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

Video is a "no way" proposition if we're talking High Definition, it's way too bandwidth intensive. For Standard Definition it could work.

Either way, you're still talking about way too much storage utilized.

Standard Def (DVD quality, MP4 encoded) 5 Mbits/second, or ~.6 MBytes/second. Not too bad, roughly 2.2 Gbytes/hour.

High Def (High Quality, MP4/H.264 or VC-1) 12 Mbits/second, or ~1.5 Mbytes/second or about 5.4 Gbytes/hour

High Def (Blu-ray or HD-DVD quality) 20 Mbits/second, 2.5 Mbytes/second, 9 GBytes/hour. The same estimates would be just a little high for Over The Air archiving.

Yes, I do write about this from time to time

 

Wow John! Nice to have you aboard here.

 

Interesting test at Anandtech with the Air;
http://www.anandtech.com/memory/showdoc.aspx?i=3287

Kinda surprised at the batterylife numbers.

PS, what's the issue with HD recording? Most capture solutions have effective encoding to make it easier (and usually your source is 720P or 1080i so lower requirements than 1080P|60), and your system memory is usually used an effective buffer.

 

Theres nothing about testing that with the Air. They speak of the Air then go on to speak about the Macbook Pro with the Memoright 128Gb ssd which we have already done here at NBR. There results are significantly different, and very questionable, I would suggest at only 61MB/s read and 73MB/s write??? They should come around 100MB/s-120MB/s mark.

 

Storage capacity is the issue with HD video, not bandwidth. At 10 Mbits/second you are talking ~4.5 GBytes/hour. It goes up from there. A 64GB drive (68.7G in base 10 terms) after OS and apps has maybe 50GB of usable space.

Yes you can encode at less than 10 Mbits, but quality will suffer. Once quality starts to suffer is it really high definition?

As for 1080p, the numbers I cited are for 1080p24 encoding, ie film. The numbers get worse at 60p.

OTA 1080i with MPEG-2 doesn't look very good at anything less than 14 Mbits/second (minimum). There's visible softening to the image.

With MPEG-4 (AVC/H.264 High Profile) or VC-1 you can get by with a tad under 10 Mbits/second, but not by much.

720p is barely high definition to me these days. Does that make me a snob? Maybe; but since I have multiple displays that are 1080p capable, and on my main display I sit where 1080p is still resolvable I'm not too worried about it

 

Are the results backward? Write exceeding read speed?

Is it possible that they are using a SATAI instead of a SATAII interface?

 

Well that review provides a resounding "meh" especially at almost $4K for the drive. Worse battery life than a 5400 rpm drive? Ewww...

Think this might be the culprit:

 

No...typically the Memoright is higher write than read as can be seen in my benchmarking earlier on. Here is my benchmark of the memoright 32Gb ssd, 104MB/s write with 88MB/s read.

I don't understand what they are saying about the pata-sata conversion though as the MacPro is a SATA interface I believe.

I could be wrong as Im not a mac guy.

It almost seems like the author mixed up his reviews in a way. If you look way back into my history as well, you will see that I did a battery comparison earlier on which showed clearly that SSD made a difference and it is common sense.

Hard drives creat ALOT of heat which results in continuous fan use not to mention the additional power being used when the HD is being pushed. This is a large contrast to the ssd being pushed with NO moving parts and ALOT less heat.

My fan only EVER comes on on my M1330 when I got my external 22" hooked up. Other than that...I never notice it except for the PCMark05 testing of course.

 

Ok, but that's at odds with how the technology works as you have to "clear" the cell electrically before you write to it

They are saying that the interface controller to the SSD array is PATA and that there is an additional layer of PATA / SATA conversion that allows for the SATA interface.

I'm a unix guy, it's platform agnostic

Assuming a good implementation this would be true. Apparently not all implementations are good ones at this time...

What is the raw power consumption for both at idle and in use. That's the answer to both the heat and the battery life issues. More power consumed == both a shorter battery life and more heat.

Not noticing doesn't necessarily equal not on

 

Ah ok, yeah couldn't figure out the issue, space, definitely, but was thinking throughput was the problem you were focusing on.

Hey I agree, and yes I am a admitted audio/video snob, kinda byproduct of working in TV for 3 years in Montreal as a cameraman; the more you know and the more you're exposed to quality hardware and content, the more you hate the little imperfections. Just an expesnive habit that's the only annoying thing.

