Best/Cheapest 256GB SSD?

What's the best 256GB SSD I can get for under $500 (preferably around $450)? Also, how much will SSD prices drop by August? Thanks!

 

Depends on what you do. If you want the most reliable, go with Intel. If you want the highest performance, consider a SandForce drive.

The market will change with G3 and newest generation SF drives flood the market.

 

One of the best deals I see

Newegg.com - Crucial RealSSD C300 CTFDDAC256MAG-1G1 2.5" 256GB SATA III MLC Internal Solid State Drive (SSD)

 

Got it, and love it. The drive boots this laptop in about 13 seconds. It may depend on your controller. If you have a SATA 3Gbps, the RealSSD ranks right up there with the rest of them. Since I don't have a SATA 6GGbps controller I couldn't tell you what to expect. Other benches have the drive lagging other of the 2nd Gen SSD drives.

In any case, here is a bench I ran on my drive from a while back -


Also note, I use a Momentus XT for data storage. While 256GB is large, I have a lot of need for 10GB+ virtual machines. Didn't want to waste SSD space on those files.

HTH

 

I got a 256GB Micron C400 from superbiiz for $455. The price and stock may vary now though. They were the only ones who had the C400.

 

Thanks everyone! If it makes any difference, would certain SSDs be better if my main focus is gaming load times and boot-up speed? I know that an SSD doesn't increase FPS, but having a level load in 10 seconds rather than 30 is VERY appealing.

 

I agree but keep in mind that SSD and gaming really do not go together really well.

Reasons:

First games are HUGE often 10GB+ if you like to have a lot of games available to play your SSD can run out of room fast. This means most small capacity SSD are out of question, or you have to pay big money for a high capacity one.

Second, load times may be faster but how much faster? Usually not too much honestly. I load into games just as fast as my friends with SSD often times and I think it has a lot to do with video card and other small factors.

But the real reason it does not matter is because in the case of multiplayer you cant start the game until everybody is loaded or a certain time limit is reached anyways. So if your the first in all that means is more time looking at the gamer count down or the other people loading instead of you waiting for the loading.

If its a single player game, most do not take very long to load and a lot of games even load in the background while you play to eliminate almost all loading times.

I just got my first SSD 160GB Intel and currently I have over 400GB of games on my C: so I am going to have to tone it down some

Starcraft 2 and Portal 2 get top priority.

 

C300 performs well on SATA II and SATA III interface. With the introduction of C400, I expect prices for C300 to drop.

 

Yeah, the space issue does worry me. Considering the cost, I'm still not sure whether I'll get a 256GB SSD (planning to get it with the new m11x R3 when it launches). I did put together a list of all the games I want on my PC, and with the OS, I have about 210GB used. Not quite filled up, but there's not as much space left as I'd like. I'll just have to see how good prices are in August...if that Intel 320 300GB SSD is less than $450 by August I may get one.
Will load times - with the exception of multiplayer, which I don't do a lot of - actually be 3x less on an SSD? Obviously, sometimes games do load in the background, but look at Mass Effect 2 - it's literally a minute just to load up another part of the ship.

 

yeah you do not want to fill a SSD it starts to kill off its performance.

Id personally try to keep 20% free at all times.

As for game load times... I dunno. SSD is faster than HDD mostly because of random seek time not the raw read/write speed, though newer SSDs are getting larger and better numbers than the first ones so they are overtaking HDD even in read/write.

Games if your hdd is defragmented should be pretty linear when loading so the HDD can keep up with the SSD pretty well.

Like I said I think this loading process is much more system dependent as a whole than just the storage drive. Look at BFBC2 AMD released drivers for the GPU that cut loading time in half. This is the GPU and its drivers it has no bearing on the drive at all, and I know I have loaded into maps faster than HTWingnut who at the time was using an Intel SSD.

Not all games load in the same way, others may very well gain a good boost from a SSD, I just wanted to lay claim that it may not be as much of a boost as your expecting and that if game load times are super important to you make sure you have a good amount of system ram, a fast cpu, and a good video card to match the SSD.

 

I tried filling Sandforce drives to 95% full, I didn't see a significant performance drop.

On what SSDs did you notice a performance drop?

 

None that I have tested personally I just remember reading its bad for a SSD to keep it near full as it cant quite function optimally if its under heavy use.

Perhaps not so much the performance but the life of the product suffers. Im sure its more proven on the older SSD's where as new ones may have tried to correct for these shortcomings.

If I recall SSD' benchmarks used to be run with the disk fresh and one after it was made full just so you can see how a SSD will perform in its used state, but GC and Trm are now a reality when they were more of a dream back then.

