What makes the Samsung EVOs so cheap?

I was wondering this after seeing Intel release their 730 line and seeing specs not seem as good as the EVO at face value.

I currently have a 240GB Intel 520 that is hurting for space so I'm thinking of upgrading to a 500GB EVO.

But I'm curious as to why EVOs are so cheap GB per $ compared to basically the rest of the entire SSD market.

Thanks!

 

They use TLC nand.

 

They are cheaper because Samsung produce everything in the SSD themselves.
NANDs, Controller, PCB, everything. Hence they have probably hefty margins on the EVO line and can adjust price accordingly

They also use NAND called TLC which is cheaper than MLC which many use.
Don`t believe anyone if they come in here and say TLC is worse than MLC.
TLC will outlast all the components in your computer and can easily withstand 30+ years even with heavy use. So it is a moot point.

I use 500GB EVO now and its the snappiest drive I have ever used.

 

Not entirely true. The Crucial M500 series are cheaper, especially in the mSATA form factor. Both offer great value and performance. The Sammy's are slightly faster, but it's unlikely you'd notice in real world use, while the Crucials offer better bang for the buck (less $/GB).

 

Slightly faster and more power efficient.

 

Spoken like a true unbiased owner While Clodfire rightly points out the fact that Samsung produces ALL the SSD components in house, as part of the cheap price factor, he throws in the untrue statement I highlighted. TLC is worse than MLC period, The consumer may never see this for endurance. Depending on their workflow, multitasking, they may see it in performance. Competition, mainly from the less expensive Crucial, has Samsung EVO so cheap. See the links for some facts.

TLC vs MLC
AnandTech | Understanding TLC NAND

Multitasking - real world difference
http://forum.notebookreview.com/solid-state-drives-ssds-flash-storage/736807-my-1tb-evo-very-slow-compared-my-corsair-performance-pro-256gb-why.html

Idle power consumption doesn't necessarily equal longer battery life. M500 beats EVO here
Crucial M550 1TB SSD Review - Benchmarks - Power Testing & Final Thoughts

 

Exactly. Some people here are the worst around here to recommend drives based on enterprise workloads despite 99% of those who come here and ask want consumer performance drives...
Not gonna mention names though

Its absolutely hilarious that you link to a thread where one guy have issues with his EVO. I`m sure you can`t find any about Intel SSDs or Crucial SSDs. Faulty drives, user errors, hardware issue, it happens to all drives. But I want sink that low that I go searching for these threads

There are many reviews out there that will show that EVO is among the fastest drives out there, even on high Queue Depth, and since its one of the cheapest drives and TLC have life expactancy to outlast pretty much any scenario except some peoples 1% workloads, its a no brainer to buy them.

You can start here:
Samsung 840 EVO 750GB SSD Review - Real world tests | Myce.com

I`m not biased at all. I have owned many drives over the years and I can say out of my own experience that this is the snappiest drive out there.

 

Cloudfire you either have a reading comprehension problem or you just see red and disregard what's written if it differs from your opinion. Your statement disregarded facts and was opinion. Showing just one user review defeats your argument. The owner noticed a performance difference in his workflow. His i ssue was performance! This, likely. happens after the faux SLC layer is full leaving only the TLC. And we know from the 840 (TLC only) about lower performance. His experience wasn't faulty hardware or user error. How is showing a real user experience, in a heavy workflow, sinking to some low level? It wasn't a link about a faulty drive. We all agree that all manufacturers produce some lemons.

Did I say that consumers may not see a difference in endurance? It is a fact that TLC will die before MLC

Again, you link to empty drive tests that are useless to consumers. I've given you better links for your argument.

Disregarding facts is bias, sorry. Come on, out of your experience it's the snappiest drive you've used not out there. That's hilarious, a salesman till the end! Oh, you did mention names You just can't help yourself.

 

lol there we have it again.
"Workflow"

Great you found a user with a workflow where the EVO didn`t cut it. Congratulations. Want me to find a Sandisk Extreme II where it sucks? Nah, wont bother. Its just too silly tracking down threads on a forum.

I have already posted review that shows EVO being the top performer.
There is tons of information about TLC outlasting any scenario for 99% of the users for 50+ years. Doesnt matter that MLC live longer. Its a moot and stupid point.

You should know that its the controller and the firmware that mostly contribute to the performance of the SSD. Not TLC vs MLC. Samsung have tons of experience with TLC. They know how to build drives that perform. Which you can clearly see if you read reviews.

