So... X25-M or Vertex.

Got 3x30GB Vertex here:
HDBenchmark2.jpg

HDTune2.jpg
 
Yep, the Intel update for my X25-M was just something I burned on a disc and did outside of the OS. Never needed to reinstall my OS and all my files were OK after the fact. Also, not that it matters much, but my WEI score jumped from 7.4 to 7.7 after I did the firmware update...

As for the price, all I know is that the 80GB X25-M was about $600 back in November/December last year. You could get one from newegg for like $280 with a combo deal when I got one in March this year (haven't checked prices since). So they've already dropped by over 50% of their price as far as I was concerned. I'd guess they'll be approaching the $200 by year's end.

Were you running RAID? Even an X25-E fresh out of the box will only net you a 7.5 and the numbers are up across the board vs an X25-M. Maybe you are using a newer build? I'm on 7201 after using 7100 and 7000 before. OCZ Vertex EX also tops out at 7.5.
 
The price of NAND isn't getting cheaper any time soon, it's on an upward trend for the last year and a half. When a cheaper manufacturing process becomes stable enough for mass production, prices will drop. There's no scuttlebutt at all at this point for possible transition times.
 
The Vertex does not suffer from performance degradation. The Intel is faster in random writes, but like anything it's all what price/performance ratio you are comfortable with.

http://www.bit-tech.net/hardware/storage/2009/05/15/ocz-vertex-ssd-review/11

And as Dan_D mentioned in another thread, don't mind the negs they say about Intel because the new firmware fixed that.

all SSDs have issues with degradation. its just the matter of degree. intel just released the firmware regarding this issue on the X25 SSDs.
 
The price of NAND isn't getting cheaper any time soon, it's on an upward trend for the last year and a half. When a cheaper manufacturing process becomes stable enough for mass production, prices will drop. There's no scuttlebutt at all at this point for possible transition times.

This seems to be likely happen before the end of Q4, when all the manufacturers are beginning soon to use their newest processes. But the price should also be dropping pretty continuously over and a long the time, but slowly. Hate to wait. :p
 
When comparing the vertex and intel, will those differences be that noticable in performance?

If I'm coming from a single raptor as my boot drive, I am sure to notice a big difference between my raptor and either of these SSD's.

But if I was switching from a vertex to an intel, would I actually notice any difference?
 
When comparing the vertex and intel, will those differences be that noticable in performance?

If I'm coming from a single raptor as my boot drive, I am sure to notice a big difference between my raptor and either of these SSD's.

But if I was switching from a vertex to an intel, would I actually notice any difference?

YES. I just switched from a single 300GB Velociraptor (I used to run a two drive 300GB Velociraptor RAID0 array before one drive died) to an Intel X25-M 80GB RAID0 array. While the RAID0 X25-M array is a bigger jump than a single drive provides, I can tell you from the performance numbers I'm seeing, that you will definitely see a difference in performance.
 
YES. I just switched from a single 300GB Velociraptor (I used to run a two drive 300GB Velociraptor RAID0 array before one drive died) to an Intel X25-M 80GB RAID0 array. While the RAID0 X25-M array is a bigger jump than a single drive provides, I can tell you from the performance numbers I'm seeing, that you will definitely see a difference in performance.

I wonder why Astralite and thebeephaha stated earlier in this thread that a RAID 0 setup would either degrade or not improve SSD performance? After reading of your testing, I was preparing to setup 2 X25-M's in RAID 0. I would never have thought it possible for a RAID 0 setup to degrade performance compared to regular SATA???

As you eluded to, an ICH10R (for example) would appear to have more throughput running in RAID 0 than in SATA.

Again, I am only referring to use of the X25-M's so as to remain on-topic. ;)
 
I wonder why Astralite and thebeephaha stated earlier in this thread that a RAID 0 setup would either degrade or not improve SSD performance? After reading of your testing, I was preparing to setup 2 X25-M's in RAID 0.

As you eluded to, an ICH10R (for example) would appear to have more throughput running in RAID 0 than in SATA.