 

Sorry missed that, read a bit of the intro, went bleh, and skipped to the tests results (I usually check specs for CPU or graphics tests or something involving drivers, timings, and multiple other facets that would affect outcome, but didn't bother for this). OOoops

Anywhoo, interesting point Anand made about the voltages which made sense, but if it conflicts with your experiences, then that's interesting in the potential of it being a little more setup sensitive (chipset, configuration) than I originally thought.

It would be similar to some 5400RPM drives outperforming 7200RPM, now we have to make sure we know what configuration to mix if it affects the benefits we were hoping for (personally I want both performance and battery life improvements, but right now I need battery most of all for long flights since the Fujitsu has only small battery options).

 

Nope, think of the opening to the ST series:

"Space... the final frontier".

I think I made the case for SSD when it's a mobile solution -- less working parts to go "thud" in the middle of the session. But you do run out of space quickly with HD content.

Real men measure their screen sizes in feet My main display is 9' diagonal in 16:9 and 11' in 2.35:1.

I do quite understand about the imperfections, I hate hearing bad ADR punch-ins and I hear them far too frequently.

Cheers,

 

K...the reviews up, up, up!!! Click on it in my sig block!

 

I would think this is a record for the SSD I just reviewed. Its in the Dell 3 day sale at 40% off the m1330/M1530.




If anyone was going to jump, now would be a good time as they come standard with 4Gb RAM, WifiN and LED is available once again!

 

can you buy this if you don't have a dell?? and is this the sata 2 samsungs??

http://accessories.dell.com/sna/pro...tdetail.aspx?c=ca&l=en&cs=cadhs1&sku=341-6670

 

The link you sent does not mention it as being the Ultra Preformance which is crucial. Last I had heard, it would be a bit before you could order the new SATA II separately from Dell.

 

Looks like they changed the offer again -- when I go on the Dell website they're charging $750 for the 64G SSD option on the xpsm1330

 

Ok ok... so, let me get something straight... the latest version of SSD is the SATA II SSD??? made by different manufacturers. And the biggest capacity is the 64GB for now, right?.. the 128GB will be out the next year I read it somewhere.....
Somebody know what type of SSD the M15x have?... is it the SATA II? , from samsung?...

thank you very much for your help

 

it's samsung but not sata II yet... sata II samsungs are in mass production right now shipping to companies like alienware, as per something i remember reading on samsung site....

so expect it within the next month or so?... unless they're holding it back

 

Ummm...check my review in my sig block. It is on the Samsung SATA II. They are produced and available in Dells...which would mean Alienware as well. I know of atleast 3 people who have received them in their new systems already.

Also, Samsung will have a mlc 128Gb ssd out within the next few months.

 

last time i chcked ( a few days ago) with aw tech, they don't have the sata ii's in stock yet... plus the aw config pages don't say anything but solid state drive... and the prices are the same as before.. no ultra performance or anything..

 

Please see my review. There are even pictures....

I don't know who you are checking with but:

1. They are listed on the site as configurable options;
2. My system has a Samsung SATA II 64Gb ssd in it while I type;
3. I know of others who have received them IN their new systems after order.

They went into production a month ago and shipped immediately after that.

Matter of fact...see attached. Its a pik of the ssd details. Notice the important stuff like production date, SATA II and so on.



<a href="http://img604.imageshack.us/content.php?page=blogpost&files=img297/682/dsc00567ml2.jpg"

 

yours are dell.. he asked if aw uses the sata ii.. not if dell uses sata ii. when i checked with aw tech support a few days ago, they still don't offer sata ii.

 

AW is owned by Dell. I was just trying to clarify that they are manufactured and shipping, albeit, you may be correct in that AW does not have them available as of yet. Sorry for any confusion

 

@ JL1989 & Les Thank you very much for your response.... After reading the Les revie about the SATA II 64GB Samsung I'm very sure I will get it no matter what on my future M15X or maybe M17X... I don know how I'm goint to pay the $1,200 but i don't care, I want the SATA II SSD!!!...

By the way I asked AW and the told me the M15X has the SATA I for now.... but like Les said they are currently shipping, so, it's just a matter of time... and i think we will se the SATA II in AW's soon

 

on the other hand, les, would you mind telling me.. how many times that green bar passes by on startup with sata II/64 bit.?

tanks =)

 

I have never counted the green bar passing. Is this some new benchmark I am not aware of yet?

 

hmm.. no, i've seen some talk of how many times the green bar passes by before....

but my rationale behind it would be....

1.) green bars appear to pass by at a constant rate on most machines
2.) the time it takes to get past the bios vary from machine to machine, also if it has to go through a respawn option too...
3.) there are inaccuracies with how people time the "start up" of their system - human errors

so i think that one would get a better sense of how fast it actually starts up by how quickly it loads... which the green bars show to a certain degree.(in my opinion), and that takes the human errors of timing, and the bios time away.

