need help deciding...

i am looking at getting an SSD for my Envy (in sig). i have been eyeballing the Samsung 830 and Crucial M4, both in 128GB and 256GB flavors. i can get the 128GB Samsung right now (99.99 on amazon/newegg) or i can wait a few weeks and get the 256GB of either.

so my question is: for those that have 128GB SSD, are you hurting for space? i'm a very impatient person, and frugal, too, so i want to get the 128GB Samsung right now, but i'll wait and get the 256GB if people are finding space to be an issue.

 

I personally regret getting the 128GB Samsung, I suggest to go for the 256GB.

 

If space was the only issue; that is something you could (potentially) work around.

What you can't work around is performance (specifically; sustained performance over time) and that is where you need the higher capacity drives. Especially given that a modern O/S (Win7x64) plus programs needs at least 50-80GB to be productive with...

My viewpoint is why pay an SSD premium to only achieve HDD results? (And with the right (wrong?) workload, an SSD can be much worse than a HDD in a matter of a few weeks...). Spend more to get the highest performing option (new/unused states) and don't fill your drive to much more than 70% to keep enjoying those speed benefits indefinitely.

To abuse the SSD (like you can a HDD...) and keep performance almost as high as when you bought it: Partition it to much less than full capacity IMMEDIATELY and consider the unused capacity as a performance/reliability insurance cost of utilizing an SSD.

(NOTE: To simply use/fill the drive to less than 100% capacity is not the same as partitioning it (when new) and ignoring (indefinitely) the 'unused' partition).

In my systems: for O/S (C: ) use: I use 100GB out of 240GB-256GB and I still see the performance drop a little (less than 5%...) from brand new/out of the box performance to fully installed/configured setups with all Windows updates, programs/updates and drivers/utilities installed (about 80GB worth of Win7x64 O/S + software). On Scratch Disk SSD's - I use the SSD's at 50% capacity (partitioned, of course...). On strict DATA drives - I have used the SSD's up to 80% full (before I move/delete older data to multiple HDD's).

The above strategy gives me the (SSD) performance I paid for; sustained over an indefinite period of use. The cost? SSD's cost me more than double the $$/GB than most people; but the workflow advantages (daily and continuously...) more than offset this (relatively small) one time cost.

If you are just eager to jump on the SSD bandwagon; get the 'toy-like' capacities of 128GB or less: if you have a real need (or desire) for the highest, sustained, performance of your storage subsystem: aim for 240GB capacity or more (with all the controller's channels filled and each channel interleaved optimally too...) - and partition accordingly.

One other thing to consider (if you haven't already): if you value battery life: consider the Crucial M4 or Intel 520 series over the power hungry Samsung 830's (power hungry when being used at anything over 'idle' conditions...).

Good luck.

 

i'm convinced. i'll wait it out for a 256GB SSD, or larger. i looked at the intel 520 series, but the cost is a little more than i care for and i'm leery of the controller intel chose to use with it. i know sandforce is fast, i'm just not convinced of it's reliability - even in intel's hands. what is difficult for me to decide on are the controllers used between the crucial and the samsung. i don't know anything about either of them. i know samsung and crucial both make fantastic RAM, but that's really all i know.

battery life is always a concern, but not enough to make me dismiss the samsung. 90% or more of my time is spent plugged in, so i don't see it being much of a deciding factor.

 

While battery life may not be important plugged in - but heat build up should be (depending on how well your notebook is designed regarding cooling).

 

my 7200rpm hitachi typically operates 40-45C. i understand the tolerance is 55C. do you think this will be a problem with the samsung?

 

For an HDD - that's an optimum temperature (according to Google HDD statistics) for an SSD - I think that is stupidity (in a notebook format, without a cooling fan over the SSD).

See:
Crucial m4 256GB 2.5 Inch / 0009, Samsung Serie 830 256GB 2.5 Inch / CXM03B1Q - SSD Comparison - SSD Reviews, Data Sheets and Comparison - ssdreview.com


Look at the bottom of the linked page: over 12 C hotter running (and I'm pretty sure this isn't in an enclosed notebook's chassis).

Yeah; I think this will be a problem with the Samsung.

(Even if the SSD is running within spec's - it is needlessly (vs. other SSD's) heating up the rest of the notebook's components - Bad, bad Samsung!!! - leading to premature system failure, performance cpu/gpu throttling or other (hot surfaces) undesirable effects).

