drive and get that if it's ~$300.
>>53104760
Is this a fancy way of saying raid 0?
>>53104838
more like 2 drives partitioned into 1.
>>53104838
No, not at all. It stands for Just a Bunch Of Discs. It shares some of the fun features of RAID 0 though, in that if any one of the discs go the whole thing is fucked. Don't think there are any performance benefits to it.
Does this mean 120GB SSDs are going to be available for less than 20$ soon?
>>53104982
They will just take them off the market like they do with mechanical drives once they go below $50~
>>53105060
SSDs don't vanish from the market like that. Smaller manufacturers take up the flash the big ones don't want so you end up with cheap, small capacity SSDs whereas HDDs are all made by a few big players with limited factory space.
Samsung's gonna be pissed.
http://www.newegg.com/Product/Product.aspx?Item=9SIA6ZP3JU3350
>>53104673
>$500/4TB
fuck that noise, I can get a 4TB HDD for barely over a hundred bucks. My backups and chinese cartoons take up a lot of space, but do not need a bazillion IOPS.
When SSDs equal and/or beat HDDs on capacity per dollar, then and only then will HDDs die.
>>53104673
>Who gives a fuck about vaporware helium tech
Doesn't something have to not exist for it to be vaporware?
>>53105780
Yeah, but mechanicals are loud, take up space, and the large ones still use a fair bit of wattage 5-10w, so there are advantages to SSD storage. If I had the dosh, I'd spring for them.
>>53105185
HDD are pretty much dead in like 2 years
HDD prices have been stagnant like the last 3 years.. Once a 2-3TB hits $100 which wont be long from now, HDD will be done
>500MB/s SSDs in a NAS
>1Gbit ethernet
Fuck this shit
>>53105906
3TBs can be purchased for 85usd, so are you talking about?
http://www.amazon.com/dp/B005T3GRLY/?tag=pcpapi-20
>>53105964
I bought that exact drive on Black Friday... in 2012
>>53105723
>$2000
>>53105964
>seagate 3tb
enjoy failures
It's possible to recover data from a dead HDD but what a about a dead SSD?
>>53106013
Other brands have a similar price and SSDs are far from reliable.
>>53106020
The equipment required is way more expensive
>>53105922
>old, deprecated USB 3.0 is 5 times faster than gigabit ethernet
ayy
>>53106020
Yeah, just the same.
It's normally controllers that die, it can just be bypassed. Chips don't tend to just take a shit, and even if they do. Still recoverable.
>>53105780
It won't take long now. Even Seagate and W8 see the writing on the wall, with WINDOWS acquiring SanDisk and Seagate making their own SSDs.
I doubt traditional HDDs will remain competitive at the turn of the decade.
>>53106035
it's a HDD you pillock
and WD/Hitachi is better. Hell, toshiba is better
>>53105922
>not running multiple GbE links in parallel
>not getting 10GbE
>>53106085
Where did I claim otherwise? You are just making yourself look bad.
>>53106064
I got the best of, right now.
M.2 NVMe
Pioneering the future
>>53106064
I hope you're right, but it's undeniable that they're not yet there. And it's possible that they won't get there, because there's still a pretty big $/GB gap to close, and fabs are expensive.
>>53106064
>WINDOWS acquiring SanDisk
wut?
>>53106110
>SSDs are far from reliable.
>SSDs are far from reliable.
>SSDs are far from reliable.
>SSDs are far from reliable.
>I never said it was a SSD
>>53106135
You do realise that most SSDs have shitty firmware on their controllers?
>>53106124
Apparently he thinks Western Digital is Windows
>>53106135
>>53106167
SSDs are more reliable than HDDs. There's a reason why their warranty and MTTF are much longer even on consumer devices.
>>53106167
Mate I've fucking lost track of what you're on about. Are 2 people pretending to be 1 or something
>>53106167
Sure if you buy from a garbage company. My samsung 840 (non pro) 120gb has close to 8tb written to it by now and zero problems. Still in use as a dedicated OS boot drive on my everyday machine.