 

i just bought the new 100 read / 80 write OCZ from newegg.. can you suggest a free benchmark utility so i can get the whole spectrum of reads/writes vs IO size and ill gladly post it here. without quantifying it, the drive is definately fast with reads and writes...

 

Les:

Any word from MTron on the new 1000 series launching? It seems like this product is becoming vaporware.

Cheers,

 

You have a point. I am in contact with their engineer regarding a question I had but he is not allowed to speak about unreleased product. I was assured that it would be released late March and then early April but it seems they went into a cocoon and i am not sure why. I have put a word in with my contact however we work with her maintaining contact when there is something to promote.

She has assured me of the exclusive and until she has one to send, that will be the last I hear of that. Sorry.

 

Anyone got any recommendations for good express card SSD drives?

 

It depends on the capacity you need, and the speed you need

Actually, most of the so-called "ExpressCard SSDs" use the USB2.0 functionality of the ExpressCard format. That means poor speeds and latency. But it's alright for music or video storage, for example. These are sold by Lexar, Transcend, and other minor brands.
Then, you have some real PCI-Express SSDs. The only ones I know are from Sandisk (costs you an arm and a leg), and the ones sold by futurestorage.co.uk, up to 8Gb

Up to you now

 

Well I just got a XPS 1330 with the new Samsung Sata II 64GB SSD. I had bought the same notebook with the Sata I SSD a few weeks ago, but returned it and got the newer drive. This gave me the chance to compare them side by side.

The new Sata II drive is more than twice as fast. Here are some stats:

PCMark05 HDD - XP Startup: 52MB/s vs. 24MB/s
HDTach: 125MB/s vs. 48MB/s
PCMark05 Score: 7,619 vs. 6,309

There is only one weird issue with the new Sata II drive. When I run HDTach or HDTune there are large blocks where the read speed drops to 85MB/s. For example, from 0GB to 6GB the drive only reads 85MB/s, but then shoots up to 150MB/s from 6GB to 12GB (see image below). Does anyone know why this is happening??

 

Note, the 85MB/s blocks seem to be increasing as I add more data to the hard drive?!? At least that is the only thing I can think of. After loading some programs there are now 4 blocks of the drive that drop to 85MB/s instead of 2 blocks. This probably shouldn't effect the drive, but I have been using the standard Vista defragmenter. Any ideas????

 

I have seen this and observed this through one of the Samsungs. The result, actually is extremely good and neither Samsung nor Dell can understand why it is happening. My response from Samsung was as such...

"what you experienced appears to be the result of where the data was when it was accessed and what type of controller was used – the new RBX is superior to previous controllers."

I, similar to you was receiving absloutely incredible scores with these two programs... I am assured that this shows only a benefit and no concern for the SATA II. I will knock on his door again though and see if Samsung Korea had any further luck addressing the big question...why.

EDIT: I sent along a copy of your HDTach as well if thats ok and can you post the picture of your PCMark05 score as that would make you the highest score ever reached with this machine? I am going to believe yours is newer with the Penryn.

 

Here is one more piece of information that may or may not be related to this issue. I downloaded a tool from the link below that shows information about the drive. There is a "Caution" related to "Reallocated Sectors Count". Does anyone know what this means?


http://crystalmark.info/?lang=en

 

Thanks for the quick reply!! ...and please feel free to send any images to Samsung or Dell or anyone else. Below is an HD Tune image:

Do you defragment your SSDs? I use Diskeeper on my normal hard drives but didn't know if it is worth the time on SSDs or if there could be a negative effect.

 

There is no need to defragment ssds and this has been tackled by authors of all levels al over the net. I will admit that I still do, however, when I run a Disk Cleanup every month or so. I cannot answer why and believe it is unnecessary.

I sent in the DiskMark result and received a response from Samsung that they will probably be back to me early next week.

Your HDTune and HDTach are less than mine which I have attached. This is actually amazing and Samsung is looking at it the same way. My average read is 144MB/s with max at 191MB/s

 

Wow!! Nice speed. Are you using Vista 32bit? I wonder if there are other settings that can effect the speed. I noticed in the device manager that caching was unchecked. Is it checked on your drive?

Can you run the app from the link below and see if receive a "Caution" related to "Reallocated Sectors Count"? Are you using firmware PS105D14?

http://dl.crystaldew.info/CrystalDiskInfo

At the link below you can download and run Crystal Disk Mark. I receive a consistent 100MB/s read with this app so did the guy at this review. This is probably a more accurate and true read speed...HDtune/tach could just be freaking out.

http://crystalmark.info/download/index-e.html

Sorry for all the questions.

