Best SSD in Terms of Price vs. Performance

Hey all,

I'm getting either a Macbook Pro this summer or an Envy 14. Regardless of which laptop I pick, I will most likely be getting a sub-$200 SSD to go along with it as a boot drive.

Currently, I am stuck between the Intel X-25M 80GB and the OCZ Agility 2/Vertex 2 50/60 GB.

I was wondering which drive would bring the best performance/price balance. The X-25M seems to be a very good drive, but all of the reviews show the two OCZs to be faster in every area. Which do you feel would be best to go into a system that has a 500GB 7.2k HDD for storage?

 

I don't know much about the OCZ agility/vertex drives, but in my opinion the more import info you need to be looking at is the random/read right speeds, because that's what's going to make your OS snappy. The X25-M is highly regarded to my knowledge for fast random read/write speeds. I know on paper, sustained write speeds on the X25-M look slow in comparison, but unless you are going to be moving large files around often it's not going to make much of a difference. I went with the X25-M 80GB G2, and I don't regret it one bit, I highly recommend it! Just my two cents...

 

The Sandforce based drives have better random R/W speeds but they cost more than the Intels.

 

MyDigitalDiscount has the 60GB Vertex 2 E for 165$ after rebate...

In terms of price/GB this makes the Vertex 2 and Intel drive equal. The Vertex 2 in faster in every way so in terms of price/performance that is the way to go. Also note the second generation Intel controller is proven and the SandForce 1200 controller used by the Vertex is new technology. To some people this may be a negative.

With that being said I've just recently ordered the 60GB Vertex 2 E instead of the 80GB Intel drive after reading reviews and forums for the past week or so! I hope it works out for me!

 

Thanks for letting me know about that deal. I'll look there in the future for stuff like that.

 

X-25m Imho.

 

Same here 10char.

 

Why though?

 

Brainless regurgitation?

 

The Intel drives are nice drives but the Vertex 2 is faster in every possible way. If you can get them for the same $/GB then the Vertex 2 is obviously also the best $/performance as well...

Everyone will recommend the Intel drive over the Vertex 2, but won't give any reason. (I've noticed this on other forums as well!) Really the only negative to the Vertex 2 is again the fact that the SandForce1200 controller is new technology. Also, because of the way the SandForce1200 controller handles writes to the drive differently via compression it should last longer than the Intel drive. I believe the MLC drives losing their ability to write data after ~ 10,000 writes to each individual cell sounds like it won't be much of a problem for the common user though. Whether or not you believe this may be of concern to you is really an opinion you have to form after doing some research though!

If you (The OP) haven't checked out the OCZ forums regarding the SandForce1200 controlled SSDs I'd say that would be a good place to start as far as educating yourself about possible problems.

http://www.ocztechnologyforum.com/f...Agility-2-and-all-Sandforce-Controller-Drives

I myself was going to get the 80GB Intel X-25M drive but figured I'd save about 50$ right now and then upgrade to a larger sized drive Q4/2010 or Q1/2011 once Intel puts out retail drives using their 25nm memory chips.

 

Yeah, I think that people just promote whatever they have. It doesn't bother me. The Intel drive is great, but I've come to the conclusion that the Vertex 2 is just a faster drive.

I've looked at the anandtech review for both the Agility 2 and the Vertex 2 to gain info on the SF controller. It seems to perform better than anything else in its price category right now.

 

anything faster than intel x25m g2 is no brainer

its like driving a formula 1 on a 40mph street

you got lots of horsepower that will never be used

intel x25m g2 is FAST ENOUGH so going faster wont do any difference.

i would choose intel x25m g2 80gb over OCZ Agility 2/Vetex 2 50/60gb anyday

MORE GB so more software/photos/movies/games/etc
PROVEN RELIABILTY, you dont have to mess with firmware updates
FAST ENOUGH FOR EVERYTHING no more loading times/ blazing fassssssst

 

That ^ 10char

 

I'm not necessarily worried about that. Sandforce is the future. The Intel drives have been out for quite a while by now.

 

Exactly. That's why I spent 170$ on the Vertex 2 right now and might upgrade in 6-9 months when the new Intel drives come out with the 25nm flash memory as well as the new generation controller as opposed to spending 230$ right now on the Intel when I know they're upgrading the controller/storage size in around half a year.

Also, there's literally no sense in not buying the OCZ drive because the Intel drive is already "fast enough". In the same sense Intel may as well not increase the 70MB/s write speed that's currently the weak link of the G2 drives when they come out with their new drives. Something tells me they will be though.