 

I haven't noticed any heating issues with my SSDs, and I can't say first-hand if Samsung or Crucial drives heat up (though Google says they do). As for the current Intel drives, I've had absolutely no issue with my 330 since buying it. Not sure if Intel actually did anything special with the firmware though since they're not telling anyone what exactly they did, and other non-OCZ SF-2281 drives are getting good ratings on Newegg as well. However, at Intel's pricepoint, I'd just get another (reliable) brand with the SF-2281 (I assume Intel didn't do anything special with the firmware until shown otherwise).

 

Actually they did, some systems that BSOD with SF-2281 drives even after the fix don't BSOD with the Intel drives, Anandtech mentioned it in their 520 review.

Source: http://www.anandtech.com/show/5508/intel-ssd-520-review-cherryville-brings-reliability-to-sandforce/

I've still seen or heard of a few cases of isolated BSODs on the 520, but it still looks like as if Intel did something different.

 

tiller, some ssd's like the Intel 320 are over-provisioned with extra nand so you can fill them completely and be ok.

 

Not if you actually use them (when over 60% filled), like I do.

 

There haven't been any reported cases of Samsung's 830 "overheating" or causing any overheating of any other components. You are simply speculating and have nothing to back up what you're stating. Why don't we have the all 830 users report on temperatures under load to squash your false theory? They have been operating within safe limits and have been 99.9% bulletproof/reliable. Tiller, stop spreading false information already. Just give it up. You are simply fear mongering others into thinking Samsung's SSD is a terrible drive but nothing could be further from the truth. The 830 and the previous 470 have proven to be some of the best SSD on the market. There's a reason why people on this forum, and even across the internet recommend Samsung's drives. You make claims of poor engineering and poor design but have no proof to backup what you claim. Sure energy consumption, and to some extent the load temps according to that site, this round are higher than other SSDs but its only their second generation of consumer SSDs. Things are bound to improve as SSDs replace HDDs anyways. You don't know what could've been the reason(s) behind that temp reading.

So please, just shut up already. Lay off the Samsung hate-juice buddy.

OP, go with the Samsung 830 (or the M4), 256GB, you will not regret it. It is VERY reliable. You and many like you, myself included, spend most of our time plugged in so energy consumption is not a huge issue. Even on battery its a non-issue.

When I'm proven wrong, I will retract my statement. Only time will tell.



EDIT: I apologize for being immature but what you said really grinded my gears.

 

That site measured 5C higher temps with the Samsung which is nothing. I never had any heat problems when I had it. The Samsung is really a very good drive if you don't mind the higher power consumption.

 

Sorry, not 'hating' Samsung. Just hate poor engineering (it should not go rewarded)...

I am not making anything up - I gave the links to back up my claims.

Also; please note that I am not saying that 'idle/normal' use will cause anything to overheat - I am saying that if I buy a bleeding edge component like an SSD; I intend to use it fully (not just let it sit there idle 99% of the time...). And if the storage subsystem is being stressed - you can be sure I'm also stressing the cpu fully too.

Under those conditions; the relatively higher power consumption (vs. other SSD's) can only translate into higher temperatures - period. This is poor design because it doesn't give a correspondingly higher performance in real world use (just heat...).

Now, while I appreciate that you apologized for being immature (apology accepted...) - I would further suggest that you simply ask me to clarify what I'm talking about and engage in a meaningful exchange of information - instead of simply name calling.

I posted at least a year ago that an SSD manufacturer can achieve anything with any SSD design if they were allowed to use as much power as they wanted. Samsung has obviously chosen this route to increase the stability of their SSD's and many people see this as a positive.

I, on the other hand see how Crucial, Intel and Plextor can offer that same kind of stability without resorting to simply increasing the voltage supplied to the nand/controller as a means of cheap insurance.

Sure, the 830 series can give 'best of class' performance for certain metrics - but SSD's were (and still are...) marketed as fast, quiet, low power (vs. HDD's) alternatives for our storage subsystem needs.

Where the 830 fails (falls flat on it's face imo...) is in the 'low power' requirements under the type of use an SSD is actually best suited for (running flat out...) - not only does it fail in comparison to other/current SSD's, but to mechanical HDD's too (a 10K RPM VRaptor runs cooler - granted; with its icepack and native desktop habitat... and the power envelope is very well tamed too (ignoring idle power, of course)).

See:
Temperature And Power Consumption : Western Digital VelociRaptor WD1000DHTZ Review

If we were talking about any other performance metric (r/s r/w speeds, for example), I don't think we would be having this (kind of) conversation - so why are you so sensitive about this very real issue?

