Is your SSD REALLY faster than your hard drive? Or do you just assume so?

[Malastare]

You just zeroed your credibility with this statement.

 

...

 

Not just most, but virtually all SSD failures in the wild, i.e. outside of endurance testing in a lab, have resulted in little aluminum bricks.

 

Citation please.

 

I've been working with datacenters for ten years. They find malfunctioning SSDs on blades a few times a month (but that is apparently still 30-50% less than they did with HDDs), and the majority of their failures are detected by the OS. They dump the logs to backup, re-image a new drive, restore the logs and move on.

 

I won't bother to reply to the rest because, while most of the data you presented is correct, your credibility has been zeroed.

 

Translation: "I won't talk to you, because you refuted a bunch of things I stated as facts, but you had experience that didn't match mine, so I'm going to claim that you are the one with no credibility".

 

Noted.

[Heal-To-Full]

Anyone with a clue about PC building understands that the heat generated by one component (like a CPU or GPU) can have a detrimental effect on other components (like an SSD) in "systems with inadequate cooling."

Well, I don't.

 

Perhaps I'm stupid.

 

Perhaps my Antec P-193 case is inadequate.

 

Perhaps I indeed don't understand how the heat generated by my CPU and GPU in the back of my case, some of it exhausted by my fans, also in the back of my case, the rest carried by my D5 pump to my MORA well outside the case, would have a detrimental effect on my SSD - mounted in a separate compartment, in the lower front of my case, with fans taking in fresh cool air ahead of them, blowing right on these SSD.

 

So it's up to you to explain to me - how?

 

And why did two SSD, which are supposed to be more reliable and not to give a crap about temperature, die a sudden death, why only one HDD over my 25 years of computing and close to a hundred hard drives, a worthless Seagate to boot, has done that, while subjected to the same temperature plus its own heat?

 

 

Yeah, I know I just got really lucky with HDD (being very picky about them) and unlucky with SSD, because I've been an early adopter, buying crappy Intels as early as 2008 and not waiting for quality companies like Micron to enter the game. But a lot of people got unlucky that way. A lot more people get unlucky, because they keep buying OCZ and Worstbuy and Kingcrap and whatnot.

 

But that's the reality ordinary users face. Sh-tty Intel/Worstbuy/Kingcrap SSD that will fail with zero warning, not advanced corporate SSD in RAID60 arrays. Unless they back them up to HDD, of course, like any functional person will, but then they don't get the weight savings or the noise savings, and the "reliability advantage" is the same as just having two HDD.

Edited November 14, 2014 by Heal-To-Full

[Transcendent]

why only one HDD over my 25 years of computing

 

Wow, you were lucky, I was forever doing RMAs for failed HDD.

[Malastare]

And why did two SSD, which are supposed to be more reliable and not to give a crap about temperature, die a sudden death, why only one HDD over my 25 years of computing and close to a hundred hard drives, a worthless Seagate to boot, has done that, while subjected to the same temperature plus its own heat?

 

The plural of anecdote is not evidence.

[BuriDogshin]

Perhaps I'm stupid.

 

/thread

[Heal-To-Full]

i can read sarcasm and pretend it's not sarcasm, ha-ha

And it's still up to you to explain your statements.

 

How did heat generated by my CPU and GPU in the back of my case, some of it exhausted by my fans, also in the back of my case, the rest carried by my D5 pump to my MORA well outside the case, would have a detrimental effect on my SSD - mounted in a separate compartment, in the lower front of my case, with fans taking in fresh cool air ahead of them, blowing right on these SSD.

 

And why did two SSD, which are supposed to be more reliable and not to give a crap about temperature, die a sudden death from that.

 

And to answer the question: would it be wise for a normal home user to use a SSD without backing its entire contents up on a HDD, or would it be utterly moronic?

Edited November 14, 2014 by Heal-To-Full

[BuriDogshin]

And it's still up to you to explain your statements.

There's no need. My statements are perfectly clear, and I trust the audience to be able to distinguish between what I wrote and what you, out of either dishonesty or stupidity, are trying to claim I wrote.

 

Now, where are your studies demonstrating that SSDs are less reliable than HDDs? No unverifiable stories of personal experience please, and no anonymous posts either.

 

Or did you just make that up?

Edited November 14, 2014 by BuriDogshin

[Sivn]

For SWTOR i dont noticed a difference if its on my SSD. I put it down to the **** engine the game uses.

 

SSDs really only affect a handful of games, Skyrim and WoW are 2.