So you can keep your "workflow" to yourself and your 1%. But please stop giving out advice to people looking after a fast drive, where you post links to insane workloads or enterprise scenarios. Its just silly because it doesnt matter to the clear clear majority of people.

 

Don't write this, "Don`t believe anyone if they come in here and say TLC is worse than MLC". in a forum giving advice and expect not to hear differing views. Again, you post that nonsense, then I only need to show one example to defeat your cheerleading. The facts are that TLC is worse than MLC. It may not be a factor in many people's light usage "workflow" for endurance and/or performance. What's wrong with letting the OP know that if his "workflow" is heavy an MLC drive might be a better choice?

I wouldn't mind seeing negative reviews about the Sandisk Extreme II, as it's useful information. But it doesn't belong in this thread.

The Myce.com review is what you refer too? You might want to check that review again. Ignore the empty drive results as they are useless, no matter the "workflow."

Yes, I can, clearly, see that Samsung knows how to build drives that perform in empty drive benchmarks. I can also, clearly, see that their history shows a different story when in a used state or "steady state."

Multitasking; VMs, browsing, downloading, video/photo editing, torrenting is an insane workload? Having more information is less useful to the closed minded. So, no, I won't stop giving advice that may be helpful to some.

 

For 99% of users, TLC is perfectly fine. TLC may die before MLC, but in most cases, we're talking hundreds of TB of writes later which most users will not achieve before replacing their equipment. Even a daily "heavy" user will not exceed 100GB of writes a day. If you do, then you are using the wrong drive. Quoting performance for a worse case scenario doesn't help anyone. It only scares them thinking their drive will die on them.

Check out the write performance over time on a Samsung 840 TLC drive that was constantly between 80-100% full: http://forum.notebookreview.com/sol...0497-samsung-840-120gb-endurance-testing.html

I reached 200TB and stopped. It's proof enough to me that TLC has endurance and reliable enough for most users.

It didn't budge. I'm about to do something similar with an 840 EVO...

 

Bullrun,

Your points are valid and Cloudfire just doesn't want to see it.

Another recent post indicates that OP'ing by only 6% helped keep the SSD fast.

See:
http://forum.notebookreview.com/sol...are-17-what-ssd-should-i-get.html#post9600955


The numbers though, show otherwise.


While TLC may be good enough for 99% of the people (that don't know any better), my point is that very often MLC deals come up where you don't have to drink the Samsung Kool-Aid.

Samsung drives have always seemed laggy/slow to me and it hasn't changed with anything they've come up with yet.

And no; I'm not talking about my (maybe 'too intense') workflows; I'm talking about simply using them navigating the O/S (Win8.1x64 Pro at the moment).


TLC drives are the FWD version of a Porsche - all show and no go.

 

Amongst all the arguing I've seen here, I guess my only question remaining because I don't know how much my workflow is, I can assure it isn't near 100GB. Heck, I average 250GB of throughput on the internet a month and that's where I spend 90% of my time. So can I expect the drive to last me 10 years? The only reason the 840 EVO caught my eye was the price. But I did some reading up after posting here and I understand how MLC > TLC. But this would be the difference I guess in going for an 840 Pro vs 840 EVO. The last thing I'd want to do is upgrade to a 500GB EVO from my 240GB Intel 520 and then all of a sudden start noticing even the slightest performance drop.

 

No arguing: just different methods of presenting what we know/believe to be true.

I think you might notice the drop (does 30MB/s scare you?).

See:
AnandTech | Samsung SSD 840 EVO Review: 120GB, 250GB, 500GB, 750GB & 1TB Models Tested


See the comparison graphs for all the SSD's (the recommended SanDisk Extreme II 480GB is the only 'winner' here). And check out the TRIM checking (it works!) but with the drop to ~30MB/s when hammered (from ~400MB/s when the pseudo SLC is in play). Even then; 150MB/s 'sustained' is not what I would call exhilarating performance either though.


These graphs also show me visually how the Samsung 840 Pro can feel laggy too (consistency; it does not have).

And I do want to point out Anand still recommending 25% OP'ing (I am still recommending 30% or more).

The 240GB Intel 520 Series (along with the same capacity SanDisk Extreme I's) were in their grove as O/S drives - the 480GB SanDisk Extreme II's are the only superior replacements at this time.




The EVO's aren't cheap (enough) for me - they are extraordinarily expensive.

 

How exactly do you overprovision? Is it when installing the OS you just only format at lesser than drive size
and don't touch the rest? Does it really improve performance that much to overprovision?