Again; in a desktop environment with direct/active cooling on the SSD - I do not consider this an issue. But in a notebook system that is stressed daily (cpu/ram/ssd), this is an issue everyone should be aware of.

When I tested (internally) and bought the Intel 510 Series 250GB SSD I said the same things about it - but the 830 is even worse than that (relatively old) drive and this is simply not the direction I want manufacturers to go in (notebook's, not desktops are the future, after all...).

See: http://www.ssdreview.com/review/com...ies-250gb-25-inch-pwg2/power_consumption.html

Sure, for many, many casual users the 830's limitations will not be felt - but I suspect than many people like myself look for and adjust their purchasing decisions beforehand (which is why there are no reports of 'overheating' 830's in the wild?).

Anyway, I hope I've offered a useful glimpse into my viewpoints?

 

i have only recently (within the last 6 months) become concerned with the heat generated from my HDD in my laptop. i never considered it before. now, i watch it regularly - and i get worried, on occasion, with the temp creeps up close to 50C. since i was made aware of it, the higher temps "issue" (i use the term lightly, for now) of the samsung is something that i will have to consider. i like the performance of the samsung 830 and intel 520, but i still don't completely trust sandforce. i like the price of the cruciual m4. i'm still torn - but i have a little time to decide.

it seems to be well-known that 55C is the upper limit for HDDs. what is the upper limit for SSDs?

side note: i couldn't really care less how people give opinions. as long as i get people's opinions of products. even from "haters", "fanboys", or whatever, the information is still there and it can be seen why someone does or does not like certain products. i appreciate tiller's opinions and ire towards samsung's direction for the 830. and i also appreciate davidricardo86's defense of the product. i get to see 2 opposing viewpoints and that always helps me.

 

Just one more point (can't edit in Win8/IE10 right now!!! ARGG!!):

The best overall performing, most reliable drives for my $$$ are/were the Hitachi's for me. What makes them best performing is:

Quiet (at idle),
Low power consumption (for performance offered),
Cool running (by design),
Performance over time (in Win O/S's),
Low load noise (for performance offered),
and Price.


I feel that Intel is currently the 'Hitachi' of SSD's with it's approach to how it balances and positions it's product offerings.

Every metric is given the appropriate weighting with Intel - something that other manufacturers should strive for if they really want to seem competitive to people like myself (who value 'balance' above all).


A 'balanced' system (balanced; down to/within individual components) is the BEST system to have/use for an extended length of time (years) - being superior in ANY one aspect at the expense of another is not a configuration that tends to stick around here for long.

What I strive to keep 'perfectly' balanced is: PLATFORM/CPU/RAM/OS - yeah SSD's are not in that equation - they're simply 'tweaking' an already powerful setup that is spec'd for the most bang for the buck (productivity-wise) - but that doesn't mean that any downfalls of an SSD won't affect the balance of the system as a whole though (especially in a space/power/cooling limited notebook)...

 

Tiller, I'm trying to see it from your point of view and maybe I was being stubborn but I will admit you do make some valid points. Now you have me wondering about how this might affect the life of my drive and the lives of other people drives. Right now it may be too early to tell, but how will things be like 2-3 years from now or after Samsung's warranty is up. I'm still interested in seeing what max temps other 830 owners are getting. According to Samsung, the 830's operating temperature is 32ᵒF to 140ᵒF (0 degrees C to 60 degrees C). I've been able to get mine up to 42 degrees C. Is there a program I can use to stress test my SSD? I want to try and reproduce that same temp reading from the link you posted. How do these reviewers come up with the energy consumption readings (watts)? Is there a program available for that too?

And thanks for keeping your cool, sorry about that once again.

 

Be careful about stressing an SSD with writes. Just like over-benchmarking on an SSD, things like that will place unnecessary writes against the drive. With a limited number of writes for NAND Flash memory, it is something you don't want to do in and out every day, 12 hours a day, 7 days a week.

What about a full virus scan against your SSD? If you have enough data on there, it would go through reading all the files on disk. If you have a large installation, this may take some amount of time. During the scan, you can use a tool like HWMonitor or HWInfo to monitor the temp of your drive.

You could try some large benchmarking tools, but most of those will be writing things to the drive.

Another thought is to copy over an extremely large amount of data to the SSD, and then copy that out to another drive (or external HDD) over and over again. Trying to see if reading from the 830 over and over again can heat things up.

 

Endurance isn't anything anyone needs to worry about.

