Samsung Announces New SATA 3.0 6 Gb/s SSDs

Status
Not open for further replies.

ralfthedog

Distinguished
Oct 23, 2006
39
8
18,535
[citation][nom]Legendkiller[/nom]What's the different between 20NM SSD and 30NM SSD? I Dont know why NanoMeter matters so much because i dont see improvement for smaller NM...[/citation]

The difference is 10NM. :)

It lets them make more chips per wafer. From the consumer perspective, you might see a price drop if they can keep the yield up.
 

ikefu

Distinguished
Aug 11, 2009
251
0
18,780
Traditionally, the smaller the circuit goes the higher the speed they've been able to squeeze out of it as well. Unfortunately, smaller sizes have also been linked to a diminished number of write cycles before failure.

So higher speed but shorter life. However, for the average SSD user that life spans is still many years. Only the heaviest of SSD write cycle users will find the shortened life as a problem.
 

shadamus

Distinguished
Sep 16, 2010
120
0
18,710
Sweet. I've got two of their 470-series drives (128GB in an i5 laptop, and 256GB in my i7 laptop), and have not had a lick of trouble with either one. I would be willing to replace the 256GB 470 with one of these after I see a bit of a track-record...
 

rantoc

Distinguished
Dec 17, 2009
1,859
1
19,780
[citation][nom]ralfthedog[/nom]The difference is 10NM. It lets them make more chips per wafer. From the consumer perspective, you might see a price drop if they can keep the yield up.[/citation]

Sadly it also traditionaly have resulted in less erase cycles per cell
 

becherovka

Distinguished
Apr 8, 2009
163
0
18,680
Less erase cycles per cell might not effect you for 20 years depending on low/normal usage, by then who cares. But I guess it will affect some people.
 

razor512

Distinguished
Jun 16, 2007
2,130
68
19,890
What I find strange is people question the MTBF of the drives but they don't question the write cycles.

I have personally killed ssd's in a few hours using spinrite (x25 v)

Why doesn't tomshardware actually test how many GB of data can be written to the drive before killing it?

the MTBF is extremely inaccurate, what makes you think the advertised write cycles are any better?
 

kureme

Distinguished
Apr 21, 2010
282
0
18,810
[citation][nom]razor512[/nom]What I find strange is people question the MTBF of the drives but they don't question the write cycles.I have personally killed ssd's in a few hours using spinrite (x25 v)Why doesn't tomshardware actually test how many GB of data can be written to the drive before killing it?the MTBF is extremely inaccurate, what makes you think the advertised write cycles are any better?[/citation]
By the time Tom's finishes these tests, the results will probably be irrellevant.
 

Auriacan

Distinguished
May 12, 2008
3
0
18,520
[citation][nom]razor512[/nom]I have personally killed ssd's in a few hours using spinrite (x25 v)[/citation]

Why on earth would you use spinrite on an ssd? That's just asking for the drive to be killed. Spinrite is for magnetic spindle drives, not for ssd. The constant reading/writing of sectors that spinrite performs would just destroy the flash memory, killing any chance of keeping your write cycles safe. Even ask Steve Gibson (the creator of spinrite) and he will tell you running spinrite on an ssd is a bad idea.
 

subasteve5800

Distinguished
Sep 15, 2010
316
0
18,810
[citation][nom]Auriacan[/nom]Why on earth would you use spinrite on an ssd? That's just asking for the drive to be killed. Spinrite is for magnetic spindle drives, not for ssd. The constant reading/writing of sectors that spinrite performs would just destroy the flash memory, killing any chance of keeping your write cycles safe. Even ask Steve Gibson (the creator of spinrite) and he will tell you running spinrite on an ssd is a bad idea.[/citation]

I'm pretty sure his goal was to see how long it would take to kill the SSD.
 

alidan

Splendid
Aug 5, 2009
5,303
0
25,780
[citation][nom]razor512[/nom]What I find strange is people question the MTBF of the drives but they don't question the write cycles.I have personally killed ssd's in a few hours using spinrite (x25 v)Why doesn't tomshardware actually test how many GB of data can be written to the drive before killing it?the MTBF is extremely inaccurate, what makes you think the advertised write cycles are any better?[/citation]

lets assume 5000 writes on a at 500mb write speed.
1gb takes 2:42 to burn out.
80gb takes 224 minutes
128gb takes about 358 minutes
256gb takes about 716 minutes
512gb takes about 1433 minutes

now these are insane numbers assuming max write is going at all times, and that you are getting max data at all times.

in all likelyhood the numbers quoted would be about 2 to 3 times in a more real world insane write.

and no user would ever do something like this, and i mean consumer, not server or such.
 

dalmvern

Distinguished
Jun 15, 2011
673
0
19,060
I am glad that Samsung is finally coming out with the 6 Gb/s SSD, but it really annoys me when they claim that "its high-performance allows users to download up to five DVD video files in less than a minute." If they had said you can transfer 5 DVD files from one SSD to another in less than a minute, I would accept that, but to begin with, I doubt any of us have the internet connectivity and bandwidth to download 5 DVDs in less than a minute.

Less hype, more benchmark please.
 

alidan

Splendid
Aug 5, 2009
5,303
0
25,780
[citation][nom]dalmvern[/nom]I am glad that Samsung is finally coming out with the 6 Gb/s SSD, but it really annoys me when they claim that "its high-performance allows users to download up to five DVD video files in less than a minute." If they had said you can transfer 5 DVD files from one SSD to another in less than a minute, I would accept that, but to begin with, I doubt any of us have the internet connectivity and bandwidth to download 5 DVDs in less than a minute.Less hype, more benchmark please.[/citation]

if you are connected to a backbone line, you may have FAR higher bandwidth than the ssd can provide.
 
Status
Not open for further replies.