 

I've seen quite huge load speed increases in MMOs..Particularly World of Warcraft which is my main game..So I'm guessing the increased or rather decreased load time the SSD would provide depends heavily on the type of game..IMO of course, I don't have any personal experience on this, just a video on youtube of a WoW player's load times on HDD and when he switched to SSD.. Oh and take a look at the Intel 320 300GB, IDK how much it costs though..

 

The less free space, the harder the controller would struggle to find whole erasable block when there are new write request, it would impact both endurance and performance(increase in WA factor). This is how the whole SSD thing works, no brand can escape that. In fact, the original SF used 17% OP to compensate its compression which would leave more holes. It was later reduced because it bump up their $/G price.

If the usage is read heavy, it doesn't matter though and even 100% is not going to hurt read operation.

 

Isn't it better to leave unallocated space when you partition it than to allocate all of it but not fill it with data?

 

It makes no difference to the SSD which only sees files.

 

With Intel on a new drive if you partition it leaving free space at the end of the partition, it's supposed to add that to the overprovisioning pool. This supposedly helps maintain peak performance, reduces overall wear, and increases the lifetime of the drive. IMHO, for a typical user, this is unnecessary, although sacrificing 3-5GB of space is a personal choice.

Other SSD's have not made such claims, and think it's unique to Intel. Also you can apparently do this after a secure erase, but you don't get the same returns as you would if you did it with a factory fresh drive.

Previous Samsung drives would suffer from this, as would original OCZ drives. I had a 256GB Samsung previous gen drive for a little while and did some testing and not only did performance decrease greatly, it never corrected itself even after a secure erase. 4k read/write was worse than a 7200RPM HDD. Suffice it to say I did manage to get my money back.

I know Intel does not suffer this from what I've seen so far, nor any of the other current gen products.

Also check out this great write-up

http://forum.notebookreview.com/sol...how-prevent-ssd-degradation-without-trim.html

But I think a lot of this stemmed from lack of trim.

A full drive can lead to excessive wear of the cells not occupied since even regular wear leveling can't maintain a 90% full drive by manipulating the data efficiently.

 

Source: http://www.tomshardware.com/reviews/intel-ssd-320-crucial-m4-realssd-c400,2908-4.html

 

I have my doubt on the figures. Read speed should not be affected, GC or no GC. What I am seeing is the speed go all over the place and a read speed lower than write speed is simply unbelievable.

 

And 4K speeds are more interesting than seq. speeds in my opinion.

Please mention the source out of courtesy for the website and our readers.

 

Believe what you like. I found it here if you are interested in reading it
Cost Of More Space, m4's Over-Provisioning : Crucial m4 And Intel SSD 320: The Other SSD Competitors

It is funny how you people always downplay data saying otherwise about SSDs than what you expect. I see it all the time here in this forum. Shure 4K is important, but this figure just show how much better the 510 is when filled up.

 

I just bought a Mushkin Callisto Deluxe 240GB for 499$ on eBay. I hope it's a good deal !

 

I am not downplaying anything. I just cannot find a good explanation of why the read is affected. I am well aware of different firmware has different strength and weakness(why I said Intel has two different line now cater for completely different usage, a practice that is used in the linux file system analogy, there isn't one that is the best but 'best for certain thing'). But a huge different in read is very difficult to explain.

 

Looking at everything, I think I'm gonna get a 750GB HDD with my laptop and see if I really need to spend money on an SSD. If I want one, I'll probably just wait till some good Black Friday deals pop up.

 

SSDs IMO are for people who don't need the extra space for storage and want to have a notebook/desktop that would perform faster than any regular HDD could provide..But I honestly do not think it's wise to grab an SSD(or multiple SSDs) in order to have equal storage space as an HDD unless you're just really rich or have a ton of cash and don't know where else to spend it..Cause then I'd recommend grabbing a WD Scorpio Black 750GB instead..

 

Saw the C300 256GB on Ebay new from what looked like a trusted seller for $400 shipped, so same price ratio as the 128GB for $200 newegg had just 2x capacity and 2x price.

 

Isn't EBAY trusted seller an oxymoron kinda like Military Intelligence?

I can not imagine a vendor I trust less.

Perry

 

Hmmm considering that I have made hundreds of ebay purchases and I was in the Marines I think your comment was almost directly an insult.

 

Do not insult the TechnoViking Marine!

My dad owns a house in Scottsdale, AZ. You can camp out there and stalk this guy...

 

OMG I'm gonna be a lot more careful when talking to Viking now..LOL

 

We are not downplaying anything. Why don't you foot the bill and buy one yourself to test. My tests show completely different sustained times after my drive is filled up, even faster than rated.