>>53106198
Ah, OK. Hmm, that seems bad. SD have always been more generous imo than WD, so we'll probably see their prices rise. Also, worrying we're seeing HDD market players in SSDs, I'd like to see oligopoly be staved off for as long as possible.
>>53106135
Maybe you should try to read what I wrote.
>Other brands have a similar price
http://www.amazon.com/Hitachi-7200RPM-Desktop-SATA-Drive/dp/B00SG4XD24/ref=sr_1_2?s=electronics&ie=UTF8&qid=1456092179&sr=1-2&keywords=hitachi+3tb
Same capacity and almost the same price.
> and SSDs are far from reliable.
Did you forget the OCZ SSDs, SandForce firmwarebugs, guaranteed raid failure and corruption on power loss (consumer SSDs)?
>>53104673
Will they still retain their data for a long period of time when powered off?
>>53106249
>8tb written
That's like less than one year of usage.
>>53106124
Sorry, was on my Android phone with a seriously aggressive autocorrect. I miss my windows phone, the keyboard alone is enough to make me want to ditch the S6 for a Lumia 950.
>>53106121
It gets cheaper over time and there are still many different ways to attack the problem of density. The 3D v-nand that let Samsung make such a huge leap in $/GB? That was on 40nm.
>>53106306
Not him but do you want to look at issues with individual lines? We could do that all day with HDDs. It would be far better to look at returns, at which point both HDDs and SSDs have such low return rates that there really isn't any point in even considering it.
>>53106053
In reality tests have shown it's not much faster than 1GbE. Also, enjoy your single device access and not direct remote access
>>53106105
LAG still only does a max of 1GbE per device. MPIO can do a little bit better but not applicable outside of virtualization
>>53106320
I had a msata 16gb that was in a plastic bin for 5 years and the data was still intact. It booted arch fine when I plugged it in this year.
>>53106119
>Pioneering the future
we ads now.
>>53104838
they're just concatenated
raid0 is like;
ABABABABABABABABAB
where "A" is blocks of the first drive, and "B" is blocks of the second
JBOD is;
AAAAAAAAABBBBBBBBB
just one after the other
>>53106035
>SSDs are far from reliable.
it's not 2010 any more
http://packet.company/blog/?category=850+Pro+Endurance
>>53104673
>when in just a few more years SSD will be even cheaper/reliable.
They have so far to go.
I bought a 256Gig EVO for $85.
I could buy a 1Tb spinning rust for less than that.
SSDs are not going to get cheap soon.
More likely the smaller sizes will vanish and larger sizes will be all you can get. The price points will move upwards, of course.
>>53106320
Yeah. I'm no expert, but all those SSD horror stories are usually from drives deployed in corporate datacenters (ie fuckton of constant writing 24/7)
>>53106119
Enjoy your proprietary firmware, faggot.
>>53104760
All I need is 2TB so I am just looking forward to the $250 2TB drive
>>53108710
>proprietary firmware
As opposed to?
The future is bright. I'm going to buy a fuckload of 4TB NAS drives this year, but I'm really hoping my next server upgrade is full on flash.
http://www.kitguru.net/components/ssd-drives/anton-shilov/toshiba-128tb-ssds-will-hit-the-market-in-2018/
>Just got a 850 evo for 95$ with tax
>Prices will go down even more
I can never win.
>>53108939
Don't worry anon, I bought a 1TB EVO when it was $400. I've seen them go way down in the upper $200 range.
>>53108878
SD and others were promising 8TB SSDs last year, we just recently hit almost 4TB drives (3.8), so color me skeptical.
>>53108989
Samsung will push 4TB drives this year. It's just very optimistic but stupid extrapolation. They'll pass HDDs in capacities very soon though with Meme 3D NAND and shrinkage. There's also the lookout for xPoint.
you're all fucking idiots for getting SSDs for long term storage purposes
the scary part is, there are actually idiots putting out more than $100 for storage beyond 2TB
>>53109094
Cry more.