Still though I have to stress that the 50-60 dollar difference between the two drives is what was my deciding factor. There is the 20GB difference in drive space but I am currently using < 40GB on my laptop so I assume the 60GB drive will be plenty for me!

 

stop living in the future

jk LOL

 

since when intel x25m g2's write performance become insufficient

afaik 70MB/s write is fast enough for everything including saving RAW DATA"UNCOMPRESSED" in realtime(FRAPS,etc)

 

If I get an MBP, I'm going to use a 60GB Vertex 2 as a boot drive. I'll put the HDD in the ODD bay and keep that as extra storage.

I'm definitely going to keep an eye out for what Intel does with their next line of SSDs. Hopefully they bring the price down some more, if that is even possible in that short of a time period.

It can always be faster.

Also, I don't really need the 20GB for a boot drive and the Vertex 2 is $60 cheaper than the X25-M because of that.

 

Vertex 2 is a little bit faster than Intel in real life. I doubt you'll notice the difference without using a stopwatch though.

OCZ has a lot to prove with reliability. I've seen a lot more more reports of failed OCZ drives than failed Intel drives. Not exactly scienitific I know but the difference is significant to me.

+ OCZ reliabilty track record
+ OCZ firmware support track record

I wouldn't really recommend buying the Intel either because Anand has been saying it will be replaced mid year.

Keep in mind that an SSD will get slower when it fills up. An empty Vertex 2 is a little bit faster but a Vertex 2 with 50GB will be a lot slower than a Intel G2 with 50GB on it.

 

This man know's business. ^ I hate to sound like a broken record, but Intel is the brand, but I, like Phil advised, would wait until the G3 is announced.

 

The idea of the X25-M being "fast enough" is entirely arbitrary. Who gets to decide what is "fast enough"? Why is the X25-M fast enough while Indilinx-based drives aren't? It reeks of fanboyism to me.

Sure, I'd take an Intel drive over a SandForce, but that has nothing to do with speed - it's because the Intel drives are cheaper per GB. However, that doesn't mean I wouldn't take every bit of speed I could get at a reasonable price.

On the whole space vs performance topic, does anyone a link to data on the space vs performance degradation curves for these drives? Without the numbers, I don't see room for good judgement on the matter.

 

Not sure who said anything about the Indilinx drives. They can still be a good buy if the price is right.

I said you won't notice the difference between an empty Intel G2 and Vertex 2 without using a stopwatch. I still think that's true.

I do think you'll notice speed differences between a Vertex 2 60GB with 50GB on it and a Intel G2 80GB with 50GB on it. The Intel will be a lot faster I expect.

 

It's mostly this argument I was responding to:

Stating "fast enough" as an absolute and using that terrible F1 analogy just got to me there. Sure, no-one has ragged on the Indilinx drives either, but if I was going to make an argument as to what was "fast enough" I'd at least have an explanation as to why some alternatives are not "fast enough".

@ Phil: Your points are not what I was focused on, and you're probably right about the stopwatch.

I'm not so sure about your point with regards to spare area, though. Consider this, as well as the fact that this test is being very unfair to SandForce because the data is entirely random.

 

I currently have an X-25 E in my Latitude e6400 and a RunCore 128GB ZIF PATA drive in my Viliv that uses the Indilinx Barefoot controller. The BF controller is a proven controller, but the Intel seems more consistent. But those two can't be compared because there are MAJOR differences between them.

 

I'm not sure either but going by Anand it seems a full Sandforce drive looses about 20% in performance. The full Intel drive looses about 10% in performance.

The Intel and Sandforce we're comparing here is 80GB vs. 50GB so the Sandforce will be full quicker than the Intel. Especially since the OP only has one hard drive this is something to consider.

I'm still surprised that Anand seems to believe the IOmeter benchmarks are a reliable indicator of real life performance. There's a good chance real life performance would look a lot different.

 

Since it's necessary for further comparison, here's the normal (presumably empty-drive) results for the benchmark they used for that test - the results of the SF drive are 0%, 30% and 50% better than the Intel, depending on the benchmark. Sure, that benchmark isn't necessarily reliable, but at least it's a real-world benchmark.

 

Your link seems to point to the same page as before.

Unfortunately IOmeter benchmarks usually aren't accurate predictors of real world impact. It's known the Sandforce controllers perform very well in IOPS, it's not the full story though.

For example: Toshiba's Daikoku controller does quite bad in IOPS yet it manages to outperform Intel in many situations when it comes to real world performance.

 

I fixed the link.

Where have AnandTech said they use IOMeter for "AnandTech Storage Bench"? IOPS is just a unit of measurement - what are you suggesting they measure?

