[dazzerfong]

  Really should not.
 
I received a warning AGAINST using SDD - for any application that requires multiple writing/erasure/format of the card/disk. Solid state elements are still not fully reliable and the more you tax them, the likelihood of them to self destruct is greater. Leading to detrioration of the card at first, up to the total failure in the end.
 
In the SSD case, burning-in could well mean burning-out - quite literally so. 

Maybe a while ago, but not really anymore. HDD's have mechanical components which will fail, or worse suffer from a crash. I'll be honest, I've given up on chasing the most 'reliable' system: they're so neck-and-neck nowadays. However, if power loss is frequent, SSD's suffer more from it.

 

[Thad-E-Ginathom]

... ... ... either works - or it doesn't.

 
 
Now that's the kind of audible difference that I like to deal with.
 
...and it doesn't just mean sound or silence: I have had plenty of unpleasant experiences with dropouts or distortion with computer playback. The answers,  though, (apart from "throw this machine away: it is  never going to make a good audio box.") although seldom simple, are always mundane, and never, ever, magical.

 

[maverickronin]

  Are most individuals in this thread from an engineering/science background?
 
I want to know if these magical effects only affect those that are less scientifically inclined? Or if it's a basic matter of common sense, in which case, I fear for the rest of head-fi.

 
I don't have any formal training in engineering or science.  I work in IT/tech support.  I'm pretty much all self taught and a "jack of all trades, master of none" type.  I'm a skeptic, maybe even with a capital "S", as that would appear to be the best way to achieve a proper understanding of reality.
 
I don't really think that just having a background in science or engineering will save you from this kind of stuff.  Most of what's formally taught just specific knowledge and just knowing things won't necessarily save you from believing in nonsense.  After all, there are tons of people in the "audiophile" industry who believe tons of crazy things despite having such a background.
 
The issue is knowing how to think, rather than what to think.  People believe in all this craziness for the same reasons that they believe other crazy thing like alien abductions, grand conspiracy theories, and "alternative" medicine.

 

[bfreedma]

  Really should not.
 
I received a warning AGAINST using SDD - for any application that requires multiple writing/erasure/format of the card/disk. Solid state elements are still not fully reliable and the more you tax them, the likelihood of them to self destruct is greater. Leading to detrioration of the card at first, up to the total failure in the end.
 
In the SSD case, burning-in could well mean burning-out - quite literally so. 

 
This was true of the first generation of SSDs but is no longer the case.  The introduction of proper garbage collection techniques, TRIM, for example at the OS level alleviated the issue.  Most manufacturers now build garbage collection, wear leveling, and over provisioning into the SSD controller itself, further reducing the risk of failure due to multiple writes to the same cell.
 
Every test I'm aware of show that SSDs are now significantly more reliable than spinning media disks, both for home use and at the enterprise level.  It's reached the point where the failure rates are no longer determined by the actual data volume written to create failure but are now calculated by the estimated wear rate of the solid state cells based on multiple petabytes of read/write cycles.
 
Bottom line - solid state memory is about as reliable as any storage media you can acquire.  As always, random failure is possible and all data should be backed up - regardless of primary storage technology.

 

[superjawes]

Burning in SD cards.. Shouldn't get worse than that, give it a scream in some threads and have fun.. Lol

Really should not.

I received a warning AGAINST using SDD - for any application that requires multiple writing/erasure/format of the card/disk. Solid state elements are still not fully reliable and the more you tax them, the likelihood of them to self destruct is greater. Leading to detrioration of the card at first, up to the total failure in the end.

In the SSD case, burning-in could well mean burning-out - quite literally so. 

Basically, no.

It is technically true that writing to SSD memory will eventually lead to the drive's death, but it takes a lot of data to kill a consumer SSD...and I do mean a lot.

Tech Report set out to kill some SSDs by writing boatloads of data to them (the linked article was written after the final sample died), and while most users probably aren't going to write more than a couple terabytes per year, every one of TR's samples wrote HUNDREDS of terabytes before failing (note: that is well above what the drives are rated for).