I got about 1.5TB combined from SSDs and about 16TB combined of HDDs.
>>53109048
Kinda doubt it, their current 3.8TB drives are over 2K and enterprise only. That's not going to fly in consumer space and they'd be hurting their enterprise profits with a viable 4TB offering, and they'd still have to contend with being more than HDDs in a more space centric part of the market.
Is xPoint Intel's new cells? Because as I recall they plan on pricing them between ram and flash, so it's pretty much nonviable outside enterprise. Or are you do you think it'll put downward pressure on ssd pricing?
>>53109231
good goy,buy more SSDs
>>53109241
>they'd be hurting their enterprise profits with a viable 4TB offering
Not him but no, that's not how it works. I could go on and on about the benefits of the enterprise hardware and the shortcomings of the consumer version but instead I'll ask one question.
There are both consumer and enterprise versions of 4TB HDDs. Why do enterprises buy the enterprise version since it's more expensive?
>>53109348
I can understand HDDs having a higher cost enterprise product, since they're mechanical and have a lot more failure points, and so there's a lot more to be concerned with in terms of quality control and such. With SSDs though, I assume TLC is TLC and Vnand is Vnand. Are the controllers a lot more advanced? Larger caches? I assume it's not just warranty length, since all the warranties in the world won't matter if your data's gone (backups aside).
>>53104673
>$500 4TB SSD
>>53109854
Enterprise class SSDs are almost always SLC (Single Level Cell) flash, which is much more expensive per byte to make, but makes it up in the fact it can withstand an absolutely enormous amount of writes, as the controller only has to pick up 2 voltage levels, high and low.
MLC and TLC flash sacrifices write endurance for capacity, as each one crams more bits per cell, but the controller has to differentiate between multiple voltage levels. As the SSD wears out, it becomes harder to find the difference between the voltage levels in order to read the data back.
>>53109094
Indeed. If there's one thing an SSD can never do better than a mechanical drive, it's store data for long periods of time without failing. All of those non-moving parts fail after a prolonged time of doing nothing.
>>53108710
>he thinks hard drives don't have proprietary firmware
>>53109926
I see, I guess I was thrown off because Samsung's 3.8TB drive is vnand, so I assumed they used the same cells across their product ranges, since you see vnand in both enterprise and consumer drives.
http://www.newegg.com/Product/Product.aspx?Item=9SIA6ZP3JU3350
>>53109948
Define 'long term'?
I seriously have not had this issue at all... I mean, realistically what are you doing that causes this?
I have flash drives from 10 years ago and they work. Just fine.
>$500 4TB SSD
>not even 2017
Mushkin has some fucking huge balls to do something like this, holy shit.
>>53110028
Vnand can be SLC, the controller just has to be set to only write one bit per cell, instead of 2 or 3.
Its definitely not TLC though, the write endurance for it is terrible. MLC is most likely for that particular model of SSD, as I am not aware of SLC flash being dense enough to pull that much storage off quite yet.
>>53110044
I was being sarcastic. For long term storage, an SSD will always outlast a mechanical drive. Always
A fucking sleeve bearing fan will outlive most any consumer HDD that's made now.
>>53108397
>corporate datacenter
>being powered off
it's about the flash cells losing their data when you use the drive as cold storage
Hard disks will eventually become like tape drives. Used for archiving only.
As for SSDs, laptop manufacturers are slow as fuck for adaptation. They'll probably drag on HDDs for years because they save $5 on the price.
>>53104916
>>>53104838
>No, not at all. It stands for Just a Bunch Of Discs. It shares some of the fun features of RAID 0 though, in that if any one of the discs go the whole thing is fucked.
If it's JBOD and a drive dies, you can still recover whatever is on the good drive. As long as the controller isn't retarded.
>>53105804
Helium vapor?
>>53109899
>mfw I bought a 1TB drive for about $500 a few years ago
Sure, Mushkin's performance is guaranteed to be shit, but it's at least something.
