Sandisk MLC SSD's

Toytown

Gawd
Joined
Jan 13, 2005
Messages
996
http://sandisk.com/Corporate/PressRoom/PressReleases/PressRelease.aspx?ID=4478

The G3 SSDs are more than five times faster than the fastest 7,200 RPM HDDs and more than twice as fast as SSDs shipping in 2008, clocking in at 40,000 vRPM1 and anticipated sequential performance of 200MB/s read and 140MB/s write

Available in capacities of 60, 120 and 240GB*, the unit MSRPs are $149, $249 and $499, respectively

Not sure if that's true or not as i thought the Intel SSD was a lot faster.

The G3 SSDs provide a Longterm Data Endurance (LDE) of 160 terabytes written (TBW) for the 240GB version, sufficient for over 100 years of typical user usage. (2,3)

LDE (Long-term Data Endurance) - an industry metric, introduced by SanDisk, that quantifies how much data can be written to an SSD in its lifespan expressed in terabytes written (TBW). Data is written using typical PC transfer size pareto, written at a constant rate over the life of the SSD and data is retained for at least 1 year upon LDE exhaustion. Based on SanDisk internal measurements, as typical client PC user writes 4GB/day.

I think i prefer this measurement of life compared to MTBF, however it only gives the details for the 240GB version, am i to presume that because the 60GB version has 4 times less space it will also have 4 times less lifespan (less cells to wear level against), the fact that they only point out the lifespan on the 240GB version makes me think so.

Its also interesting to see the 4GB/day quote, so i decided to start up perfmon and measure how much data i do actually write to my HD's a day, i'll post back with my usage data tommorow.
 
Five times faster than the fastest 7200rpm drive...riiight. I didn't know drives these days only did 40mb/s. :rolleyes: Gotta love marketing BS. The Intel SSDs are faster too.
 
virtual RPMs? WTF?

If it's a solid state disk then it won't actually have any revolutions, right?

There's not really any good reason to use vRPM as a metric then. If it's a completely different design then why bother comparing to the old standards?
 
If it's a solid state disk then it won't actually have any revolutions, right?

There's not really any good reason to use vRPM as a metric then. If it's a completely different design then why bother comparing to the old standards?
It's just marketing.
 
I may be crazy.. but I kind of like the vRPM metric. Makes you appreciate just how much faster a future SSD in the future would be compared to disk. These drives may suck.. but it's still a cool idea.
 
But RPM wasnt really an indicator of harddrive speeds anyway. I have some 7200rpm's which are beat by newer 5200rpm drives, due to smaller platter sizes and technology improving.

IMO they should have just stuck with MB/s transfer speeds for consumer info.
 
Ok so i got my data for the last 24 hours, during this time i didnt download any big files, have a torrent program running or run any kind of disk defrags, i also didnt load any games, it was kept to browsing/watching videos (streamed over net)/emails/msn etc
My system is running Vista x64 and has 4GB.

Total bytes read = 7.16 GB
Total bytes written = 12.03 GB

So assuming that the 60GB has 40TB of write capacity then i have :
3325 days of writing use (40000 / 12.03)
9.1 years of writing use (3325 / 365)

Thats still not to bad and better than i thought it might come back with, obviously the higher capacity drives will have a longer life.
 
But RPM wasnt really an indicator of harddrive speeds anyway. I have some 7200rpm's which are beat by newer 5200rpm drives, due to smaller platter sizes and technology improving.

IMO they should have just stuck with MB/s transfer speeds for consumer info.

This is true, but as soon as SSDs become mainstream, disk platter technology will probably remain stagnant and you can pick a Vraptor as the basis. This is of course just for S&G.
 
This is true, but as soon as SSDs become mainstream, disk platter technology will probably remain stagnant and you can pick a Vraptor as the basis. This is of course just for S&G.

Still, vRPM is an attempt to show equivalent performance on a completely different technology where real RPMs are really only relevant to performance if things like areal density are taken into consideration; which is pointless since areal density doesn't affect performance on SSDs - at least not like they do with hard drives.

Apples and oranges.

Looking at Sandisk's formula, they base their vRPM completely on IOPS. Why don't they just give us the IOPS then? Damn marketers.
 
Back
Top