Again, I am only referring to use of the X25-M's so as to remain on-topic. ;)

Yes, SSD's do gain from RAID applications just as mechanical hard drives do. It would even seem that they gain more from RAID0 than mechanical disks do as they work really well in parallel based on their design. This topic was covered in the Ask an Intel Solid State Drive Engineer article. Specifically this question and answer addressed running RAID with the X25-M:

Q. How well does RAID-0 work with SSDs? Can RAID-0 be done inside the drive?

A. Let me answer the second question first and point out that SSDs are already massively parallel. The current Intel SSDs use 10 parallel channels to access the flash media, so in a way, it's already a 10-way RAID-0 on the drive.

Doing RAID-0 with multiple SSDs works very well, but make sure your RAID card is up to the task. SSDs are individually capable of enormous amounts of throughput, potentially much more than the RAID controller was designed to handle.


In benchmarks SSD's almost score linearly as you add drives. So a four drive X25-M array benches twice as fast as my two drive array does and my second drive doubles my performance numbers over a single drive. My own testing results show this to be the case. Though I can only verify this with two drives. Now benchmarks are benchmarks and real world tests are another thing. Can you tell the difference between a RAID 0 array and a stand alone X25-M? Not really sure. I haven't run my OS on just one drive.
 
Great information, Dan.
Thanks as always.

I just am wondering what the basis would be for anyone stating that there would be "degradation" using RAID 0, or no improvement? If there is any validity to the Intel SSD Engineer's article (which I believe there is), then there is no way that RAID 0 would not help.
 
Great information, Dan.
Thanks as always.

I just am wondering what the basis would be for anyone stating that there would be "degradation" using RAID 0, or no improvement? If there is any validity to the Intel SSD Engineer's article (which I believe there is), then there is no way that RAID 0 would not help.

I'm not sure. I know that with the Raptor and Velociraptor drives they improved in sequential read performance, but their seek times went up. The seek times were actually one of the best things about those drives so one of their major advantages was eliminated by running them in a RAID array. I do not know how a RAID0 array would negatively impact the SSD's from a performance standpoint. I don't believe that they do.
 
Well, I just spoke to an Intel tech person, and they told me that, yes, the performance would be faster with a RAID 0 on ICH10R, but that since the ICH10R was designed before the advent of SSD's, there would be some possibility for errors, which would otherwise not exist in single-SSD config's.
 
Well, I just spoke to an Intel tech person, and they told me that, yes, the performance would be faster with a RAID 0 on ICH10R, but that since the ICH10R was designed before the advent of SSD's, there would be some possibility for errors, which would otherwise not exist in single-SSD config's.

I have yet to run into any problems with running the RAID configuration with the SSD's.
 
Dan, it's well and good to talk about numbers, but do you feel that the X-25M provides a tangible and worthwhile benefit over a velociraptor when it comes to everyday usage? I got an OCZ Vertex when they first came out and it turned out to be a real dud... waited through a couple of firmware revs before I finally returned it, but even in peak operating condition, I didn't feel it provided any benefit other than the lack of noise. I've been tempted to try an X-25 for quite a while, but the price tag and no return policy combined with my OCZ experience have made me a bit gun-shy, so I'd really value your opinion.
 
Dan, it's well and good to talk about numbers, but do you feel that the X-25M provides a tangible and worthwhile benefit over a velociraptor when it comes to everyday usage? I got an OCZ Vertex when they first came out and it turned out to be a real dud... waited through a couple of firmware revs before I finally returned it, but even in peak operating condition, I didn't feel it provided any benefit other than the lack of noise. I've been tempted to try an X-25 for quite a while, but the price tag and no return policy combined with my OCZ experience have made me a bit gun-shy, so I'd really value your opinion.

Yes I do. I've noticed improved responsiveness in almost everything I do. Especially general Windows usage. It is one of the most noticable upgrades I've ever made to a computer. I plugged my VR back in briefly to do some work as most of my software was still on it and it was easier than reinstalling everything. I ran the VR again for a few days and the whole time I felt like my computer was responding with all the speed of a glacier.
 
