1. Home
  2. Electronics Design
  3. SSD vs. Rotating Media
  4. Page 2

SSD vs. Rotating Media

That is why I'm looking at a new SSD for my laptop. But this new tech looks pretty good. I watched the full presentation with some forward skipping and I noticed a few things. They brag how it is faster than NAND Flash (uh, what wouldn't be faster than that?) and denser than DRAM... well isn't Flash denser than DRAM, by a lot?

So they carefully avoided saying the technology is not as dense as Flash and not as fast as DRAM. So it won't be replacing DRAM directly. Since it is not as dense as Flash it likely won't be replacing that unless it is a *lot* faster which they seem to be saying it is. But clearly it will be at a price premium over Flash SSDs. It will also be sold in DIMMs to plug into motherboards but not as main memory, rather to bypass the slow SATA interfaces.

I guess I won't be plugging a 3D X-point SSD into my laptop for $250. But we'll see what my next computer has for storage... assuming it isn't a phone. Heck, the main use of this memory may well be in the server world and it may never make it to the consumer market.

--

Rick C
Reply to
rickman
Loading thread data ...

There's nothing wrong with storing hard drives in an offsite safe :). Their fire and other environmental hazards resistance is similar to plastic disks.

A word of caution though: I talked to some storage guys who claim that both magnetic disks and SSD disks in cold storage aren't as reliable as I thought. Apparently they should be powered up occasionally for maximum endurance, for different reasons: magnetic ones to exercise the mechanics, and SSD to let the firmware do its scrubbing/sector replacement.

Reply to
Przemek Klosowski

That kind of money easily gets you 8TB of HDD storage...

Reply to
Rob

You are right. I have a 5.25" full height harddisk drive and it is only 5MB. MEGAbyte. Capacities of up to about 80MB were common in that form factor, and I also once used an IBM 62RW100 that had 820MB. That was a beast of a harddrive in those days, with very fast seektimes and an impressive 2.5 MB/s transfer rate over SCSI.

9GB was a common (high-end) capacity in 3.5" full height disk drives. We once had a Compaq 6000 server which had 4 of those in RAID-5 for a database and then two 4.5 GB drives in RAID-1 because you had to have separate drives for the transaction log.
Reply to
Rob

In 2.5 inch form factor? I could only find 4TB in laptop size. Still, the point is the SSD is affordable at the common capacities used in home computers today.

--

Rick C
Reply to
rickman

On Thu, 28 Apr 2016 10:55:40 -0400, rickman Gave us:

SSD sucks. You want mSATA or M.2 They are way faster. Why do it half-assed unless you are related to Thompson or Larkin or krw or Terrell.

Reply to
DecadentLinuxUserNumeroUno

You deserve punishment for that. I sentence you to spend the rest of your life being the AlwaysWrong mad-dog miserable ranting jerk that you have been so far.

Oh, how's that business thing working out? Got any fat orders?

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

Good thing the common man has you to look out for them. In other words, you are just being a troll. You are not so bad when you stick to the technical. When you drop down to wallow in the mud you seem more comfortable though.

--

Rick C
Reply to
rickman

On Thu, 28 Apr 2016 08:58:13 -0700, John Larkin Gave us:

You wouldn't know.

Reply to
DecadentLinuxUserNumeroUno

I didn't say it would be in 1 drive...

Reply to
Rob

M.2 is made for laptops. In order to save pins, the maximum drive size is currently 512GB. This is about a year old, but covers the device and bus options fairly well:

You'll need a motherboard with PCI-e 3rd or 4th generation in order to get a major boost in performance with M.2:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

On Thu, 28 Apr 2016 11:10:34 -0700, Jeff Liebermann Gave us:

m.2 is made for PCIe, as in direct access thru the PCI bus as opposed to tertiary to it like SATA is, and it is extremely fast. Many newer motherboards have PCIe and or m.2 slots on them.

Stop guessing as you go.

Reply to
DecadentLinuxUserNumeroUno

On Thu, 28 Apr 2016 11:10:34 -0700, Jeff Liebermann Gave us:

No shit. So you know how to google. Even the early versions are "a major boost in performance" compared to a canned up "SSD" form factor old technology crap.