Can you give me a link for what you're getting at with regards to the Daikoku controller?

 

I prefer benchmarks like Techreport does them, where they just measure finishing a single real world job in seconds.

The Daikoku controller is used in Kingston V+ 64GB as benchmarked here:
A look at new SSDs from Corsair, Kingston, Plextor, and WD - The Tech Report - Page 1

 

I think that these numbers demonstrate that AnandTech's Storage Bench isn't biased against Toshiba's controller in the way that you think it would be.

I can't say for certain how their approach works because they're keeping it secret, but here's my best guess as to what's going on - first of all, they turn a real-world usage scenario into a fixed sequence of I/O operations on a fixed set of data, presumably something like a timeline of disk accesses. Having a sequence like this is a great idea, because it ensures that the comparison will be completely fair between all sets of drives. They run this sequence and monitor disk usage during that period of time. One important question is this - do they ignore idle time during playback in the way that TR's DriveBench does? If they do, you're correct in saying that this introduces a bias towards drives that are better at handling many operations at once (Intel's drives in particular). However, TR's results, as compared to AnandTech's, suggest that there is a definite difference in methodology.

Now, this is where the million dollar question comes up - if you're monitoring disk usage over a fixed period of time where the disk is idle some of the time, how does this transform into a benchmark measurement, and most specifically one that isolates disk performance alone? Consider how TR's benchmark works - they just measure the total amount of time taken and divide the number of operations (a fixed constant) by the time taken. My best guess here is that AnandTech's approach measures something along the lines the total amount of time the disk spends processing requests, and divides this into the total number of operations involved.

It's true that their benchmark isolates drive usage so you can't calculate how your drive speed will impact overall system performance, but this has both advantages and disadvantages. The biggest problem with your preferred style of benchmark is that the data points are so close together that random aspects of system performance put their reliability into question. Besides that, you would have to have individual tests for a large number of applications to offer decent scope, and it's completely impossible to reliably test multi-tasking without some kind of disk access recording/playback feature.

While a single-application test is good for demonstrating that in some situations the difference between two drives has little impact on overall system performance, it's quite limited in scope, and leaves out one of the biggest advantages of SSDs over conventional drives - multi-tasking.

 

I would recommend the Vertex 2 - I have one myself and have found it to be the best SSD Ive owned.

 

why did you say saying fast enough "fast enough" as an absolute and using that terrible F1 analogy"?

you sounded like a broken audio disk, can you speak like a normal person

intel x25m g2 is fast enough because it has
+no more LOOOOONG loading times
+ultra fast boot/reboot shutdown times
+can multitask(scanning for virus/spyware/malware while watching movies or playing games, installing software while playing games, copying files while a burning cd/dvd)
+has good random read and write speed
+has good sequential read and write speed
+doesn't degrade like crap like other ssd overtime
+has more GB than other ssd so you will have more free space
+more space means the ssd will be faster because the free space will be used to utilize the ssd performance

also using that terrible benchmarks analogy just got to me there.

sure it is faster in benchmarks but
-will my computer boot faster?
-will my computer copy/paste small and large files be faster?
-will my computer installing/ uninstalling software and games be faster?
-will my games load faster?
-will my computer multitask better?
-does those higher benchmark numbers mean anything in real life?(intel x25m g2 vs OCZ Vertex 2)

if you still doesn't get what i mean "fast enough" i suggest you watch this video comparing x25m vs WD RAPTOR 10,000RPM
YouTube - WD Raptor vs Intel X25-M, programs

also can you stop making an essay?

jk
LMAO

 

I'm going to be running this in a two drive setup. I'm not too worried about 20GB of extra space.

There's these two things called punctuation and capitalization. You should probably figure out how to use them in your writing before criticizing anyone else.

 

criticize?

i am just replying, i suggest you read better ok?

also i clearly stated why i think going anything faster than x25m g2 is pretty pointless

also as phil state, i wont be surprised if the x25m g2 will be faster in real life "more free space=better performance"

 

No, I suggest that you re-read what you wrote. You criticized him for writing too much and "sounding like a broken record."

When Anand starts using something in his personal setup, I believe that the Vertex 2 has to be somewhat better than the Intel X25-M. When you guys claim that the Intel drive is going to be faster in daily use, all of the data seems to go against that, even in the testing methods that attempt to "foil" the SandForce's compression methods.

 

I just hope the SSD manufacturers don't follow your advice and suddenly decide they simply don't need to make their SSDs any faster.

No one is forcing you to read that post, so what are you complaining about?

