The new SSD Thread (Benchmarks, Brands, News and Advice)

EX is SLC and is out already, Vertex 2 is MLC internal RAID0

 

Looks like Newegg has some SSD deals.

G.Skill FALCON FM-25S2S-64GBF1 2.5" 64GB SATA II MLC Internal Solid state disk (SSD) for $179 after promo code EMCLSMS36 (with FS)

Super Talent UltraDrive ME 2.5" 32GB SATA II MLC Internal Solid state disk (SSD) for $130 after promo code EMCLSMS37 (with FS)

Actually I recently bought the Intel 80GB when it came down to $315. I am going to put it in my notebook and dual boot Vista x64 and Windows 7 RC.

 

Read and write throughput is hypothetically doubled (but in practice rarely is full doubled). Seek times are usually worse, but generally on for example a pair of desktop HDDs with a 12ms seek you might drop to 14ms, which is mainly a function of the drive that has the "longer" seek and some overhead to synchronize things as the first blocks come off the drive. SSDs have seek times in the 0.#ms range, waiting for the slow disk and overhead will still not make the seek time even close to that of a mechanical disk. I don't think I've seen a stripe set that had a seek time that even broke the 1ms barrier.

 

It looks like the Vertex EX drives are shipping without a firmware. http://www.ocztechnologyforum.com/forum/showthread.php?t=56703

 

Oh OCZ. This is why you have such a bad rep. I'd rather you release it later and with few problems than as fast as possible. On another note:

http://vr-zone.com/forums/431588/super-talent-upgrades-ultradrive-series-ssds.html

Super Talent's new Indilinx FW update.

 

LOL

Man, that's funny. Tapadong Barefoot? haha...

 

WOW That is FAST!!!

 

Good stuff on the Super Talent, but I have learned there is a lot more to SSD rating than the speed.

 

it should be done for all xp installations. doesn't matter if it improves performance, it lowers wear'n'tear (tm).

 

Has anyone tried the new TrueCrypt?
http://www.truecrypt.org/docs/?s=version-history
"The I/O pipeline now uses read-ahead buffering, which improves read performance especially on solid-state drives, typically by 30-50%. (Windows)"
http://www.ocztechnologyforum.com/forum/showthread.php?t=56547

 

While IMO not that great of a review, it does have comparisons for:

Intel's X25-M, OCZ Vertex, Samsung, and Super Talent ME.

http://hothardware.com/Articles/Fou...dux-OCZ-Corsair-Kingston-Super-Talent/?page=1

 

my sh*tty sony vaio tz398u came with a 64 GB SSD and a 300 GB HD; the ssd is faster. windows 7 boots up in 14 seconds.

 

Err, is it just me, or does that Intel rebrand perform pretty poorly in comparison.

Edit:

Nevermind, its IOMeter performance is... amazing.

 

Just got mine installed

 

It should be the same thing.

 

That's stupid fast

 

With so many benchmarks being used to test SSD, and the results changning from test to test for each SSD, has anybody really put together the "big picture" in what it all means, and how a layman would decipher those benchmarkes into what is the best SSD?

It was so easy with HDDs

Lets see for SSD there is

Crystal Mark
HD Tach
HD Tune
IO Meter
ATTO
Sandra
PCMark Vantage

Its odd since the winning SSD in one test is not the winner in the next, so its harder to relate the synthetic tests into real life performance.

So who thinks they have a good idea of how to read these tests and what each one represents?

If you were trying to evaluate an SSD what tests would you be looking for the most?

 

My test is to put an image of my current computer that I do all my work from on to the SSD and do what i normally do. Maybe test a few large files opening/copying, but pretty much just turn off defrag and indexing and let er rip. Problem is, I have not yet bought an SSD. Not since my debacle with an early Gskill Jmicron 60GB. I did the above and it worked. great. For about 5 minutes. Then the 5th time I opened Outlook, locked up tight, and that was pretty much it. from what i understand, some have dowen the same with the Intel 80GB and it works like a charm. So, I hope to try again in the future.

 

Crystal DiskMark and 4K Writes in IOMeter.

The rest don't matter.