>>53105906
>HDD prices have been stagnant like the last 3 years.. Once a 2-3TB hits $100 which wont be long from now, HDD will be done
Why would I spend 100 on a 4TB SSd when a hdd of same size is $10?
>>53110757
HDDs are terrible for long term storage. There are so many small and delicate moving components that can seize up after a few short years.
>>53106105
>>not running multiple GbE links in parallel
>>not getting 10GbE
Link aggregation and NIC teaming is only for fault tolerance and load balancing. You will still never get more than 1gb from a single remote host.
>>53110993
That's because they don't tell you to grease the bearings every few years like you are supposed to they only fail due to friction. Everything else can be fixed by Wapping controller board.
>>53111214
I could have sworn I found a version that increases bandwidth
>>53110832
>performance
I've bought fancy Samsung Pros, and I've bought ultra budget tier Kingstons and Crucials.
As far as I'm concerned, they're all too fast for me to notice.
The only time I know there's a difference is when I run benchmarks on them, because the numbers get shoved in my face.
>>53110993
you are fucking retarded? ssds literally corrupt themselves after being left unpowered
mecahnical drives have been built for fucking decades. they are excellent for long term storage
>ssd for storage
>>53111360
you are correct. however, kingston is literally the worst manufacturer to purchase from as far as speed goes and reliability.
crucial isnt that bad actually.
>muh hdd
>muh ssd
I don't care, I just want something that can store stuff and won't shit itself after 10 years.
I don't even care if the read/write speed is shit, because they're just videos, audio and images; I can just copy the ones I want into active usage.
>>53111407
lol
>>53111404
You should learn to read before you call others retarded, though I must also admit fault for using the word storage and not archival. So, let me drill this one home, HDDs are not suitable for data archival.
>ssds literally corrupt themselves after being left unpowered
That has absolutely no bearing on the matter at hand.
>>53108878
>QLC
>>53111360
That's about right. They're all rocking responsiveness, the rest is more for bragging rights. I wonder how the NMVe SSDs fare however. It won't be a big jump like from HDD to SSD.
>>53111429
Crucial used to up there with Intel and Samsung, they were sadly left behind. I hope they can get some Micron goodies to get their feet back up.
>>53111486
>implying there's anything wrong with shittier MLC nowadays
Even Samshit is dishing out TLC enterprise drives. It's not 2008 anymore.
Speaking of 2008,
>yfw a 32GB drive was $500
Isn't the bit density limit approaching the maximum on spinning disks? The fact that WD had to go helium to push to 8TB shows it.
Meanwhile Toshiba is showing off 128TB and we're still on 2.5" form factor, imagine what they can do if they made 3.5" SSDs.
Regarding Mushkin, I don't trust JBOD. It's nice to see a cheap 4TB drive but I will wait for a controller that can actually support 4TB without silly tricks.
>>53111478
>That has absolutely no bearing on the matter at hand.
So you're telling me that if I want to archive something, I need to keep it plugged into something forever?
What if I want to lock it in a safe?
>>53111557
It's kinda crazy when you think about it. In a decade or two we could be carrying around the entirety of today's internet around with us on our wrists.
>>53111584
I... what? Who said anything at all about using an SSD for archiving anything? Please go back and actually read. If you want to archive data use tape. If you have nothing of value to say, don't say anything.
>>53111478
hard drives are fine, tape is better
>>53111633
>hard drives are fine
See >>53110993
>There are so many small and delicate moving components that can seize up after a few short years.
Hard drives are not ideal for data archival, tape is what you'd want for that. But again, I'm repeating myself because apparently people can't actually follow a conversation yet still feel the need to say something.
>>53111699
what the fuck is wrong with you? i have drives that 10 fucking years old which still spin up.
stop acting like a fucking retard, there are like 2 'delicate' parts of a drive and they dont fucking 'seize up' you dip
>>53111488
I moved from a SATA 3 to M.2 interface (same motherboard Asus X99-A, I just haven't bought an M.2 SSD yet and recycled an existing Crucial MX200 to boot)
Boot times are pretty much the same. So are program launch times and compile times.
