Vertex Frist Impressions

N

notsane

Gawd
Joined
May 23, 2007
Messages
583
Hey, this drive make things much snappier!

I just did a fresh install onto a Vertex 60GB SSD. Using Vista 64bit. I disabled defrag and indexing. Here is my first benchmark.
Vertex_1.jpg


I also noticed that my "windows experience" scores are all even for the first time. pretty balanced component selection I suppose.
BalancedSystem.jpg
 
Those sequential write scores seem awfully low, even with the smaller cache on the 60gb. What's going on there?
 
othellomcbane said:
Those sequential write scores seem awfully low, even with the smaller cache on the 60gb. What's going on there?
Click to expand...


The cache is the same on all 4 vertex models now 64mb. pre-release versions were smaller.

I hadn't done any reg tweaks before the benchmark. Also Benches are less reliable when repeated over time on SSDs, so we shall see if things improve once I have everything set up correctly.
 
That "less reliable when repeated over time on SSDs" thing is a bit of a myth... if that were the case, it would mean that SSDs themselves will get slower over time just from regular usage which is also a bit of a myth.

If an SSD isn't as fast and posts the same speeds a year in that it did the day it was first used out of the box, something is seriously wrong...
 
thebeephaha said:
I don't get it.
Click to expand...

The Vista rating system maxes out at 5.9, so all his hardware is not getting a 5.9 rating, its just as far as it goes. This has been changed in Win7 though and goes higher, but over all the Vista rating system is useless.
 
I am not sure why the reg tweaks and ramdrive are not helping benches much. I made so many tewaks at once, I can't tell... Hardly scientific method.

The partition is aligned correctly (semi correctly anyway). I suppose I will wait for the firmware out any day and install firmaware, realign the partition and re-install vista.

hrmpf!
 
If you're using Vista or Windows 7, and you formatted the SSD with those OSes and used NTFS, it's aligned, period. Not sure what the "semi" part there means, however. :)

Benches look fine to me for a single SSD, I don't see what the problem is or what you're thinking is wrong.
 
The only tweaking I did with my G. Skill SSD was moving the pagefile to another drive. I haven't seen anything that really makes the registry tweaks and such look worthwile.
 
I know what he means about the alignment. Vista doesn't get it exactly right for SSD and there are a ton of articles online that explain this. Now I kind of expected the Vertex to defeat or keep up with the Intel X25 especially in write speed but the Crystal mark scores arent showing that. I wonder if the larger cache on the 120gb Vertex drives will make that much of a difference. Either way those read speeds are bout as good as a pair of raptors.

Edit: your reads at 4K are pretty good. Here is an intel X25 80gb for comparison.

SingleIntelSSDX25M80.jpg
 
Copyright said:
I know what he means about the alignment. Vista doesn't get it exactly right for SSD and there are a ton of articles online that explain this. Now I kind of expected the Vertex to defeat or keep up with the Intel X25 especially in write speed but the Crystal mark scores arent showing that. I wonder if the larger cache on the 120gb Vertex drives will make that much of a difference. Either way those read speeds are bout as good as a pair of raptors.

Edit: your reads at 4K are pretty good. Here is an intel X25 80gb for comparison.

SingleIntelSSDX25M80.jpg
Click to expand...
Thanks so much for this benchmark to compare to! The intel drive has much higher scores is some areas, so that is very interesting (makes me wonder if I should have just gone Intel). Supposedly the new Vertex firmware is going to make things even better. Since I will need to re-install to update the firmware anyway, I will try the alignment mentioned in the OCZ sticky while I am at it and report back.

Don't get me wrong, I am happy with the performace and I think things are very snappy in general (much better than my WD 640 AAKS).
 
I saw the OCZ thread where the guy did the alignment and got some nice speed increases. Give that a shot and please come back with your scores. The Vertex drive is much cheaper then the intel. You can just about get a 120gb Vertex for the price of the 80gb intel. Plus the 120gb with extra cache may perform closer ot the intel speeds or better.
 
I think on the newer SSDs, the alignment doesn't help much

I got a masterdrive ox 32gb yesterday, and tested it out(before and after alignment etc) and the results were the same.

Read speeds are indeed close to 150mb(140mb/s average)

Write speeds start out at 90mb/s then drop(caching) to about 40mb/s after 30% of the benchmark average works out to about 50mb/s

I bought this SSD mainly to load some games on, so writing wont be a problem =) need read IOPS
 
Copyright said:
I saw the OCZ thread where the guy did the alignment and got some nice speed increases. Give that a shot and please come back with your scores. The Vertex drive is much cheaper then the intel. You can just about get a 120gb Vertex for the price of the 80gb intel. Plus the 120gb with extra cache may perform closer ot the intel speeds or better.
Click to expand...

Will do, but going to wait for firmware so I only update once!

FYI ALL Vertex drives have the same cache now. Only the pre-release versions of 30GB and 60GB had less cache.
 
Digital Viper-X- said:
I think on the newer SSDs, the alignment doesn't help much

I got a masterdrive ox 32gb yesterday, and tested it out(before and after alignment etc) and the results were the same.

Read speeds are indeed close to 150mb(140mb/s average)

Write speeds start out at 90mb/s then drop(caching) to about 40mb/s after 30% of the benchmark average works out to about 50mb/s

I bought this SSD mainly to load some games on, so writing wont be a problem =) need read IOPS
Click to expand...