Why not go for something slower than the X25-M G2, then?

 

Actually I think it is. Anand's gaming benchmark makes a good example compared with this. (and that's compared with the 160GB Intel)

Or compare these benchmarks with these, very different pictures. At Anandtech the Intel G2 80GB seems faster in most benchmarks while on Tweakers the Kingston is faster by quite a margin. (both benchmarks measured in IOPS).

It's not very important though, I just prefer benchmarks that are based on single jobs where the SSD is the bottleneck, measured in seconds. These give the best reflection of actual real life performance. Like Techreport and Laptopmag do.

It's quite easy to measure multi task performance in a real life way, like done by Laptopmag here:

http://www.laptopmag.com/review/storage/intel-x25-m.aspx?page=6

 

i am pointing about price/performance ratio not performance/performance

LMAO it was meant as a jk(JOKE) if you dont get it, that makes it even more funnier

 

I would be remiss if I didn't point out that this is one person's opinion. Just because a particular piece of hardware is used in a production setup does not mean that it is the "best". Being the "best" is very subjective: he may have chosen a particular piece of hardware for any number of reasons, performance notwithstanding, parameters based on what he felt was the best for his computing needs.

In any event, in an effort to obtain the widest array of unbiased information when making a purchase decision, one shouldn't resign themselves to what one individual has in their particular system. In some (or nearly all) cases, this can lead to the misconception that just because one person is doing something, that's the way everyone else should do it as well.

 

I

I feel as though his needs from an SSD are identical to mine. He uses laptops as portable tools that do not need massive amounts of storage space. I am planning on using whatever SSD that I buy in a similar way; I will use it only as a boot drive. The fact that he went all Sandforce immediately after the Vertex 2 review seems significant to me. Before this, Anand pretty much believed that any SSD that wasn't from intel was sub par.

Also, the statement about how following one person is a bad thing. Are you saying that I should go intel like everyone else and their dog?

 

Another good example of real world testing,
Corsair F100 100GB
Corsair V128 128GB
Crucial C300 256GB
Kingston V-serie 30GB
OCZ Vertex2 100GB
Seagate Barracuda XT 2TB
WD Velociraptor 300GB
WD Scorpio Black 320GB

Hard-Drive Roundup June 2010 - Load Times

 

No, not at all. What I am saying is that, in an effort to make a wise purchase decision, you should research your options thoroughly and not base your decision on what one guy is doing.

Don't get me wrong, I think Anand is a terrific guy. He's extremely competent and knowledgeable, very tech savvy. However, relying only on one opinion for a purchase decision is doing yourself a disservice.

And just for the record, I bought the Intel X25-M 80GB G2 SSD a couple weeks ago. In my case, capacity was more important than the very incremental performance gain that the Vertex offers over the Intel product.

 

That's what I get for typing stuff up on a phone using a virtual keyboard. Basic spelling errors that wouldn't happen on a keyboard occur as a result of it.

The reason why I was defaulting to Anand was because of his comprehensive testing and competence in general. He's very good at explaining things at a pretty deep level as well.

 

It's not just Anand though... basically every review I've read of the Vertex 2 has put its performance above that of the X-25M.

Also it seems as though people have missed that the $/GB of the Vertex 2 and Intel X-25M drives are the exact same if not in favor of the Vertex 2 at higher capacities if you purchase from MyDigitalDiscount...

 

That's because MDD overprices the Intel drives.

 

I meant if you purchased the 60GB Vertex 2 at MDD... compared to 220$ for the 80GB X-25M which is the cheapest I could find it.

 

They also have shoddy shipping/billing practices too, and that comes from a previous infuriated customer...me. But that's besides the point..

$0.02

 

80GB for $220 is still cheaper per GB than 60GB for $180.

 

OCZ Vertex 2 50GB $189, $3.78 per GB ($3.15 per GB if it's 60GB)
Intel G2 80GB $195 $2.44 per GB

 

What he was talking about was the $165 deal that MDD has running for the Vertex 2 60GB. It makes the cost go down to $2.75/GB, the exact same as the X-25M when it is priced at $220.

Many of the prices on Google Products are outdated. Normally, $220-225 is the price that you can find just about anywhere for the X25-M, just like $190-200 for the Vertex 60GB everywhere else.

Newegg - OCZ Vertex 2 60GB - $194 - $3.23/GB

On Amazon, with rebate, the Vertex 2 60GB costs $189.

Newegg - OCZ Agility 2 60GB - $179 - $2.98/GB

Newegg - Intel X25-M 80GB - $220 - $2.75/GB