 

I like ATTO myself in addition to those, it seems a pretty good bench.

One thing I was able to pull from that review link posted above is how some things take multiple tests to find the real strengths of a SSD..

That Kingston Model for instance had a much slower write speed compared to the rest, but sequential write, in the random write tests found in another test it was over 2x faster than the rest.

So the question is would you generally be writing to random places or sequential? Since SSD are not defraged I would think that over time it would become random rather than sequential if you read/write alot fragmenting the data. This would mean in real life tests it would be the faster even though the benchmarks put it in last place.

Those are the kind of things I am trying to learn about, so I know how to read tests given from reviews and choose the best SSD for the money.

 

I think the time has come to leave synthetic benchmarking behind. It does not really work anymore. There just not an accurate indicator for real life performance. Even for traditional hard drives.

There's a few websites online that are starting to understand that, like Laptopmag, Techreport and Anandtech.

Real world benchmarks. Booting, hibernating, file copy, application launching, virus scan, running scripts etc. The review done by Techreport on all the new SSDs is a very good example:
http://techreport.com/articles.x/16848/7 (several pages)

Laptopmag's review on the OCZ Vertex is another good example:
http://www.laptopmag.com/review/storage/ocz-vertex.aspx?page=2 (several pages)

PS. It's very interesting though that Techreport and Laptopmag get completely different performance of the OCZ Vertex, in file copying for example. There must be something wrong with the OCZ Vertex that Techreport had.

 

My guess is a different firmware, OCZ has had new firmware for that unit on a very regular basis. So its important to include your firmware version with a review also.

 

True yeah. They should do that for the hard drives as well.

I've sent the author of the article an email asking about the OCZ Vertex performance.

 

Just read the first review, it makes SSD look pretty bad in real life tests I wanted to get a 300gb Velociraptor and in the real life tests even the 150gb model was very close to the SSD performance for a small fraction of the cost (can get the 300gb velociraptor for $200) but thats for desktop use only, I guess in laptop land the difference from a 2.5hdd and a SSD is more substantial.

But those real life tests seem to to go against how fast all the SSD owners claim everything to be, boot time for intance. The SSD owners are claiming very fast boot times, but in that review the HDD's and the SSD got almost the same time.

Flaw in the test? or are the SSD owners under the placebo effect?

 

The Vista boot test by Laptopmag confirms that booting is not one of the strong points of the SSD. The hard drive is only one second slower on 52 seconds.
http://www.laptopmag.com/review/storage/ocz-vertex.aspx?page=4

There's a difference between Vista and XP boots though.

 

I'll be honest one of the things I was looking for was game load time and boot time, looks like the HDD is still not far behind.

I think for the laptop I will get a ssd when prices are better and performance may (will) go up by then, but for the desktop I'll go ahead and get that 300gb Velociraptor, it seems to do what I need at a fraction of the cost and gives me much larger capacity.

 

Interesting, the boot difference here between the SSDs and the Velociraptor is much bigger
http://www.anandtech.com/storage/showdoc.aspx?i=3531&p=29

PS. The Seagate 5400.6 is one of the slowest hard drives on the market. Shame they used it.

 

Can someone help me out over on this thread on trying to get HDDerase to work with Ultimate Boot CD?

http://forum.notebookreview.com/showthread.php?t=380663

 

i have 20sec in vista. no configuration done.
they clearly messed up boot times (which is easy on vista. driver initialisations can go terribly wrong == very slow)

 

It could also be that you measure differently.

Some people measure from pressing the power button while others start after the bios. Some people measure until the first appearance of the desktop, while others finish when all task bar processes are loaded or wlan connected.

The guys at Anandtech also measure around 50 seconds
http://www.anandtech.com/storage/showdoc.aspx?i=3531&p=29

 

I have a question about the vibration when it comes to SSD.

My motorcycle has EFI, so I am able to hook a laptop up to it and do my own tuning and record the air/fuel ratio when it running at a stand still or on a dyno.

One time I did put the running laptop in the saddle bag and went for a ride to record the data... The hard drive was a paper weight when I finished the ride - put a new hard drive in and the computer worked perfect.