 

Main use for SSD in my mind is for windows and apps/programs to be installed on. Documents, games andanything else is just fine on HDD.

[Heal-To-Full]

There's no need.

There is. You accused me of frying my own SSD - at the same time as claiming SSD are more reliable, which would imply, among other things, they aren't as liable to get fried.

 

"Fried" in respect to +38C being quite an exaggeration... I wish I could fry steaks at +38C.

 

In fact, I know, and know well, and am paid to know, that SSD don't give a crap about temperature until it's nearing their capacitors' degradation range. Any time you have SSD and HDD in the same enclosure, there's zero risk to the SSD until the HDD have long been rendered inoperable. HDD are completely gone from any military and industrial equipment specifically for this reason - HDD care about heat, SSD don't.

 

 

So with your statement about how I have supposedly overheated my SSD:

Anyone with a clue about PC building understands that the heat generated by one component ... perhaps we now know how you bricked those SSDs.

You lied.

 

Someone else, they might not know better, but you know better, and you consciously made that judgement call, you decided to lie.

 

So forgive my lack of tears - after being caught on that, whatever real or otherwise qualifications you possess, if you choose to disregard your knowledge and lie to suit your agenda, all of the credibility you would otherwise have is also forfeit.

 

 

Now, where are your studies demonstrating that SSDs are less reliable than HDDs?

Where's your

?

 

SSD are not less reliable than HDD. SSD are ~as reliable as HDD: http://www.storagesearch.com/reliability.html

 

So there's no reliability advantage.

 

But since there is no warning to SSD failure and the data is lost forever, they still have to be backed up on HDD. No weight savings, no noise savings, you still have the HDD, and then you add a SSD to speed it up.

Edited November 14, 2014 by Heal-To-Full

[Cernow]

I upgraded my main system drive from a WD Caviar Black HDD to a fairly decent spec Samsing EVO SSD a few months ago. It has Windows 7, apps and games installed on it. I use a regular HDD for file storage.

 

The SSD brought some huge improvements to certain things, but not to others.

 

Windows boots in about half the time it did before. Photoshop and similar apps load in about half the time too.

 

Some games benefit massively from it. Loading times in WoW and Wildstar are much faster ,as are shooters like Borderlands 2.

 

But SWTOR has gained ZERO benefit from the SSD. Planet loading times are almost identical to what they were before. Players and NPC still take time to load in crowded areas like the Fleet etc. I haven't been able to detect any tangible improvements to TOR's gameplay with the SSD. So it would seem that TOR's long loading times are due to server side issues. Which would make sense given how the majority of players tend to zone into new planets, new phases and instances etc at pretty much the same time, regardless of hardware differences.

Edited November 14, 2014 by Cernow

[BuriDogshin]

In fact, I know, and know well, that SSD don't give a crap about temperature.

Wrong. Totally wrong. Just about all sub-micron semiconductors care about temperature, because of a variety of physical characteristics that are temperature related, such as electromigration, leakage currents, thermal noise, and, especially in MLC and TLC Flash, higher bit error rates (which usually get corrected by the built-in ECC bits and re-read operations, but not always) . Do you know nothing about semiconductor device physics? Nothing at all?

 

And now you write:

SSD are not less reliable than HDD. SSD are ~as reliable as HDD: http://www.storagesearch.com/reliability.html

 

But earlier you wrote:

SSD reliability is abysmal - being a budding technology, you have much more chance of a SSD failing than a HDD

 

It appears that when pressed to find some data to back your shenanigans, the best you could do was find that study saying, you claim, that SSDs are just as good. Which means you pulled your earlier claim from that dark stinky place that many of us suspected you had pulled it from.

 

And you dare to claim that other people have zero credibility? That's laughable.

 

Seriously, haven't you embarrassed yourself enough for one thread?

Edited November 14, 2014 by BuriDogshin

[Heal-To-Full]

Wrong. Totally wrong. Just about all sub-micron semiconductors care about temperature, because of a variety of physical characteristics

No.

They do care, yes - but all of that care is internalized. To an external entity, such as a PC builder, things are very simple: SSD don't give any more of a f-ck about temperature than your other electronic components.

 

They won't fail until other components start to fail from overheating.

You claim I fried my SSD, located in the same enclosure as my HDD (which still do fine) by overheating them with my CPU/GPU.

 

Well, back it up. Give me one source indicating that SSD fail at lower temperature than HDD.