In short, as long as you're using an SSD under semi-normal conditions, you don't have to worry about killing one with data writes.

 

[castleofargh]

  Are most individuals in this thread from an engineering/science background?
 
I want to know if these magical effects only affect those that are less scientifically inclined? Or if it's a basic matter of common sense, in which case, I fear for the rest of head-fi.

 
I was in a math oriented curriculum until I was about 20years old(turn that into whatever study that is in your country), then I went into a photography school...(what an idiot) I never did anything even remotely related to science or engineering again and forgot all about ..   what?

 
my mother has always impressed me with how well educated and bright she could be(lol mother complex lol). but she just doesn't know how to skeptic. the concept itself is like a foreign object. she's kind of skeptical when she first learns about something, meaning she doesn't just intake whatever people tell her. she's not so much the gullible type. but once she accepts an idea, it's almost impossible to make her come back on it.
 
me on the other hand, I could have triple checked an information and believed it all my life, if a homeless guy on the street tells me it's false, I might just end up checking again and factoring whatever the guy told me.(that's why analogsurviver makes me think so much) I always have that lingering "anybody can be wrong, maybe I missed a part too" at the back of my head. it's annoying as hell at times, but it also makes me improve so I'm cool with it.

 

[arnyk]

  Really should not.
 
I received a warning AGAINST using SDD - for any application that requires multiple writing/erasure/format of the card/disk. Solid state elements are still not fully reliable and the more you tax them, the likelihood of them to self destruct is greater. Leading to detrioration of the card at first, up to the total failure in the end.
 
In the SSD case, burning-in could well mean burning-out - quite literally so. 

 
I switched my customer production of desktop PC's and upgrades of laptops, to SSDs well over a year ago after using them personally for my own PCs for at least another year before that.
 
The warning you received would effectively forbid the use of SSDs for any boot volume because of the read/write load caused by swapping and paging. They typically amount to about 50% of the I/O load on a well-designed system, and they are in turn about 2/3 reads and 1/3 writes. IOW a ton of writes of new data which according to the advice above would be quickly fatal.
 
The first year or so that used SSDs for my own use (as a test) I had about equal numbers of SSD failures as HD failures, but the SSD failures were plain old electronic failures of the interface circuits, not wearing out of the flash memory due to running out of life for new writes.
 
The second year of general application of SSDs found them to be running essentially without any failures at all. All my systems are RAID arrays with just one volume for both boot and data whether HD or SSD.

 

[analogsurviver]

Basically, no.

It is technically true that writing to SSD memory will eventually lead to the drive's death, but it takes a lot of data to kill a consumer SSD...and I do mean a lot.

Tech Report set out to kill some SSDs by writing boatloads of data to them (the linked article was written after the final sample died), and while most users probably aren't going to write more than a couple terabytes per year, every one of TR's samples wrote HUNDREDS of terabytes before failing (note: that is well above what the drives are rated for).

In short, as long as you're using an SSD under semi-normal conditions, you don't have to worry about killing one with data writes.

If it is master recordings, I mean #001 RAW recording, I most definitely do not want to take chances. HDDs fail at a sustainable, non catastrophic rate - if in reasonable RAID configuration.
 
Audio and video combined for the two last concert  recordings amounted to just a bit under 300GB. Two DAYS ...
 
SSDs of 2TB are simply too expensive - and anything with smaller capacity is, for all practical purposes, too small to consider to use under RAID. 
 
There is a reason why USAF flew, basicaly, 286 ( or whatever2XY) processors, despite newer, more powerful ones were available.
 
RELIABILITY - and ironing out all the teething troubles. Computer geeks might be tempted to fiddle with SSD - I will wait until it is proven by time.
 
What good did the analog tapes from the 80's compared to older stuff ? Yes, one could put MUCH hotter signnal on it, more treble, prior to saturation ; but after a few decades, these new tapes can - disintegrate in your hands to dust, whenever the old stuff still works as intended. See what I mean ?

 

[bfreedma]

Basically, no.