I also live in an area with cobblestone (bricked) roads and side walks. Due to a bad back, I will have to get a laptop roller bag - planning on getting the Dell M6400 (8.5 lbs).

Would the SSD drive be able to withstand vibrations under these conditions?

 

If the motherboard can withstand it, then I would think an SSD could too.

Not even close in access time. I upgraded from a Velociraptor and there was a very noticeable speed improvement.

Newegg has the Vertex 30GB today (until 2:59PM) for $99 AR. Look at the Shell Shocker page.

 

Many thanks

 

OCZ Vertex Series 30GB Internal SATA Solid State Disk (SSD) $99.99AR Free Ship, May. 18 6 AM

Newegg Shell Shocker, Part 1 - Valid until 3 P.M. PT) Newegg has the OCZ Vertex Series OCZSSD2-1VTX30G 30GB 2.5" Internal SATA Solid State Disk (SSD) - Hard Drive Replacement - for a low $99.99 after rebate. Free Shipping. Tax in CA, NJ, PR, TN.
$20 rebate Exp 5/18/09

 

Not sure if it has been mentioned yet, but there appears to be a $40 rebate on the 120 GB Vertex. I saw the rebate pdf from an eBay page. And there is a $30 rebate on the 60 GB Vertex.

That 8% off for eBay via live.com makes the 120 GB a pretty good deal, for those who have to pay tax on newegg purchases.

 

I saw the $40 rebate through mwave.com but with the shipping and starting price you might as well just buy it from Newegg.

 

in for a third 30gb vertex from newegg.

got two from there $80 sale a while back.

when I get the third maybe I will finally install them as 3 way RAID 0. 90GB should do me

 

30gb is way too small for somebody like me, thats ok for a business person with just the OS and a few programs, but I want room for games and other stuff too.

It works out in my mind to be more valueble to have a larger space that runs a bit slower like the VR so I can use it for all my files, than to give a very large boost in speed for only a portion of my files.

Infact since on my desktop the only real loading time I have for anything is my games, and the only hdd bottleneck I have is my video rendering the VR will help with both.

Im in for a SSD for my laptop for sure since the VR is not an option, but I am going to be more patient and wait for better prices, and that will will equate into better products and probably higher capacity too.

 

Don't forget if you use all the space on your VR the inner rings are going to be MUCH slower than the outer... So really you prolly get maybe 150ish gb of descent sequential speed and 150ish gb of slower storage...

I speculate that with the recent newegg push to sell the indilinx drives we should be seeing the G3 soon or else they just realized they marked up the drives WAY too high there for a while...

 

Slower for sure, but a 2.5" drive shouldnt suffer as much as a 3.5" drive so I think MUCH is a overstatement. But it will be MUCH cheaper

The per dollar value if you figured out the capacity of a ssd vs the price and then factor in the speed of a ssd + price is better on the VR, in addition like I said what good is a SSD if it can only speed up half of my stuff instead of all of it? Boot & Programs are already fast on a desktop like mine as all the programs stay loaded in ram, and I never turn off my desktop so I never boot. Its mainly games & video work I am worried about.

Im not trying to convince anybody else that the VR > SSD, just stating that I already determined it is better for me specifically considering what I need it to do.

My notebook is a different story, there is no VR option, and I do have to boot it up every day I use it (though this really does not bother me, boot is fast on a well cared for machine and while it boots is when I am plugging in my mouse and power cord so by the time I am sitting down ready to use the computer, its ready to use) I still need atleast a 120gb SSD for my notebook though and will probably pair it with a 500gb drive in my 2nd bay.

 

I suppose this can be accepted

 

bumpy lun,[y

 

It could be that their test computer takes a while before it starts loading the OS like mine does. I'm at the 30 second mark before the OS starts loading, then it's another 31 seconds until I'm at the desktop. Even if the SSD allowed instantaneous booting to the desktop, my bootup would still take 30 seconds. Seeing as how that won't happen, any improvement on my laptop won't seem as dramatic.....a Win load time cut in half by 50% would still only decrease boot times by 25%.

 

vertex 30gb at newegg sold out....glad I jumped on it earlier.