 

i dont understand what the above chart is supposed to show. could someone explain it?

 

We're talking about the ability for a NAND Flash cell to write data. On here - - See Flash memory - Wikipedia, the free encyclopedia or AnandTech - The SSD Relapse: Understanding and Choosing the Best SSD . Note, the Anand tech article is dated as the newer flash cells get about 5,000 cycles.

Addressing the post, I don't know about the source of the graph, but in general SSD flash writability is not something to worry about. Before this happens, you'll probably have upgraded or encountered a different problem w/ the SSD or other hardware. However, given that the number is indeed limited, there's no reason to get closer to this threshold just for the mere sake of doing it.

 

If I'm correct on where this chart came from (from the Anvil endurance testing section on the extremesystems.org forums), then this is supposed to show how little we should be caring about how long SSD's will last in 'normal' use.


(Yeah, found it):

See:
SSD Write Endurance 25nm Vs 34nm - Page 203


What I have to point out is that while I don't doubt that these high numbers were achieved, that does not mean that current/tested SSD's are actually usable (real work workloads) at these unrealistic total writes to the drive's nand chips.

By that, I mean that while the amount of data indicated was actually written - that is not the same as saying the drive will perform flawlessly in normal use (turned off, left idle, switched to a new system, etc.) with the same amount of writes to it. More to the point: the SSD may not even be able to be read properly (with no errors) with such high writes - and certainly these weren't tested to see if the SSD's could be read after a year of non-use.

These synthetic benchmarks/scores for endurance only means that the drive can only do those kinds of synthetic workloads. Remove the power (for a few hours/days/weeks/months) and see if the drive is usable for real O/S and user workloads and I would say they'll fail pretty bad.

In fact, why would manufacturers not rate them much, much higher? (And charge appropriately...).

It's because in real workloads they can be made to fail at much less than their rated lifetime TB's written value. And by fail, I don't mean simply die outright - to me 'fail' is also when the SSD performs at less than HDD speeds (for a given workload). And this can be shown easily with almost any SSD out there right now.

Getting attached to one metric or another (in this case: write endurance) without testing/checking for other just as important (or, to me, MORE important) metrics is kind of a waste of time. Sure, it gives us one more data point that we should use in our evaluations of SSD's - but the problem is that some people just see this one point as the end all and be all of their 'evaluation' of a storage subsystem solution (just like some people hang on to 4K R r/w 'scores' too...).


For more information:

See:
Endurance Testing: Write Amplification And Estimated Lifespan : Intel SSD 520 Review: Taking Back The High-End With SandForce

See:
SSD Write Endurance 25nm Vs 34nm

 

I would like to thank everyone for their input. You have not made my decision any easier

Now I just need either the 256gb samsung, the 256gb crucial, or even the 240gb intel to get marked down to 180ish on amazon, or I'll have to wait another few weeks.

Sent from my Nexus One using Tapatalk

 

More likely than not, you'll be picking up the M4. They're usually on sale all the time (since the beginning of summer), so keep a lookout on the Deals subforum for it. Intel may or may not do another sale, though I doubt it since people will bear the extra costs for the Intel name (a little for Samsung as well).

 

For the record, after a year of normal use in a desktop computer, I have a 128gb PNY SSD sitting at 97% usable life remaining, and a 256gb Kingston SSD at 96%. The PNY is paired with a 2tb mechanical "data" drive, and the Kingston is the sole system drive in my wife's desktop. Based on this, both drives are expected to last 20+ years of normal use, and iirc 0% MWI is standardized by JEDEC as being the point where a drive will retain data in a powered-off state for a minimum of 1 year.

For that endurance thread, they tested the X25-V at about 800tb or so by running an MD5 on the data and then leaving it powered off for a month, when they turned it back on and re-ran the MD5, the data was still intact.

 

well, amazon marked the crucial m4 256gb & samsung 830 256gb to $199.99

i went with the samsung (cheaped out on the shipping and went with free super saver - 1 week wait time ). since price was equal, performance was my main deciding factor. all i can do about the temps & battery life are try it and see what i get.

thank you, everyone, for your input.

Update: I've had the Samsung for about a week, now. Wow! What a difference! Boot time is down to ~15sec (from ~2min), iTunes opens instantly (~20-30sec before). It's just amazing! Temps are steadily 30-35C. Battery life is unchanged, so far. I can't speak highly enough about this drive.

Side note: I have discovered that my CPU is now holding my system back. May have to look into upgrading I'm 520m to a 720qm or something...