Good to know. I have a G-Skill 64GB SSD (J micron controller) that I am using to load games, I will also try the alignment tests on that
 
notsane said:
Will do, but going to wait for firmware so I only update once!

FYI ALL Vertex drives have the same cache now. Only the pre-release versions of 30GB and 60GB had less cache.
Click to expand...

Woa really? There was a guy on a newegg review that stated his was 32mb but I guess you can only trust a newegg review so much. He even said his box showed 64mb but the drive showed as 32mb. Not sure how he was viewing it.
 
Copyright said:
Woa really? There was a guy on a newegg review that stated his was 32mb but I guess you can only trust a newegg review so much. He even said his box showed 64mb but the drive showed as 32mb. Not sure how he was viewing it.
Click to expand...

The only source for good Vertex info, today, is OCZ forums. It's chalk full of stuff, only hours old. It's incredible the mis-information that exists, already, between that site and this one.

Also, running those benches on an OS partition will read a LOT lower than running them off of an empty partition.
 
Digital Viper-X- said:
I think on the newer SSDs, the alignment doesn't help much

I got a masterdrive ox 32gb yesterday, and tested it out(before and after alignment etc) and the results were the same.

Read speeds are indeed close to 150mb(140mb/s average)

Write speeds start out at 90mb/s then drop(caching) to about 40mb/s after 30% of the benchmark average works out to about 50mb/s

I bought this SSD mainly to load some games on, so writing wont be a problem =) need read IOPS
Click to expand...

i had this happen on an mtron pro/linux. Sequential speeds are the same....the random seeks/second were 25% higher aligned.
 
Brahmzy said:
The only source for good Vertex info, today, is OCZ forums. It's chalk full of stuff, only hours old. It's incredible the mis-information that exists, already, between that site and this one.

Also, running those benches on an OS partition will read a LOT lower than running them off of an empty partition.
Click to expand...

The intel bench I provided was running an OS. I haven't been over to the OCZ forums in a while..I will have to swing back in and do some reading.
 
Copyright said:
The intel bench I provided was running an OS. I haven't been over to the OCZ forums in a while..I will have to swing back in and do some reading.
Click to expand...
I am getting mixed messages about how much it effects the benchmark. Also about the effect of repeted benchmarks. My first sequential bench was much higher than all of the following test (by 10+ MB/s)
 
notsane said:
I am getting mixed messages about how much it effects the benchmark. Also about the effect of repeted benchmarks. My first sequential bench was much higher than all of the following test (by 10+ MB/s)
Click to expand...

No one really knows at this point why the OS partitions read so differently than even half-full non-os partitions. Mine are DRASTICALLY different as well as most everyone elses. This is on Vx64 anyway, affecting all benches.

The biggest thing benches do to jack up an SSD is write all over the place which messes up the consecutive write bench results. If I use Hyperfast after running them, it "cleans up" all of the empty areas again and restores the SSD, in a sense. IOMETER is the worst of 'em.

Thus our desperate need for things like Hyperfast, PD10 etc...
 
I thought HyperFast and all the other "SSD optimizing" software was proven to be bullocks as no software (except the firmware of the drive itself) can actually see the physical blocks on the disks, so all they end up doing, is moving data about, in a way that the software (application level) think is "defragmented" but in reality, the firmware makes it impossible for the application to have any idea of what it's actually doing.

Am I mistaken here?
 
OK, I just flashed my firmware to 1199!

I booted up to my "old" vista 64 boot drive (OS drive from 3 days ago WD 640AAKS), flashed the vertex firmware, formated and aligned (4096) using disk managment, disabled indexing and enabled advanced performance then benched. This bench is not running an OS like the earlier benches were. I will re-install the OS tonight and see if the writes go down, though I don't think it will effect them much.

WOW! Yay OCZ!

Vertex_1199.jpg
 
notsane said:
OK, I just flashed my firmware to 1199!

I booted up to my "old" vista 64 boot drive (OS drive from 3 days ago WD 640AAKS), flashed the vertex firmware, formated and aligned (4096) using disk managment, disabled indexing and enabled advanced performance then benched. This bench is not running an OS like the earlier benches were. I will re-install the OS tonight and see if the writes go down, though I don't think it will effect them much.

WOW! Yay OCZ!

Vertex_1199.jpg
Click to expand...


Looking much better. I was about to post some info on the 1199 firmware in here. I saw another guy get a nice improvement with this firmware.
 
I was looking at the 1199 firmware briefly but I did not see a way to install in Linux. Anyone?
 
Can you run test with larger size, not 100MB? I wonder how much the 64MB of cache affects scores.
 
The other versions of the firmware updaters are still in the works. This is a pre-relase for windows only, sorry.
 
Hmmm I thought I read over at Toms hardware they traced the "performance degredation over time" to the power management built into some motherboards. I believe there is an article on it over there now.
 
All benchmarks so far have been with a 63k alignment. i used diskpar as per the OCZ how to to re-align to 128k. My benchmarks didn' really improve..
Before Alignment (63k)
Vertex_1199.jpg

After Alignment: (128k)
Vertex_128_4096.jpg


The 1199 firmware did make a HUGE difference though..

OK, now off to install vista
 
You must log in or register to reply here.
Back
Top