(Except Visual studio 2015, that fat piece of shit launched a bit faster because it's so bloated)
I honestly think I should have saved my money and bought a larger SATA3 SSD instead of a faster M.2 one.
>>53111729
Yeah, I mean the reason professionals like to store all their archived data on tape is just because of nostalgia, your anecdotal evidence just proved that. The whole world should stop being so retarded and swap over to HDDs for long term storage of their data, the anon without a shift key has proven their reliability.
>>53111820
no, most professionals use tape because it has provided huge capacity long before things like hard drives and blu rays did, and once the infrastructure is in place and it works there is no need to change.
if you are backing up tens and hundreds of TB it makes sense due to cost, but if i was to buy something like 1 or 2 6TB reds and put them into a safety deposit box at a bank they have just the same odds of working when i plug them in.
you are not going to fool me about acting tough.
>>53110993
They can, but that doesn't mean they do. I have drives turning 31 this year.
>>53111729
>stop acting like a fucking retard, there are like 2 'delicate' parts of a drive and they dont fucking 'seize up' you dip
There was a time when the lubricants used on the disk platters would thicken with age.
Centrifugal force would tend to move this goop out to the rim of the disk.
That's where the heads are parked. On cold days the goop would be tacky enough to lock the heads to the disk through surface tension.
Disk drive motors are not high-torque. So at startup the disks would not rotate.
At IBM's labs in Germany they tried to address this with a microcode patch to the controller to increase the starting torque.
Result: heads ripped off.
http://gizmodo.com/optical-data-storage-squeezes-360tb-on-to-a-quartz-disc-1759359652
This will be used for archival in the future. People will use large SSD to their everyday storage needs and those discs will be used in the same way CD/DVD were a few years ago.
The problem I see is that normies won't archive anything because their SSD are going to be so large it won't matter. Because it won't, until their drive fails...
>>53112427
Stiction hasn't been a problem on hard disks in over twenty years, and hell I've only ever seen one drive made before that that has had been afflicted by it.
>>53110798
Can't you read data from failed SSDs, just not write it? I mean, unless the controller is retarded. I thought that was one of their benefits, graceful failure.
>>53111875
>but if i was to buy something like 1 or 2 6TB reds and put them into a safety deposit box at a bank they have just the same odds of working when i plug them in.
Bwahahahaha no. Just no.
>>53112118
Once again, you have proven to the whole world that hard drives are the way to go with your irrefutable anecdotal evidence! We should get this out to the rest of the tech community, so that they may learn from your wisdom.
>>53112596
I can back it up with a garage full of systems and functional 20-30 year old disks, can you care to show me the wealth of sources or other evidence he used for his one-liner that amounted to "they fail because I say so!", or your "hahaha XDDDD ur wrong" shitpost?
>>53113030
I really don't need to come with any evidence, at all. It's what the industry uses. It's either stored tape or an active datacentre, it's not a stored hard drive. You saying you have a garage full of functional systems makes no difference.
I have two garages full of five year old hard drives that all crapped out on me! See how fun anecdotal evidence is? This is why you don't trust testimonials on sleezy websites.
>>53112488
>gizmodo
>with another BS clickbait title
typical. The tech isn't commercial yet, they haven't stored 360tb of anything. They could theoretically store that much data. They actually stored a few megabytes.
>>53104760
Jail Bait OverDrive
>>53105865
>Fair bit of wattage
>10w
pick one faggot
>$500
>nope
Right now it's about 270USD+ for a 1TB mSATA, and it's just under 100 Mighty Joe Dollah for a 2TB 2.5"
Until one drops down closer to the other, guess which one I'm going with?
>we get it, they're objectively better, but fuck the current price/capacity.
>>53104673
>$500
>4TB
Why? When I can buy 8TB for ~223$?
https://pcpartpicker.com/part/seagate-internal-hard-drive-st8000as0002
Jesus fuck, now we're talking.