It is technically true that writing to SSD memory will eventually lead to the drive's death, but it takes a lot of data to kill a consumer SSD...and I do mean a lot.

Tech Report set out to kill some SSDs by writing boatloads of data to them (the linked article was written after the final sample died), and while most users probably aren't going to write more than a couple terabytes per year, every one of TR's samples wrote HUNDREDS of terabytes before failing (note: that is well above what the drives are rated for).

In short, as long as you're using an SSD under semi-normal conditions, you don't have to worry about killing one with data writes.

If it is master recordings, I mean #001 RAW recording, I most definitely do not want to take chances. HDDs fail at a sustainable, non catastrophic rate - if in reasonable RAID configuration.
 
Audio and video combined for the two last concert  recordings amounted to just a bit under 300GB. Two DAYS ...
 
SSDs of 2TB are simply too expensive - and anything with smaller capacity is, for all practical purposes, too small to consider to use under RAID. 
 
There is a reason why USAF flew, basicaly, 286 ( or whatever2XY) processors, despite newer, more powerful ones were available.
 
RELIABILITY - and ironing out all the teething troubles. Computer geeks might be tempted to fiddle with SSD - I will wait until it is proven by time.
 
What good did the analog tapes from the 80's compared to older stuff ? Yes, one could put MUCH hotter signnal on it, more treble, prior to saturation ; but after a few decades, these new tapes can - disintegrate in your hands to dust, whenever the old stuff still works as intended. See what I mean ?

No matter how you spin it, you are wrong about the reliability of SSD storage. Changing the topic to cost is a different discussion entirely.

As to "computer geeks" being the market for "fiddling with SSD", you might want to take a look at the current offerings from every major enterprise storage vendor and ask yourself if the Fortune 500 and FTSE 1000 customers they serve believe they aren't reliable.

You're also misrepresenting the reason that the military tends to be behind the curve in chip utilization. Design specs are finalized years and in some cases up to a decade before the plane/ship/whatever enters service. Nothing would ever reach production if the mil specs were reconstituted every time a newer/faster CPU became available. The reason for seeing older components is most definitely not due to the perception of reliability, because when those chips were specified, they WERE the new tech.

 

[cel4145]

There is a reason why USAF flew, basicaly, 286 ( or whatever2XY) processors, despite newer, more powerful ones were available.

No. The real reason is that the programming costs of getting the code error free with lots of redundant testing of the code to ensure that is way more expensive than the processors.

Computer geeks might be tempted to fiddle with SSD - I will wait until it is proven by time.

SSDs have been around for almost 25 years. What are you waiting for? 50 years?

 

[analogsurviver]

No matter how you spin it, you are wrong about the reliability of SSD storage. Changing the topic to cost is a different discussion entirely.

As to "computer geeks" being the market for "fiddling with SSD", you might want to take a look at the current offerings from every major enterprise storage vendor and ask yourself if the Fortune 500 and FTSE 1000 customers they serve believe they aren't reliable.

You're also misrepresenting the reason that the military tends to be behind the curve in chip utilization. Design specs are finalized years and in some cases up to a decade before the plane/ship/whatever enters service. Nothing would ever reach production if the mil specs were reconstituted every time a newer/faster CPU became available. The reason for seeing older components is most definitely not due to the perception of reliability, because when those chips were specified, they WERE the new tech.

Oh - now tell me why military is still using NiCD cells instead of NiMH, and let alone LiWhatever ? Regardless that they are inferior to their newer counterparts in amount of energy stored per unit of volume. Maybe the larger temperature envelope and robustness in both mechanical and electrical sense have something to do with this decision.
 
If the military actually uses the latest gear available at the time of specification, than this explains the track record of the early F-16s ...

 

[bfreedma]

No matter how you spin it, you are wrong about the reliability of SSD storage. Changing the topic to cost is a different discussion entirely.

As to "computer geeks" being the market for "fiddling with SSD", you might want to take a look at the current offerings from every major enterprise storage vendor and ask yourself if the Fortune 500 and FTSE 1000 customers they serve believe they aren't reliable.