You guys basically have me no longer interested in the EVO. I've now been looking at the Extreme II, 840 Pro, and M550. But with the 480GB Extreme II available for $300 on Amazon and the comparison tests on Anandtech, I'm strongly leaning toward that assuming the price remains there when I am able to purchase it within the next few weeks.

 

By all means, don`t read the links me and HTWingNut posted.

Why even bother...

 

Because you clearly don't know what real work is.

 

This is exactly what posts by tiller and bullrun ends up doing. It forces users away from perfectly good drives that really have no real world issues or "performance issues" that most people won't even notice, and end up spending more than they really need to. There's nothing wrong with wanting the absolute best, but in some cases there is no valid justification. Sure I want the best acceleration in my car, but if I can save some money and lose a little bit of acceleration with a different car/transmission/engine, it's not the end of the world. Tiller's points are valid for heavy WRITE workloads, or if you tend to fill your drive to its peak frequently. But aside from that, it really is a sky is falling mentality.

In any case you can overprovision easily by just shrinking your main partition by how much you want to OP using disk manager. Create a new partition in that space, then delete the partition. Simple. I ended up OP'ing my drives 10%. Mainly because that's what Samsung recommends, and enterprise drives are even OP'ed only 7% by the factory. Buy what's best for you, not for someone else's habits or workload.

But I give up. I'm not really posting here to convince anyone, only share my experiences. People can make their own decisions, but I'd look at more than just this forum before investing hundreds of dollars in a storage drive.

 

By all means, read Cloudfire's linked review. It's a good review. Just ignore the "real world" tests he always points to, as they are done on an empty drive. No consumer, that I know, would buy an SSD to leave empty and just run benchmarks. If that's your goal buy an 840 Pro. Give more weight to their "sustainable performance test"

Definitely, read HTWingNut's link. He did great work. His review along with Hardware.info should alleviate fears about TLC endurance in average to light usage. OP, based on your usage scenario, EVO endurance is not a factor.

Hardware.Info tests lifespan of Samsung SSD 840 250GB TLC SSD [Updated with final conclusion] | Hardware.Info United States

 

They give horrible advice. They end up confusing more than helping. And they keep on grinding on the same BS arguments that is only valid for a tiny percentage of people.
If this was a forum for server/enterprice clients I would agree with some of their arguments, but this is just sad to see them jumping on all threads and assuming everyone does this.
Their arguments make little sense for people coming to a notebook forum. TLC vs MLC, provisioning based on queue depths of 32 (seriously...), heavy workloads etc, its the same crap posted every single time anyone wants advice on a SSD buy.

I wish Phil the moderator would return. He was very active here and was good at giving sound advice. Now its just the same boxed in arguments that keeps posted in every thread.
Whatever. Its a semi dead subforum anyway.

Im out

 

I'm sorry. I'm just really confused. I thought you guys were against the Evo. I'm not the most advanced in storage terms. I must have misinterpreted what I've read... a lot. I did read through every link provided by every poster. I did not skip over any posts, I've spent a lot of hours reading over SSDs today.

Simply put, I'm hurting for storage space on my laptop. I have an mSATA slot but I refuse to use it due to it being SATA II. I'm not ditching the blu-ray reader because I do use it frequently. That leaves me with one storage slot. 240GB just isn't cutting. Especially since I keep lossless audio archives, have been downloading youtube videos from my watchlist so I can watch offline, and of course video games.

I don't want to end up buying a drive that performs lesser than the Intel 520. But if the Evo better performs it and you guys say I have nothing to worry about for the entire life of my computer, then yea, I'd rather save the money and buy that instead. If the Extreme II is a better performer than the Evo, then I currently find that tempting since it is only $50 more on Amazon.

 

Sorry, HTWingNut, I don't really get how my words are misunderstood. "The consumer may never see this for endurance. Depending on their workflow, multitasking, they may see it in performance." "Did I say that consumers may not see a difference in endurance?" "The facts are that TLC is worse than MLC. It may not be a factor in many people's light usage "workflow" for endurance and/or performance. What's wrong with letting the OP know that if his "workflow" is heavy an MLC drive might be a better choice?" Do I need to bold the words; may, may not, might, might not, if, depends on ? The use of the word "worse" was not my choice. It sounds, well, worse.

No where do I say TLC is going to die, don't buy it in any "workflow". Do you read what I wrote as that? I do say, factually, that it will die before MLC

What's wrong with saying if your usage is heavy or mutitasking, an MLC drive might be the better choice without already knowing the "workflow"?