We where $0.25c per SSD GB, $0.05c per HDD GB, now it's what, $0.12c per SSD GB?
Shit this is getting intense. Switching to SSD's cost effective now if they have twice the lifespan/warranty of a HDD.
>>53115293
HDDs are around 0.028usd/GB
>>53115329
I dunno, SMR makes me nervous, it seems like such a massive hit on writes.
>Archive
Yeah, yeah, I still think it'd better to avoid it.
With markup & increased tax in the UK you're talking at least £400. A 4TB bog standard Seagate HDD can be had for about £100. I'm guessing the price ratio will be vaguely similar in most markets. I'm still not even close to considering SSDs for bulk media storage (not considering Seagate HDDs either but hey ho). When SSD cost per GB dips below 1.5x that of HDD then "RIP HDD" might be on the cards. At or very close to parity it's game over as OEMs & normies will have no reason to use HDDs. It will probably happen, but not in the 2-3 years some claim. My guess is a gradual shift, dragged out somewhat by HDD makers slashing prices towards the end, in something like 7-10 years time depending on tech. I would not buy the drive in OP's post.
>>53109094
>you're all fucking idiots for getting SSDs for long term storage purposes
Just backup your data anon :^)
>>53115485
Kinda depends on the margins for HDDs, if they're still high enough, SSDs might be chasing them for a long time.
>>53115383
If you scared of Seagate problem(I have two Seagate one 8, second 12 year old and still haven't single issue.) you can buy 5TB Toshiba price exactly same.
https://pcpartpicker.com/parts/internal-hard-drive/#sort=a7&page=1
Whether SSDs will overtake HDDs isn't a matter of "if", it's a matter of "when". SSD is objectively superior in every single measurable quality, from durability to performance to form factor to power consumption.
The only thing keeping HDDs around is their price/storage ratio, and while HDDs are growing at very slow rates, SSDs are growing consistently. Within the next 5 to 7 years, SSDs will be the same price or cheaper than HDDs at the same storage size. At that point, why would anyone use an HDD over an SSD, ever?
>>53115653
It's the SMR tech downsides itself, not seagate.
>>53115683
Indeed can't wait to switch over in 7 years... Saying it will be this or that, doesn't really help the now, and 5-7 years is a lifetime in tech. SSDs are going to be eternally stuck as OS drives, outside of some serious work in enterprise. Who knows, in 7 years there might be something new that makes SSDs a joke anyway. Memristors or something.
>>53115810
>It's the SMR tech downsides itself
No.
>>53115846
No, what?
>>53113192
>I really don't need to come with any evidence, at all. It's what the industry uses.
Really, you bitch and moan about "anecdotal" evidence in an effort to hide your lack of knowledge on what you're arguing about and you come back with this fallacious kludge? What the fuck is "the industry", and what impact exactly does this conveniently vague entity have on the reliability of hard disks in long-term storage?
>>53115810
SSDs have massive benefits in mobile computing. The fact that they aren't susceptible to shock damage and other moving parts issues saves companies and users money by significantly cutting down support costs. Plus most users wouldn't even know what to do with 256GB of space.
I don't think you can even buy a laptop over $500 with an HDD in it. Every has moved to SSD already.
>>53115881
Ah, forgot about laptops, good point, but even so many of those cheaper laptops are sporting 64/120/240 GB drives, which I think around these parts would be considered OS drives. If you want storage you buy a USB drive to plug in. It's true that most probably wouldn't even know what to do with the space, but I don't think it applies to the discussion really. I mean, people who have no use for storage obviously won't care about prices or other metrics, so it's sort of a dead to argue over, I think.
>>53115683
>5 to 7 years
That's optimistic. And 7 years is an eternity for tech, it's hard to predict anything that far down the line. SSDs have some inherent issues with continued shrinking (endurance) that HDDs do not, so it's possible that SSD pricing will stagnate. In recent years, the cost per transistor has begun to flatten after decades of steady decreases (and may be beginning to increase), as more exotic manufacturing is needed at smaller process nodes.