4K is not just important. IT IS the most important aspect of an SSD. Period. You use sustained speeds less than 1% of the time. If you actually know anything about SSD's, you would post real world like synthetic benchmarks and not just a single picture from one reviewer's possibly flawed test.

 

4K R r/w's are not the most important aspect of an SSD used in normal scenarios. Period.

The 3x and greater (4K) performance of the V3 over the Intel 510 only comes out to 4 seconds faster (over 870 seconds) than the Intel with the 'incredibly poor 4K R r/w's' in actual realworld tests.

There is no such thing as 'real world like synthetic benchmarks' to show this - all/most favour the SSD based systems simply because the 'numbers' are so much better for them.

No/little correlation to real world at all.

Unless your 'realworld' is running benchmarks continuously (or a server).

 

Then what metrics should we use for SSD purchases, in your opinion? Anecdotally, I'll just say that 4K random read seems to be a pretty accurate measurement of performance for my real life.

 

The 'metrics' that I suggest everyone use is this:

Have a stable, fully functioning system with all HDD maintenance used and applied and note common times (by stopwatch) for your most common tasks.

Install an SSD (make sure this is your only change...) with the exact same installation and usage pattern and simply use it for a week/month, etc. and then time your most common tasks again.

If the few seconds saved are worth the huge capacity loss and much higher price tag to you - then keep the SSD. Otherwise, return it (you did purchase from a store with a 30 day return policy, right).

See - no benchmarks needed - except the only important one: how the new part affects your interaction with the (otherwise identical) system.

How do you do an 'exact' same install without the drawbacks of cloning?

This may give you a hint:

See:
http://forum.notebookreview.com/win...e-partition-strategy-better-than-cloning.html

 

Flawed test? Yes you are finding excuses. What about the review from Anandtech? They found the exact same as Tomshardware review. Or maybe they are wrong too? You should listen to tilleroftheearth. He have a lot of knowledge in this field.

Let me quote what they said
"What does all of this mean? It means that it's physically possible for the m4, if hammered with a particularly gruesome workload (or a mostly naughty workload for a longer period of time), to end up in a pretty poor performance state. I had the same complaint about the C300 if you'll remember from last year. If you're running an OS without TRIM support, then the m4 is a definite pass. Even with TRIM enabled and a sufficiently random workload, you'll want to skip the m4 as well."

Here is the review if you want to take a look
The Crucial m4 (Micron C400) SSD Review - AnandTech :: Your Source for Hardware Analysis and News
Just face it. Intel is better than Crucial with garbage collection and TRIM.

And FYI, i would much rather believe an independent reviewer that reviews SSDs objectively than a random user on a notebook forum that owns the SSD he is talking about

 

This particular page actually contains some very informative thing about SSD contollers. Strongly recommended to read through it rather than just look at the graphs.

 

The Techreport review doesn't show differences between a new and a heavily tortured C400.
Crucial's m4 solid-state drive - The Tech Report - Page 5

 

Anand and this site is measuring different 'torture'.

Anand's one is more close to the methodology used by enterprise(random small write, full LBA span) targetting the usage pattern of a typical database or shared file system type of thing. BTW, the 8G LBA span can be very misleading(marketing want to use this number). If you read Intel's addendum about their drive, you see a HUGE drop in IOPS comparing full LBA vs 8G span.

this file copy test are more sequential in nature(which is much less likely to leave holes over time thus less impact on performance).

Again, not which one is more 'real world', just that one has to interpret then applies it to each individual situation.

 

Techreport uses "a half a day's worth of IOMeter thrashing" to get the drive in the tortured state.

 

Without knowing the parameters used in IOMeter, it is difficult to gauge what it means. IOMeter is a very useful tools for all sorts of access pattern but means nothing by just mentioning it. Anand at least described it in more detail so we know what his pattern was.

 

I think they have described it a couple of times in earlier reviews. Iirc they use the workstation access pattern with 256 concurrent requests.

 

Haven't seen that so can't comment.

If we assume both are doing legit test, I can only say that it is how they dirty the drive that makes the difference. IOW, techreport doesn't contradict anand nor vice versa. Unless you have a very good explanation that indicates Anand's test is flawed, which I would be interested to hear.

 

I didn't say they contradict or that their tests are flawed.

I said that the Techreport review doesn't show performance differences between a new and a heavily tortured C400. That's all.

Personally I wouldn't worry about the C400's used performance. If 12 hours of IOmeter doesn't cause any problems, my usage will certainly never cause problems.