Reply to
DecadentLinuxUserNumeroUno

SSD's allegedly do rather badly for longer term storage. The storage element in an SSD is basically a capacitor, which will slowly self discharge. The data needs to be refreshed or you end up with a corrupted disk drive. Notice that SSD's leak more at low temperatures than at high temperatures. If you want to store them for extended periods, you're better off at high temperatures.

I've seen backup batteries for SSD's, but they're not for fixing long term storage problems. They're to keep the SSD up and running long enough for its cached to flush. I haven't seen any battery attachments (yet) that are intended to solve the alleged SSD long term storage problem.

Also, "scrubbing/sector replacement" is called "wear leveling". This is only useful when writing to an SSD, something it unlikely to happen when with just the power turned on:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

On Thu, 28 Apr 2016 19:45:16 -0700, Jeff Liebermann wrote as underneath :

I notice one of the comments under that article and quote one: "You stupid, stupid website - the JEDEC JESD218A standard was announced in September 2010, & the slides are from October 2010... What you've done is taken a blog post about data retention for court cases &, rather than doing any checking, have pretended that it's new news to create a panic & generate site traffic." So I dont know who has the right of it!? There are an awful lot of SSDs in use and not many failures it seems, at least in the circles I know.C+

Reply to
Charlie+

DecadentLinuxUserNumeroUno prodded the keyboard with:

formatting link

Agreed ! I took a duff drive apart some time ago, I accidentally dropped the platter to see it shatter as it landed on the tiled floor. I was surprised to see that it was made of glass.

--
Best Regards: 
                      Baron.
Reply to
Baron

On Fri, 29 Apr 2016 15:04:12 +0100, Baron Gave us:

Very few makers transitioned to glass platter substrates.

They are almost 100% Aluminum.

Reply to
DecadentLinuxUserNumeroUno

Den torsdag den 28. april 2016 kl. 07.25.47 UTC+2 skrev DecadentLinuxUserNumeroUno:

in the olden days the magnetic material was iron oxide aka "rust" afaik now it is some kind of cobalt material

-Lasse

Reply to
Lasse Langwadt Christensen

On Fri, 29 Apr 2016 07:38:19 -0700 (PDT), Lasse Langwadt Christensen Gave us:

Mag coatings on audio tapes have been perfected for decades, and the hard drive boys were just as quick to make high performance, high permeability coatings on the HD platter surfaces.

I have NEVER heard or seen any stories about "rust" on magnetic recordable surfaces in either the audio or data storage industries. Then there is video tapes the broadcast journalists have used IN THE FIELD, EXPOSED to all manner of environment and they are not rusting either.

Sheesh.

You guys are worse than the RoHS "lead is bad" idiots who cost the world's electronic industries billions to change over and more on worse performing CRAP (higher temp processes) and more on lost equipment (satellites and probes).

Napoleon left (as in retreated from) Moscow and returned with a mere

10,000 of his original 400k plus army. All because of Tin.
Reply to
DecadentLinuxUserNumeroUno

Good point as SSD technology has changed and improved considerably in the last 10 years. Here's the latest "B" version of the JEDEC document as of Mar 2016: You'll need to register and login in order to download it. If anyone is allergic to registering, I can email them the document. (660KB).

The changes are listed near the bottom of the document. One change stands out: Item 303.28: Refine the low-temperature data retention flow to accommodate the realities of the SILC retention mechanism in modern NAND. (...) For example, the standard client use condition is 40C active temperature and 30C power-off temperature, at which condition the data retention verified by this standard is the required 52 weeks (one year). On the other hand, if the actual power-off retention temperature were 25C, then this standard would ensure that the retention would be at least 105 weeks, or a little over two years.

Assuming you store your modern SSD at room temperature, it's good for

2 years. So, the comments are correct and the problem is NOT as bad as the article suggests. My apologies for my alarmist comments. However, for archival storage, I still would not trust an SSD that might only last 2 years. The 1/2" reel and DC600a backup tapes in my closet are over 10 years old and still good.

Graph D.1 "Expected retention (weeks) at different use temperatures" is rather odd. Retention time increases with a lower off (storage) temperature, but decreases with a lower operating temperature. Huh?

I'm off to do a political service call, treat myself to a massage, and meet a customer for a mythical free lunch. I'll read the document in detail later to see what I might have missed.

SILC = Stress Induced Leakage Current

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.