Sink your teeth into OCZ's ARC 100 SSD sizzler with tasty home-grown chips After Toshiba’s acquisition of the company, OCZ has in one fell swoop joined the elite band of SSD manufacturers who are in the extremely fortunate position of being able to source “in-house” all the internal chips that matter in a drive: namely, the NAND and the controller. OCZ ARC 100 SSD Teaming up with Tosh: OCZ's ARC …
I know people have (rightly) slated OCZ in the past for their build quality but I can't say I've had the same experience:
Intel 330 120GB: 14 months, bad sectors, Amazon refunded
Crucial M500 mSata 120GB: 10 months, dead, RMA'd and replaced
OCZ Vertex 2E 60GB: 4 years 4 months, dead, binned
Intel 520 180GB: 18 months, alive and kicking
Kingston SSD Now+ 120GB: 2 years 6 months, alive and kicking
Crucial M500 mSata 120GB: 6 months, alive and kicking
Kingston Hyper X 120GB: 2 years 7 months, alive and kicking
Talk to insiders and you'll find out there were some horrendous bugs in the firmware / silicon. The new stuff's subject to vastly more rigorous testing. Their biggest problem is going to be getting past the crap reputation they've gained (deservedly) over the last few years.
I have ten here, Vertex2, 4 Vertex 4, 4 Agility 3, and the PCIe SSD, all performing flawlessly. That's more than I can say about the competition who make up the rest of two-dozen for 3+ TB. I've fallen seriously in love with cartridge-SSD's to reconfigure on the fly here. I've done that with spinners but it's not anywhere as flexible, nor as fast (random access).
I actually found a use for a Kingston SSD (the last Kingston I will ever purchase) as a spacer under one of the feet of the washing machine. Worked marvellously in that role. Watching it slowly rot on the bathroom floor kind of served a therapeutic purpose as well.
The trouble with OCZ in the past was that they had to source their NAND on the open market, which IMO made the quality of the drives variable. Even within a particular model you would never know if you were getting good quality NAND from a known manufacturer, or dodgy stuff that was made in an unknown plant in China.
Now that they're owned by Tosh, now using NAND of a known quality, and given that the whole SSD tech has matured since the OCZ reliability debacle I'd be prepared to give them a go again.
One thing that I will not do these days is touch an SSD brand that does not have the backing of a major NAND fabrication plant. No more SSDs for me where the manufacturer is scrabbling around on the open market for NAND chips.
I haven't had any experience with ocz SSD's but my heavily overclocked OCZ 1200mhz flex ddr2 has allowed me to still currently play anything virtually on max with an overclocked q6600 dinosaur (using proper updated GPU muscle of course) it'll be a sad day when it gets listed on eBay next year.
I'm surprised by this endurance rating.
Assuming it's for the 120GB drive (worst case), that means the drive can only be rewritten 183 times, doing a straightforward division and not accounting for write amplification and whatnot.
I thought even the worst flash cells could be rewritten 1000+ times?
Is this endurance rating typical?
Seems low doesn't it, but when you consider typical consumer useage, which where this drive is positioned, then it's unlikely that the drive will be completely overwritten once, assuming you discount any swap activity and any stuff the system is doing in the background - consumers tend to take a fill and then upgrade approach to storage. Even when you take the system activity, garbage collection, resaving documents etc, into consideration then the drive is still only likely to have encountered a mere fraction of it's rated endurance over it's lifetime.
I baulked at my 1TB M550's 72TB write endurance, but then I considered how I would actually use it and realised that I'll never get close to writing 72TB to that SSD.
Another thing to consider is that in the "The SSD Endurance Experiment" - http://techreport.com/review/25889/the-ssd-endurance-experiment-500tb-update - every single drive went way beyond their rated endurance, inluding TLC drives.
Interesting. Thanks for the link.
I don't think of myself as someone who often writes a lot of data to his drive but I just checked and I'm averaging 2GB/day. Still only 1/10th of the 20GB mentioned, but again, I don't write a lot of stuff. I can imagine that if I did different work I would easily exceed that amount. Good to know that there seems to be a margin of error in these drive specifications.
"Is there a guideline for the total amount of HDD/SSD data likely to be written to the swap space (in typical use) over a given timespan, for a given amount of swap space and physical RAM?"
I can't imagine there is. "Typical use" covers too much ground.
If you're using a Mac, you can go to Activity Monitor, Disk tab, and see how much data has been written since you last rebooted. You can then run 'uptime' on the command line and do some division to see how much data you typically write per day.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