 

Consumer SSD's - I'd say buy the the one that's the least $/GB, because pretty much all modern SSDs are going to perform really well and not be a limiting factor. Endurance only matters for the heavy users, and I think if you have to ask yourself if you're a heavy user then you're probably not one! I remember working out that my little SSD will last about 75 yrs with my usage based on the 'Percentage Lifetime Used' shown in HWInfo64.

(Maybe get the Crucial M500, if you've found that to be the cheapest per GB, think I read that earlier in this thread).

 

Most sound advice given... /end thread/

 

Sound advice indeed, though I'd add a "but...": Personally, given their track record, I'd stay away from anything with a SandForce controller. Even though the SF-2281 (and likely later) controllers are now decently reliable (the first-gen sucked in this regard), I still don't really trust them. In fact, I'm actually surprised my Intel 330 is still going strong after about a year of use, though I learned from that that the advertised performance is a half-lie with these controllers, since they're dramatically slower (sequential read/write wise) with uncompressable data than compressable.

Also, there's another, much newer controller called LAMD that I can't seem to find good data on concerning long-term reliability. So I'd avoid SSDs with that controller until at least the second generation of those controllers.

 

My thoughts exactly...SF has used its share of coupons with me for this lifetime...

As for EVO vs. M500 I'd vote for the latter...although neither would be my preferred choice for a machine of my own.

My $0.02 only...

 

Well, good to see we can all be mature here.


Facts:

MLC superior to TLC because:

1) Faster programming (erase) times. Less stress on nand.

2) Better processor chips/firmware with no tricks like pseudo SLC nand which only serves fool consumers into thinking TLC is as fast as MLC when in fact it is only slightly faster (talking sequentials here...) than the best HDD right now (~150MB/s).

3) Doesn't matter if the MLC drive is $50 to $100 more expensive than a TLC drive over the term of ownership (I'm assuming more than 18 months and closer to 4 years average).


My workloads were specifically left out of this discussion (I thought...) - I see differences in simply navigating the O/S... like I already mentioned.


Finally when the price of a 'real' MLC SSD is close to (or even less) than a TLC drive; the TLC's have no redeeming qualities at all then. And my point is that with the frequent sales that happen lately, it is better to wait for a week or two for one of those sales than to buy a TLC drive 'right now'.


Technologically, TLC is interesting in a strict 'classroom' setting. Lots of thinking, theory, tricks and innovations that deserve to be recognized. I'll give it that.


But to spend real money on inferior real world products when much better options are available? That is not what I come here to recommend.

Even if it's somebody else's money and not my own.

 

Well thats kind of the impression I get from the war between Samsung vs everybody else.

NAND simply doesnt scale well with Moore's Law.
The question has always been TLC at a mature process node or MLC at the cutting edge process node? Furthermore, is it better to go TLC for higher density or MLC with huge pages (aka bigger die area)?
Performance, capacity, endurance and cost will always be engineering tradeoffs.

Simply, TLC costs less because it gives significant advantages during manufacturing (higher density, smaller die area) provided it is manufactured on a mature process node (TLC actually costs a lot more to make than MLC on cutting edge nodes due to poor yields), obviously the performance tradeoffs include lower endurance and high latencies. In a consumer product, TLC is thus advantageous for Samsung as 90% of consumers use their SSDs in burst workloads with lots of long term data retention. As I understand, the EVO uses a portion of the NAND die which is programmed as SLC to act as a write buffer. The buffer pretty much equalizes the TLC's slow write disadvantage vs MLC drives in consumer workloads while the long term data retention habit of most consumers renders the high endurance MLC NAND moot. Thus, in theory, the average person can enjoy an SSD with performance which matches their usage style with good capacity at a lower cost.

The current approach with large page MLC drives on cutting edge nodes (mostly championed by Micron) is the alternative to TLC. While large pages give the MLC similar densities as TLC but the larger die sizes must be offset by the newer process node to be price competitive (I.e. MLC lithography must always be full or at least a half node ahead of the TLC method). Combined with generally better yields on MLC vs TLC (NAND is usually binned according to endurance) this approach produces cost effective, high density MLC NAND which is slightly more expensive than Samsung TLC. However, due to the cutting edge nature of the lithography, it is difficult to immediately match the volume of usable die output of the TLC approach. While large page MLC retains the traditional advantages of higher endurance, the performance still suffers somewhat (though reprogram latency is still superior to TLC). From a consumer point of view, the real advantage of this kind of SSD is better endurance and slightly better uncached write performance. I daresay however, this approach is a harder sell than the 840 EVO simply because it doesn't align with the majority of consumer workloads as well (due to the lack of SLC cache). However, for the prosumer, enthusiast and professional, the endurance and sustained performance is a much more attractive proposition.