If you are asking if there is a huge gulf between X25-M and the OCZ Vertex, then no. The real difference is in high queue depth operations, like server use, though you wouldn't run an MLC drive for that purpose anyway. From a practical difference its pretty negligible.

However, the X25-M with fw. 8820 probably has the best steady-state performance I've seen in an SSD. The Vertex EX (30%), X25-E(20%), and Vertex (20%) all have varying levels of degradation, in that order.
 
Last edited:
Dan, it's well and good to talk about numbers, but do you feel that the X-25M provides a tangible and worthwhile benefit over a velociraptor when it comes to everyday usage? I got an OCZ Vertex when they first came out and it turned out to be a real dud... waited through a couple of firmware revs before I finally returned it, but even in peak operating condition, I didn't feel it provided any benefit other than the lack of noise. I've been tempted to try an X-25 for quite a while, but the price tag and no return policy combined with my OCZ experience have made me a bit gun-shy, so I'd really value your opinion.

Its these problems people are having and the fact there is high speculation that there will be a 50% drop in prices in 6 months, makes me just decide to wait just a little bit longer.
 
It's a gut feeling dude ...

xxx27bu.png


It would appear, Hopkins, that your gut feeling was only indigestion.
 
Can you guys that have the Intel drives confirm that the degradation is much better under control after the newest firmware?

I'm building a couple systems for some family members soon and will be doing SSD for os/apps and a WD 640 for data. The random writes are a good selling point. But it's still twice the price as a very fast Vertex and I only truly need 30gigs of a fast SSD to handle windows and programs for these guys. The ease of use updating the intel firmware via cd is the biggest selling point since I won't be able to be there to coach the updating process all the time. A simple flash update without reinstalling everything would be pretty simple for them. But if there are still issues I will just go with the Vertex. :)

What do you guys think has Intel got this under control it seems?
 
Last edited:
Can you guys that have the Intel drives confirm that the degradation is much better under control after the newest firmware?

I'm building a couple systems for some family members soon and will be doing SSD for os/apps and a WD 640 for data. The random writes are a good selling point. But it's still twice the price as a very fast Vertex and I only truly need 30gigs of a fast SSD to handle windows and programs for these guys. The ease of use updating the intel firmware via cd is the biggest selling point since I won't be able to be there to coach the updating process all the time. A simple flash update without reinstalling everything would be pretty simple for them. But if there are still issues I will just go with the Vertex. :)

What do you guys think has Intel got this under control it seems?

All the data says that it's under control. The steady state performance of the Intel X25-M is the best of the bunch. I haven't tested it myself, but even if it isn't 100% right now, Intel will continue to work on it.
 
The Intel X25-M is better at maintaining performance under steady-state than anything else you can bring up--better than the Intel X25-E, the OCZ Vertex, and the OCZ Vertex (or any Indilinix MLC variant).

The only drive that would surpass the X25-M in maintaining performance during steady state is the OCZ Summit, because the drive has a built-in TRIM function that operates during down-time.
 
Dan, it's well and good to talk about numbers, but do you feel that the X-25M provides a tangible and worthwhile benefit over a velociraptor when it comes to everyday usage? I got an OCZ Vertex when they first came out and it turned out to be a real dud... waited through a couple of firmware revs before I finally returned it, but even in peak operating condition, I didn't feel it provided any benefit other than the lack of noise. I've been tempted to try an X-25 for quite a while, but the price tag and no return policy combined with my OCZ experience have made me a bit gun-shy, so I'd really value your opinion.

Sorry to say man, you must have had something VERY wrong with your Vertex. The overwhelming majority of buyers are thrilled with their experience and performance of these SSDs, myself included. The Vertex is a rock-solid SSD. Unbelievably faster in everyday usage and feel and reality than my Velociraptors were. I've even had the original FW all the way up to the current.
 
The Intel X25-M is better at maintaining performance under steady-state than anything else you can bring up--better than the Intel X25-E, the OCZ Vertex, and the OCZ Vertex (or any Indilinix MLC variant).