>>53115942
>I mean, people who have no use for storage obviously won't care about prices or other metrics, so it's sort of a dead to argue over, I think.
That was the point of my original post. When you don't have a need for extremely high storage capacity, SSD is already a viable and well adopted solution. We've been transitioning towards solid state for the last 15 years. Pretty much any device that isn't immobile has moved to some form of solid state storage.
>>53115869
You know what.
https://www.youtube.com/watch?v=a7Qt_SMAOcY
Mechanical storage most relevant and suitable for long-term archive if HDD too power-hungry for you, you can switch to LTO.
>2016
>using moving parts
Wew
>>53116159
Well there needs to be some moving parts in my computer to push the electricity right?
So let me get this straight. Instead of creating a better controller, just just grabbed a bottle of elmer's glue and some duct tape and stuck some drives together and called it good? BRILLIANT! I'll just wait for the 2TBs to come down in price.
>>53109048
xPoint is just Intel vaporwave as usual.
>>53110757
tapes won't be replaced any time soon, buddy
16tb SSD NAS reporting in
>>53116701
Nigga what? Post that shit.
1tb SSDs are still over $200.
>>53105723
>but mooooom you don't understand i NEED a 2000 dollar 3.84 TB SSD for muh gayman rig!!!
>>53110993
The point of archival storage is that you're not using it 24/7 once it's full. If you shut down the storage device and perhaps rewrite everything once a year then they can last several decades. Honestly SSDs, HDDs and tapes are all equally suited but the price matters more for archival purposes.
>>53116368
m.2 SSDs are starting to enter this niche. 2500MB/s is like one third of the sequential read performance of a single DDR3 stick and they are pretty affordable for what they offer.
>>53106035
SSDs are significantly more reliable. Unlike hard drives a good SSD will fail gracefully (still be able to read the data after it locks writes).
>>53106345
http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead/2
Out of petabytes.
>>53108989
>>53109048
>>53109241
http://www.pcper.com/news/Storage/Now-thats-big-SSD-Novachips-comes-4TB-and-8TB
8TB available now. Just expensive.
>>53111488
>>53111731
M.2 SSDs won't provide any noticeable change in a desktop (over a SATA ssd). Servers however are a different story.
>>53112521
That's correct
>>53110757
HDDs don't last long enough to replace tapes.
>>53110985
A 4TB HDD will never be $10, unless you're getting it used. There's a minimum price of the components required to make a hard drive.
>>53117786
>Honestly SSDs, HDDs and tapes are all equally suited
Go die in a fire. HDDs have moving parts that require lubrication. This lubrication breaks down over time. The only viable archival storage is tape.
>>53104673
HDD shills on suicide watch!
>>53104982
1TB usb flash drives when?
I want HDD to die.
No matter what I do after 2-3 years mines always start to sound like a chainsaw when I boot my computer.
>>53111404
Source for that one? Or purely anecdotal from /g/?
>>53119514
https://www.youtube.com/watch?v=zO4JIg-68gg
>>53110201
This actually isn't true for sub-20nm TLC flash. The charges in the cells dissipates after an extended period of being powered down (weeks or months).
This is why Samsung's 840 line of drives had performance issues after a period of usage. Their fix was to have the drive periodically rewrite existing data so that the charges are refreshed.
The 3D NAND technologies don't seem to suffer from this issue.
>TLC NAND makers delved too deep
>Balrogs in the quantum foam
>>53114360
>4 terabytes of jailbait
Aaw, shit nigga. Lifegoals.
>>53104673
>$500 4TB SSD
THE ABSOLUTE MADMEN
>>53111532
>>yfw a 32GB drive was $500
>mfwmf is 2008
>>53122440
Man i remember when i got my first SSD and put windows on it (was 64gb). It was like using a pc for the first time.
>>53122440
>OCZ
the cancer is dead
>>53122810
Nigga OCZ is one of the reasons SSDs fell in price. They might be bad, yeah, but they're based on another level.