In conclusion, the 840EVO is cheaper due to lower manufacturing costs, don't forget, Samsung also controls the whole product stack (and Samsung's fab output is.....formidable) thus economics of scale also come in to effect. The drive is also more popular due to superb availability and the fact its performance envelope aligns well with consumer workloads (the Cached performance simply eclipses the regular peak performance of Large page MLC) with the disadvantages being mostly moot.

 

Thank you Marksman30k. Very well thought out and well written response. I couldn't agree more... Consumers want best bang for buck. Most don't even know or care what an SSD is. All they know first time using it is how much faster the SSD is than their traditional spinner especially in a notebook. They also can't appreciate spending half the cost of their notebook just for their storage drive either for the larger capacities despite the apparent advantages. The enthusiast, hobbyist, and professional will always look for the best performance even if cost is substantially more. But for the Joe Layman user, they want cheapest, period.

 

Marksman30k and HTWingNut,

I agree 100% with both of your statements.

My point? What is wrong on educating these unwashed masses (we're not the only ones reading what we write)?

As pointed out, most consumers don't know what an SSD is nor what workload they are using. Blindly recommending to them/everyone that a TLC SSD is a good match because they don't know much about storage subsystems (HDD or otherwise) or their workflows is not a good idea in my books.

I am more of the opinion that you can never have too much facts. Give that and let the person make their decision. Don't coddle them and keep them in the dark unnecessarily. If I present something that is not understood - I expect a question to be asked by the OP and further conversation can ensue. That is how information is spread.



When clients I have presented all the facts to decide on the 'wrong' action, I don't try to persuade them otherwise. But I do present them with the facts of why they want to consider another course of action (and some of them even do think of better (for them) options than what I originally suggested). Those that are just not paying attention to the facts presented (and are only seeing lower $$ signs in their head) eventually get bit in the behind by their decisions. The next time, they usually listen closer to what they're paying me for.

On forums though, people can jump in and offer opinions (and I do too, no doubt) which dilutes the real information the OP should be paying attention to.


My final recommendation to the OP? Always buy the most advanced/proven hardware you can afford without going into debt while considering the cost over the expected term of ownership.

You'll never go wrong paying a little more even if you don't end up using all the benefits offered. The part will still be useful to someone else down the road if you decide to sell it then, or you may repurpose it and continue using it yourself. Either way; you'll have enjoyed the performance from the first day of ownership and everyday it is used thereafter.

An inferior part (especially for close to the same cost) may or may not give you the same warm fuzzy feelings about it once it's time to replace it; it depends on how your workflow changes over the term of ownership. And change is the only constant that is guaranteed.

 

@ Tilleroftheearth - I respect your opinion and know you have vast experience in the matter, and push storage drives to their knees with whatever it is you do. However, your posts come across that if users buy the EVO (as a specific example) that they will fail at any moment and will have intolerable performance issues. This only instills uneasiness and insecurity that doesn't realistically exist.

I can appreciate buying the best you can afford, but not every user needs this. Upgrading from a laptop hard drive to *any* modern SSD is like improving the acceleration of your minivan to that of a fighter jet. Choosing which SSD is like choosing which fighter jet.

 

Oh so true. Just my anecdotal experience, but I can't personally tell the difference between my Intel 320 and 330 most of the time, and even with large file transfers I'm still limited to HDD or USB 2.0/3.0 speeds. I also can't tell the difference with the Samsung 840 I installed in my sister's laptop, and as far as she knows (and cares), it's a freakin' rocket ship, speed-wise. The only concern she had was the storage capacity, but 250GB was plenty for her wants/needs. And even with the pessimistic seven year lifespan, the TLC NAND will last much longer than either of us expect to keep her Elitebook, not to mention MLC's century and a half lifespan (roughly).

 

My dad has a 5 year old C2D notebook that he turns on maybe three times per week and uses about 40 GB of space, OS included. He uses the web, Picasa and Office.

If tiller had his way, he would be running a 1 TB SLC SSD, over provisioned to 75% and 16 GB of RAM, rather than the 128 GB TLC SSD and 4 GB RAM he has now.