 

 

 

It appears that when pressed to find some data to back your shenanigans, the best you could do was find that study saying, you claim, that SSDs are just as good.

Yes. Because the study refers to high-end corporate SSD.

 

Most SSD the readers of this thread might buy is cheap OCZ, Intel, Worstbuy or Kingcrap junk.

Which is well known to fail for no better reason than lazy lunch break coding.

Edited November 14, 2014 by Heal-To-Full

[Malastare]

effect on my SSD - mounted in a separate compartment, in the lower front of my case, with fans taking in fresh cool air ahead of them, blowing right on these SSD.

 

Cosmic rays? Random EM radiation sources? Who knows.

 

I don't care about what caused your problems. What I care about is that, on average, over the thousands of SSDs and HDDs installed in the world today, that SSDs have a lower failure rate than HDDs. You did your study of ... 5 hard drives and 3 SSDs? Well, other people have done studies of thousands.

 

Sorry. I'll trust them over you.

 

And to answer the question: would it be wise for a normal home user to use a SSD without backing its entire contents up on a HDD, or would it be utterly moronic?

 

Backing up an SSD to an HDD is better than not backing up at all. But backing up an SSD to another SSD would be better, as SSDs are more reliable. However, it would also be more expensive. I'd probably just go with the HDD. Chances are pretty rare that they'll both die.

Edited November 14, 2014 by Malastare

[Orizuru]

When comparing a single SSD to a single HDD, I think its safe to say the HDD is slightly less reliable. However, it takes significantly more SSDs to equal the amount of storage you can get from a single HDD. If you begin to look at it from the view point of a 1000 Terabyte array of drives the reliability of the two drive types tends to level out until it's almost equal. Edited November 14, 2014 by Orizuru

[Heal-To-Full]

Backing up an SSD to an HDD is better than not backing up at all. But backing up an SSD to another SSD would be better, as SSDs are more reliable.

SSD are equally reliable... at best.

 

SSD are more durable to environmental conditions, but are more liable to fail without warning due to poorly tested coding in their firmware. This will get better, but that will take time. So far, they have no reliability advantage even if you pick the best SSD. If you buy OCZ/Intel/Kingcrap, they're worse.

 

Few people can afford to back up their SSD to another SSD - which BTW has to be disconnected most of the time so as not to fail simultaneously. Few people can move all of their HDD data to a SSD. Even fewer can do both at once.

 

So the reality is, installing a SSD in your PC won't make it more reliable, or lighter, or quieter, or more energy-efficient.

It will make your PC faster in running the software you choose to install on the SSD. A little faster for some, a lot faster for others.

 

Is that worth the cost? If you're already good on all the important parts (i.e. have a quad+core CPU and a discrete GPU that isn't a GT710 and a good display), definitely.

But don't expect it to magically turn your PC into something unbreakable and silent.

Edited November 14, 2014 by Heal-To-Full

[BuriDogshin]

SSD are more durable to environmental conditions, but are more liable to fail without warning due to poorly tested coding in their firmware. This will get better, but that will take time. So far, they have no reliability advantage even if you pick the best SSD. If you buy OCZ/Intel/Kingcrap, they're worse.

Citation to a published study or it's just shenanigans. Which you seem to be full of.

[Ivy_Blackrose]

an SSD will only improve LOAD times it really does not improve gameplay performance

[DarkSaberMaster]

Cliff notes?

[MSchuyler]

SSDs are faster than HDDs. Even a slow mid-range SSD will load faster than a high-speed HDD. The difference will be reliable and measurable. It might be a matter of 5 seconds or 30 seconds. The faster the SSD and the slower the HDD, the more it will matter. It's up to each user to decide just how much that is worth to them.

 

Sadly, your experiments weren't rigorous enough to say anything more than that, and in many cases, the lack of rigor undermined their validity.

 

No, they didn't. I take yours as a typical misperception and representative of others here. The question at hand is NOT whether SSDs are "faster" than HDDs, but thank you for pointing that out once again, in color, the same thing I said in my post three or four times. Maybe someone really did miss it. Maybe everyone needed to be reminded half a dozen MORE times--just to make sure. Thanks again.

 

The ISSUE is how much time you REALLY SAVE by using an SSD. It doesn't really matter if the test meets YOUR definition of "rigorous" or not. I'm living in the Real World, not a laboratory where we can parse these things out to a hundredth of a second and separate out processes OTHER than disk I/O because that is clearly missing the point. Disk performance is NOT separate from everything else that is happening, something I took special pains to point out and address already. Your repeating it as insightful while not admitting I ever said it does not change that fact.