You're also misrepresenting the reason that the military tends to be behind the curve in chip utilization. Design specs are finalized years and in some cases up to a decade before the plane/ship/whatever enters service. Nothing would ever reach production if the mil specs were reconstituted every time a newer/faster CPU became available. The reason for seeing older components is most definitely not due to the perception of reliability, because when those chips were specified, they WERE the new tech.

Oh - now tell me why military is still using NiCD cells instead of NiMH, and let alone LiWhatever ? Regardless that they are inferior to their newer counterparts in amount of energy stored per unit of volume. Maybe the larger temperature envelope and robustness in both mechanical and electrical sense have something to do with this decision.
 
If the military actually uses the latest gear available at the time of specification, than this explains the track record of the early F-16s ...

What part of my design cycle description escaped you? And when did the topic change from SSD storage to CPUs and now batteries which, obviously, have completely different operating envelopes? And of course the military utilizes equipment available at the time of the development of the specification. How do you build a plane with a CPU that doesn't exist, let alone convince the military to fund it as a private contractor?

You are a master of changing the subject - let me try the same.
How are your CD Mat test files coming along?

All kidding aside, you should really stop bloviating on technologies and processes you fairly obviously aren't familiar with.

 

[analogsurviver]

No. The real reason is that the programming costs of getting the code error free with lots of redundant testing of the code to ensure that is way more expensive than the processors.
SSDs have been around for almost 25 years. What are you waiting for? 50 years?

Yes, you could say that the programming to get error free operation is the reason. Sometimes, real operation brings about conditions no programmer could reasonably forsee - yet the equipment should not go berserk even under FAR OFF conditions. And finding such exceptions that are under certain conditions fatal does take time in order to write programmes that are capable of dealing with the real thing.
 
CDs have been around for more than 25 years - and some no longer can be read - by anything. Any vynil record ever made that has been handled with normal care and stored under reasonable conditions is still playable.
 
SSDs can not claim such test of time. Would you trust them your entire photo collection ?

 

[bfreedma]

Quote:Originally Posted by cel4145 No. The real reason is that the programming costs of getting the code error free with lots of redundant testing of the code to ensure that is way more expensive than the processors.SSDs have been around for almost 25 years. What are you waiting for? 50 years?Yes, you could say that the programming to get error free operation is the reason. Sometimes, real operation brings about conditions no programmer could reasonably forsee - yet the equipment should not go berserk even under FAR OFF conditions. And finding such exceptions that are under certain conditions fatal does take time in order to write programmes that are capable of dealing with the real thing. CDs have been around for more than 25 years - and some no longer can be read - by anything. Any vynil record ever made that has been handled with normal care and stored under reasonable conditions is still playable. SSDs can not claim such test of time. Would you trust them your entire photo collection ?

No media should be trusted as the sole point of storage. Are you suggesting you don't feel the need to back up data you store to spinning platters?

You're on quite the fishing expedition to attempt to support your position. Catching nothing but old boots so far....

 

[analogsurviver]

What part of my design cycle description escaped you? And when did the topic change from SSD storage to CPUs and now batteries which, obviously, have completely different operating envelopes? And of course the military utilizes equipment available at the time of the development of the specification. How do you build a plane with a CPU that doesn't exist, let alone convince the military to fund it as a private contractor?

You are a master of changing the subject - let me try the same.
How are your CD Mat test files coming along?

All kidding aside, you should really stop bloviating on technologies and processes you fairly obviously aren't familiar with.

CD mat test - please see a few posts back, from today.
 
I know I would not want some untried CPU in my latest gear - I would prefer to opt to something less capable but available long enough to be trustworthy. An example : when Windows 7 became available, I decided to upgrade from XP. But, the early Win7 was so horribly inferior to XP in audio, that I changed back to XP in less than a week. Win7 did, eventually, catch up with XP in audio - but it took almost a year to do so.
 
I never mentioned using , at the time of spec, unavailable products in the process of development. It is the ultimate recipe for disaster.