 

up online after bios, including msn loaded and firefox, around 20sec. bios, 7 sec..

 

bios 30secs? somethings terribly wrong then. if its at the loading barm try disabling all components like bluetooth etc to see if it makes a difference. it should. did here at least. 2min difference actually because of crappy cardreader.

 

what's an actual good price for the intel 160gb? i see around 593$ (i see CHF) on ebay.

that's quite nice i think. still waiting for an actual buyable 160gb 1.8" (but this one could fit my pc just well and my mtrons could get put to other uses )

 

does anyone know if super talent even have a support forum like ocz, corsair? I tried looking for it, but I couldn't find it.

 

Here's a little benchmark I like to run.

http://www.mediafire.com/file/fiy0n2yyehe/MVI_4784.avi

Bootup with a batch script in the Startup folder that loads 18 programs immediately on boot. List of programs I used:

start /d "C:\Program Files\GIMP-2.0\bin" gimp-2.6.exe
start /d "C:\Program Files\Microsoft Office\OFFICE11" Excel.exe
start /d "C:\Program Files\Microsoft Office\OFFICE11" WinWord.exe
start /d "C:\Program Files\Microsoft Office\OFFICE11" Powerpnt.exe
start /d "C:\Program Files\Internet Explorer" iexplore.exe
start /d "C:\Program Files\MPC HC" mplayerc.exe
start /d "C:\Program Files\Nero\Nero Burning ROM" nero.exe
start /d "C:\Program Files\TechSmith\Camtasia Studio 6" CamtasiaStudio.exe
start /d "C:\Program Files\Windows Media Player" wmplayer.exe
start /d "C:\Program Files\WinRAR" Winrar.exe
start /d "C:\Program Files\Winamp" winamp.exe
start /d "C:\Program Files\Mozilla Firefox" firefox.exe
start /d "C:\Program Files\Audacity 1.3 Beta (Unicode)" audacity.exe
start /d "C:\Program Files\DAEMON Tools Pro" dtpro.exe
start /d "C:\Program Files\FurMark_v1.6.5" furmark.exe
start /d "C:\Program Files\HD Tune Pro" HdTunePro.exe
start /d "C:\Program Files\Iometer.org\Iometer 2006.07.27" iometer.exe
start /d "C:\Program Files\SPSSInc\Statistics17" statistics.exe

-Time BIOS was visible (0:09) to desktop visible (0:31): 22 seconds
-Time BIOS was visible (0:09) to fully loaded (0:52): 43 seconds
-BIOS visible (0:09) to 16/18 programs loaded (0:42): 33 seconds (Camtasia and Daemon Tools took an extra 10 seconds to load after #16, SPSS, loaded)

My P5Q Pro is a bit sluggish off the line which accounts for the delay. A boot time with that kind of program setup is simply impossible on a mechanical drive. Even interpreting the timing conservatively, at 0:35, the computer was fully usable and responsive as illustrated by the fact that 12 of my 18 programs had already initialized themselves.

 

I already knew that app launch time is one of the strongest if not the strongest SSD points. It was even mentioned in the reviews linked here.

But thats not a real world test becuse nobody is going to open 17 programs at once

This was also one of those things I said way back when, 90% of those programs other than the large adobe ones open basically instantly already on a mechanical hdd but even if you managed to take 1 second and make it .1 seconds what did you really gain?? A piece of mind maybe but nothing that really is significant from a statisticians point like mine.

Also I can tell you that the 18+ programs opening at once test is squed in the SSD's favor because it can pull data from multiple points without any loss of speed. If you looked at the real life tests on this page: http://www.laptopmag.com/review/storage/ocz-vertex.aspx?page=3

You can see the no stress and stress tests are almost the same for the SSD but the HDD takes a huge fall.

Opening 18 programs at once is inducing stress making the mechanical hdd take that huge fall, but in real life people are going to open one program now, another program in 5 minutes. So it will be the non stressed state making the hdd & ssd gap much smaller than what that test trys to show if you ran the same test for a hdd.

So it is a good test to show what a SSD can do, but not a good comparison tests for going against a HDD.