 

I SHOWED you how a nearly FOUR HOUR game session saved less than TWO MINUTES in all that time and that the ONLY difference between the two was using an SSD instead of an HDD. All other variables were as equal as they could be.

 

Does ANYONE here not get this part? You save less than TWO MINUTES! Just ballparking it here and not worrying about the "rigorous precision" some people feel compelled to ask for, that is well LESS THAN 1%. (0.7348%)

 

And you are telling me with a straight face, without falling down on the floor clutching your stomach laughing at the absurdity of it, that THIS amount of "saved time" means we all ought to rush out and install an SSD and everything will be all better and the game and your PC will run SIGNIFICANTLY faster?

 

For a LESS than ONE PERCENT difference?

 

Folks, the Emperor has no clothes. Put another way, putting lipstick on a pig doesn't do anything. The idea that using an SSD instead of an HDD will significantly improve your game play is a myth. The seconds you save are not even enough to give you a potty break. If you THINK it's worth it, well, in your mind it clearly is, but not in any objective sense of the matter.

Edited November 14, 2014 by MSchuyler

[DOHboy]

Perhaps you think that SSDs are the same as SD cards and thumb drives. They're not. They have sophisticated controllers and some of the biggest strengths of SSDs are the random access read/write times and their blazing fast latency.

 

No, will concede I got my info swapped on the random read/writes, for some reason (don't ask me why) I thought a slow moving head was faster at accessing data than a hashtable/offset, chalk it up to brain fart.

 

The point of SSDs has always been, to me at least, fast access of data, slow(er) writing of data. They have improved significantly over the years, but reading data quickly has been their forte. The bit about the CD/DVD-rw was more if you are writing the entire disk every time, vs reading the entire thing every time.

 

I guess a better example would have been DB that is large and fairly static that is accessed multiple times is better on an SSD. A Database that is constantly being updated and modified is probably better on an HDD long term.

[BuriDogshin]

Cliff notes?

SSDs rool, HDDs drool.

[BuriDogshin]

The ISSUE is how much time you REALLY SAVE by using an SSD.

Well, for some people, sure.

 

Not for me. For me it is how much annoyance I avoid. To wit:

Waiting for a PC to boot annoys me. Waiting for an application to load annoys me. SWTOR load screeens annoy me. And an SSD, because it is faster, reduces my exposure to these annoyances.

 

IBM back in the 70's did a study that showed how even minor amounts of waiting for a computer to respond, in the study, a one-quarter-second wait after hitting return when using a command line interface, caused a significant drop in productivity. People's minds would wander off the task they were performing during the wait, and then when the wait was over they had to remember what they were doing. So i think a lot of the perceived benefit, and perhaps the real benefit as well, of SSDs is in the substantial reduction in waiting around that they enable.

[Heal-To-Full]

Not for me. For me it is how much annoyance I avoid. To wit:

Waiting for a PC to boot annoys me.

It annoys me too! Fortunately, the last time I watched my PC boot was in August.

[MSchuyler]

I challenge you all to do a test yourself. The only caveat I throw out is that it must be a "Real World" test. Laboratory results are virtually meaningless. Play the game using your SSD. Play your game using your HDD. Tell us how much time you saved and document your work. Just WHERE did you save the time? Don't worry about someone else jumping in complaining about your test protocol. Anyone can sit back in their armchair and point fingers, but they aren't doing any real work. THEN decide if it's really worth it.

[BuriDogshin]

I challenge you all to do a test yourself. The only caveat I throw out is that it must be a "Real World" test. Laboratory results are virtually meaningless. Play the game using your SSD. Play your game using your HDD. Tell us how much time you saved and document your work. Just WHERE did you save the time? Don't worry about someone else jumping in complaining about your test protocol. Anyone can sit back in their armchair and point fingers, but they aren't doing any real work. THEN decide if it's really worth it.

 

So, when I get a queue pop for a flashpoint, and have to wait around while someone with a slow internet connection, system, or HDD finishes loading in, which happens to me a lot BTW, do I count that as time saved because of my SSD?

 

Ah, but a bigger problem: I don't want to install an HDD into my system just to test SWTOR performance.

I manage to get by on 300+ GB of SSD and my NAS boxes. I do not want to run SWTOR from any of the NAS boxes, even if it is linked to the PC by gigabit Ethernet. They are not built for speed.