Sent from my iPhone using Tapatalk

 

Exaggerate much?


Sent from my iPhone using Tapatalk

 

Not by much...

 

Ok, after reading over everything newly posted and the unfortunate arguing I never meant to set forth, I appreciate everyone's input. What I've taken away from this, best bang for my buck and performance, I should get the Evo and right now because of it's price, it's definitely the top of my list. If I have the money to spend on a different drive and decide to, that's my prerogative but unless the price is close to the Evo, not likely. Seems like the chance is over 90% that I won't even notice a performance difference between my current Intel 520 and whatever new drive I get. I don't know my workload, and because of that, it's probably fairly safe to assume I'm not a heavy user in that aspect. Over the course of time, based mainly on how I watch my SSD fill and delete data under "This PC", I probably write on average 5GB a day. I have my days where it is less than 1GB because all I do is browse the web and download new course notes/lectures and then the contrast where it is 30GB+ but I think it easily averages out to about 5GB a day. Which interestingly enough my internet data usage is on average 250GB a month (I don't use Netflix but do use YouTube) which assuming data throughput on the internet relates to writes (which I know is not remotely accurate) then chances are my 5GB a day range is correct. And based on a thread linked from this thread, that would mean (assuming on the low side for TLC), it should last me over 32 years. Yea, I think I don't have anything to worry about.

Again thank you everyone!

 

Seems like a good choice to me!

 

^ A million times this.


Sent from my iPhone using Tapatalk

 

Amen!

/close thread/

 

...... Moved........

 

Further, it seems the general public is happy with the 840 EVO. A combined total of 716 4- and 5-star reviews versus 48 1- and 2-star scores at Newegg. On Amazon, the ratio is 1,541 to 43.

 

Playing devil's advocate here, most of those are "new" drives without any true long term usage. But everything I've read make it appear to be a good choice.

 

If an SSD is going to fail, the odds are much greater that the cause will be a manufacturing defect or design flaw (a la Sandforce) rather than wearing out the NAND. We're 7 months, give or take, into its lifecycle and there are, to my knowledge, no reports of widespread failures. Yes, that's not "long term" but I highly doubt we're going to see the masses turn on Samsung.

 

Not only would the above be true; if I could have my way, I would pay for it myself for your dad to use like that.


Here is a perfect example of why I recommend the way I do:

Samsung 840 EVO 500GB SSD on sale @ $70 off for $349.99;

See:
Samsung 840 EVO Series Solid State Drive, 500GB at Memory Express


SanDisk Extreme II 480GB SSD on sale at $210 off for $289.99;

See:
Sandisk Extreme II Solid State Drive, 480GB at Memory Express


I hope now everyone can see my point? I'm not recommending to spend more to get more - I'm saying spend less to get more, better and faster. Even if it does cost you less than 20GB of capacity (after OP'ing and formatting).

I don't see how anyone can argue that.

 

Amazon has the Evo 500GB for 251.99 right now.

 

The only point I see is that you're not a very good comparison shopper.

The 500 GB 840 Evo is also $259.99 at Newegg, $279.99 at TigerDirect, $299.99 at Fry's (if you have a store near you and don't want to wait for shipping).

I hope you're not this free wheeling with your clients' money.

 

These SSD discussions are second only to windows 8 threads in terms of fun reading.... Man I am a sick puppy.

 

Maybe the take-away from all this is:
If you buy a high performance notebook that its intended use is high performance then don't short yourself with a cheap SSD.
If you buy a cheap notebook that is going to be treated with general use then get that cheaper SSD.

If you need a good static data repository SSD then a cheaper one will work fine, however, backing up any SSD to a HHD is still prudent.

 

This!

The system I'm writing this on now , cheap, 4GB RAM, single core, has a value SSD (Intel 335) in it. It's really only for surfing, light downloading. No need for ponies under the hood when the suspension, transmission and rear can't support it.

 

Even the $30 savings over the life of the system (~4 years) is peanuts. Worse comparison; if sustained performance is (ever) required (yeah; that 30MB/s 'sustained' for the EVO is not what I want to explain to a client if/when they decide to reuse/repurpose the system in question) during it's lifecycle.

Most of my clients don't hire my services for to simply get them good enough, for a savings of dinner at McDonalds for a teenager or two.

They expect their systems to run like mine (in responsiveness at least) and trust me to spend their money wisely for the expected lifecycle of the system in question. That doesn't happen by being penny rich and pound foolish.