|
Lovely day, turn my computer on today, and SSD decides to die out on me. There goes everything. Thank god I had stuff backed up on my seconday drive so I can just plug it into my smaller OS drive.
So I send my RMA in and decide since it crashed I would update my comp.
Go onto newegg and buy some stuff. Pretty enthusiastic 
I'm getting a 890FX GD65 AM3 Mobo
G.Skill Sniper 8GB 240 DD3 1600 ram
and Bought a new Processor Which was the most usless update i consider doing
Since i went from
940 Old
to 965 New
Ohh well maybe Ill sell my old one for something cheap like 80 bucks or something, cuz its pretty kick ass.
Was watching day9 stream yesterday was really funny, anyone know which episods he talks about the porotss side with timings and improvements.
Thnx for tuning in.
PS: no new art, COMP crashed but you already now that if you read the text up top
|
How long have you had your SSD and how often did you write 0_0
|
On April 09 2011 07:05 paper wrote:
How long have you had your SSD and how often did you write 0_0
I bought it on 01/19/2011, And took like 3 days to get here.
Installed it on 01/22/2011 worked perfectly until 4/8/2011
"when you say How often did you write 0_0 "
I'm a little confused by what you mean.
|
dead SSD != SSD ran out of writes
A SSD without writes would still be readable. A dead SSD is not readable because there is something physically wrong with the controller, memory chips, or something else (which by the way is not affected by the amount of writes).
You should have waited for AMD's next generation processors scheduled to be released in June or got an Intel platform instead.
|
SSDs are built on Flash memory which unlike a hard drive can only be re-written a limited number of times. (This is not technically true as a hard drive will eventually give out as well, but the number is so high that it would take many many more years of constant re-writing to achieve that same kind of failure that an SSD can see).
If you use your computer in such a way as to be constantly writing to the hard disk you will wear out and eventually kill your SSD. That said it should still take quite a long time, so I would be interested in finding out how long it took.
|
Terrible CPU uppgrade.
125$ for 400 mhz ? LOL you should have waited for Bulldozer if you like AMD... or buy a 2500K.
|
On April 09 2011 07:30 Rayeth wrote:
SSDs are built on Flash memory which unlike a hard drive can only be re-written a limited number of times. (This is not technically true as a hard drive will eventually give out as well, but the number is so high that it would take many many more years of constant re-writing to achieve that same kind of failure that an SSD can see).
If you use your computer in such a way as to be constantly writing to the hard disk you will wear out and eventually kill your SSD. That said it should still take quite a long time, so I would be interested in finding out how long it took.
Worst case household scenario: 25 years-ish (90% space filled, 10GB/day written)
Regular household case scenario: 100 years-ish
And that's with the 28nm (or 2x, can't remember).
So no, it's not the write cells being unusable.
|
On April 09 2011 07:33 Boblion wrote:
Terrible CPU uppgrade.
125$ for 400 mhz ? LOL you should have waited for Bulldozer if you like AMD... or buy a 2500K.
For real, get yourself a hot new peripheral or something, don't squander your hard earned money
Myself I just recently put the following in to my PC to give you an idea:
Thermaltake FRIO Cooler
HyperX Memory 6GB @ 1600
OCZ 120GB SSD
Deathadder Mouse.
Some new Fans/PCI Mounted Cooler for GFX card.
That was my last big buy for the PC.
|
On April 09 2011 07:35 MichaelEU wrote:
Worst case scenario: 25 years-ish (90% space filled, 10GB/day written)
Regular household case scenario: 100 years-ish
And that's with the 28nm (or 2x, can't remember).
So no, it's not the write cells being unusable.
I come from a server environment where you can end up with a lot more than 10GB/day but usually don't have that fill percentage. Either way is silly in a home environment as it is rare that anyone will average that much data a day.
Either way I knew it was quite a long time, much longer than anyone will ever keep a computer these days. Thanks for the info. =)
|
On April 09 2011 07:49 Rayeth wrote:Show nested quote +On April 09 2011 07:35 MichaelEU wrote:
Worst case scenario: 25 years-ish (90% space filled, 10GB/day written)
Regular household case scenario: 100 years-ish
And that's with the 28nm (or 2x, can't remember).
So no, it's not the write cells being unusable. I come from a server environment where you can end up with a lot more than 10GB/day but usually don't have that fill percentage. Either way is silly in a home environment as it is rare that anyone will average that much data a day. Either way I knew it was quite a long time, much longer than anyone will ever keep a computer these days. Thanks for the info. =)
You're right, it should've been "worst case household scenario", edited it out, thanks  . For which 10GB/day is a pretty large amount. I don't come from the server environment but do they use SSDs there? I doubt it but I don't know a thing.
|
I thought the new generation of SSDs didn't have that issue?
|
Most 25 nm flash is rated (many say conservatively) for 3000 write/erase cycles. 34 nm flash is often rated for 5000 write/erase cycles. That's what's used in SSDs currently, with 34 nm flash being phased out.
Think about the amount of capacity an SSD has. If you've got a 100GB SSD, are you going to write 100,000GB of data to it ever? That's a gross oversimplification of the issue, but considering the redundant spare flash on each SSD, write amplification, and spreading of writes through the drive by the controller, the amount of data it would take to run out of writes is probably still going to be around that order of magnitude or maybe a little less. It's just not happening under home computing usage.
That said, as mentioned before, SSDs dying don't have to do with running out of writes. It's not the endurance of the flash that is causing the failures.
On April 09 2011 08:23 MichaelEU wrote:Show nested quote +On April 09 2011 07:49 Rayeth wrote:On April 09 2011 07:35 MichaelEU wrote:
Worst case scenario: 25 years-ish (90% space filled, 10GB/day written)
Regular household case scenario: 100 years-ish
And that's with the 28nm (or 2x, can't remember).
So no, it's not the write cells being unusable. I come from a server environment where you can end up with a lot more than 10GB/day but usually don't have that fill percentage. Either way is silly in a home environment as it is rare that anyone will average that much data a day. Either way I knew it was quite a long time, much longer than anyone will ever keep a computer these days. Thanks for the info. =) You're right, it should've been "worst case household scenario", edited it out, thanks . For which 10GB/day is a pretty large amount. I don't come from the server environment but do they use SSDs there? I doubt it but I don't know a thing.
They use lots and lots of SSDs in many server and other enterprise environments. I/O demand is typically a lot higher for those kinds of applications than in home computing. Many of those systems must handle a lot of concurrent read and write requests, which owns even those huge arrays of 15000rpm hard drives. Flash-based storage is even more important for that kind of workload.
I use my school's computer cluster (mostly a couple thousand Core 2 Quad-equivalent Intel Xeons linked together) a lot for running simulations and other odd jobs for research, and when they moved the home directories to an SSD array...huge difference in accessing some of my files and results.
|
|
|
|
|
|
EPT