The only drive that would surpass the X25-M in maintaining performance during steady state is the OCZ Summit, because the drive has a built-in TRIM function that operates during down-time.

TRIM may be coming to future firmware revisions on the Intel drive. No one can say for sure right now, but it is the rumor.
 
the X25-M with fw. 8820 probably has the best steady-state performance I've seen in an SSD. The Vertex EX (30%), X25-E(20%), and Vertex (20%) all have varying levels of degradation, in that order.

I used the wiper.exe tool on my OCZ Vertex and now Falcon and it gets the ATTO benches back to fresh, is that a real indication that performance has been restored to a fresh drive in the real world or is there still some degradation not shown by ATTO?
 
Last edited:
Trust me you are not getting better performance from a mechanical drive.. I went from the Velociraptor to the Vertex and the difference is stunning, apps launch as if shot from a gun, games load in the blink of an eye and reboots at light speed.. no contest
 
Last edited:
Trust you? What kind of reason is that?

Well let me put it another way, trust every single reviewer or owner of an SSD or keep deluding yourself, your choice pal.

You can raid a 1000 mechanical drives and still never get near a single SSD with a Seek Time of less than 0.1ms.

2jff4uo.jpg
 
Last edited:
am I reading this right? are they really this low price? they used to be $375, they really dropped this much?

intel x25-M at 229.99 with free ship? (newegg)





Intel X25-M SSDSA2MH080G1 80GB SATA II MLC Internal Solid state disk (SSD) - Retail
Get a SSD converter w/ combo purchase, limited offer

Image Viewer

$229.99
Free Shipping*


http://www.newegg.com/Product/Product.aspx?Item=N82E16820167005
 
Well let me put it another way, trust every single reviewer or owner of an SSD or keep deluding yourself, your choice pal.
There are no delusions. As I believe I've mentioned in this thread, I own some SSDs, as well. In testing

You can raid a 1000 mechanical drives and still never get near a single SSD with a Seek Time of less than 0.1ms.
Random access time is only a part of overall in situ performance. Once the SSD drive starts falling behind in writes, the fast seek time is overwhelmed by the the terrible write performance.
 
am I reading this right? are they really this low price? they used to be $375, they really dropped this much?

intel x25-M at 229.99 with free ship? (newegg)





Intel X25-M SSDSA2MH080G1 80GB SATA II MLC Internal Solid state disk (SSD) - Retail
Get a SSD converter w/ combo purchase, limited offer

Image Viewer

$229.99
Free Shipping*


http://www.newegg.com/Product/Product.aspx?Item=N82E16820167005
Checkout my post # 21 in this Bing cashback thread, Intel X-25M 80GB for $200 shipped after cashback.

But then again some drastic price cuts are coming soon, and the Gen 2 drive will be released as well...can't hurt to sit the fence for a lil while longer.
 
There are no delusions. As I believe I've mentioned in this thread, I own some SSDs, as well. In testing

Random access time is only a part of overall in situ performance. Once the SSD drive starts falling behind in writes, the fast seek time is overwhelmed by the the terrible write performance.

There is truth in what you are saying. Seek times are indeed only part of the equation. However you have to keep in mind what kind of controllers enthusiasts and average computer users are going to have in their systems and what kind of drives they'll most likely use. They aren't going to have badass LSI MegaRAID controllers and several 15,000rpm SAS drives in RAID-0, RAID-5 or anything along those lines. Not to mention those types of configurations excel at certain types of usage patterns which generally speaking, are not what they'll be expereiencing with a gaming PC. I think an SSD or a pair of SSD's will probably give them the most bang for their buck at present provided they choose wisely when they purchase an SSD. Not all SSD's are created equal as is the case with most things.

What is your exact controller/drive setup anyway?
 
Last edited:
So mikeblas,
How many enterprise drives does it take to get ~20MB/sec 4k random writes, and how much do the drives cost? One Indilinx drive can do this. Mine does, and I have had it for several months now and installed several OS's on it, so it is running steady state. What single enterprise drive can do 250MB a sec sequential reads and 150MB a sec sequential writes?
 
Back
Top