 

Ummmm thats my review! (Check my Sig link) eheheheh LMAO. Im using 64bit.

The two sites are owned by the same company.

Oh...and your disk caching will remain unchecked because this is all controlled by the AHCI driver.

 

LOL...small world.
Is the firmware of your drive PS105D14?
Are you using the Microsoft driver from 6/21/2006?

 

T9300 @ 2.5

PCMark05 --> 7,619

http://service.futuremark.com/compare?pcm05=1520433

 

After doing a little research, I think the large chunks of slower performance may be due to "bad blocks".

In the image below I am receiving a caution for Reallocated Sectors Count:

When the hard drive finds a read/write/verification error, it marks this sector as "reallocated" and transfers data to a special reserved area (spare area).
This process is also known as remapping and "reallocated" sectors are called remaps. This is why, on a modern hard disks, you can not see "bad blocks" while testing the surface - all bad blocks are hidden in reallocated sectors. However, the more sectors that are reallocated, the more a sudden decrease (up to 10% and more) can be noticed in the disk read/write speed.

The last part of this definition seems consistent with the performance drop. The drive should read about 100 MB/s but drops to 85MB/s during the large blocks (i.e. suspected bad sectors).

 

What are your readings from benchmarking software other than HDTach and HDTune. These seem to be the only two with the odd scores. What was the result from CrystalMark, ATTO and so on?

If I were to consider the bad sectors theory, I would wonder why the one I was using isn't similar, the low point still being in the 100MB/s range.

I don't consider this as being a drop in any way, but rather spiking and high scoring. I would say the 85 mark is your base read...just a thought. In my experience, I would suggest there is a reason for the low base but would like to first confirm it through other benchmarks.

I would be real curious as to your ATTO result in the read and write.

 

I guess that is possible. However, I would rather the drive read 100MB/s across the board than range from 82-148MB/s.

Can you share a few stats:

• A PCMark05 link
• Firmware of your drive PS105D14?
• Are you using the Microsoft driver from 6/21/2006?
• # of Reallocated Sectors Count via Crystal Info

Below are my Atto and Crystal scores.

 

MTRON SSD FAILURE

Well guys, the thing I dreaded the most (computer wise that is) has happened. My MTRON 32GB SSD failed on me this morning!

I started the computer as usual and XP started to boot when everything froze. When I re-booted, no operating system was found.

I removed the disk from my computer and put it in my Ultrabay HD adapter and used my other computer to inspect it. Disk management under XP showed the drive as un-initialized and all drive space as one contiguous partition.

I can tell so far that the MBR and partition data is gone. I am currently running recovery software to see if anything is recoverable.

I am so frustrated. I have had no problems using the drive for some five months and suddenly it just dies. If my recovery attempts fail I will have lost approx one weeks of work.

Some background on my usage pattern:
I have used the SSD for both work and for my home DVD project. Work usage has minimal impact as I mainly use Office applications and have the page file disabled (I have 3GB of RAM under XP/Office). However, my home DVD project has been somewhat disk intensive. I started to rip and encode my sizeable DVD collection for my upcoming HTPC project. Since December I have ripped and encoded approx 120 DVDs, which corresponds to approx 1 000 GB of data being written to the MTRON SSD.

Initially I held the same view as Les, that SSDs should be fairly safe (considering the claims made by the manufacturers) but now I am not sure anymore. I am not new to computers and have more than 16 years of x86-experience. My XP installation ran both COMODO Firewall Pro as well as McAffe Virus Scan. I have had no virus warnings (for the last 2 – 3 years) and the last piece of software installed was the plug-in Adblock Plus in Firefox a couple of days ago.

While nothing conclusive can be said yet my tentative position is that the MTRON SSD did not fail due to user or software problems. I will be able to say more when the results of my recovery attempts are known.

I will keep you posted.

 

Why would you use the SSD for scratchpad when whatever your discs are stored in won't fit into 32GB. With no additional compression, and nothing on the drive that would amount to < 4 DVD-9 discs. Even with additional compression you wouldn't eke much more than say 10-15 discs onto the SSD. Why not use your final destination disk as your scratchpad?

There are differing levels of experience. Personally, I've been using PCs for more years than you; but my main experience is in *nix platforms. Thankfully the technology fundamentals are quite similar when we're discussing hardware.

At this point, there's insufficient evidence to say it's anything more than a manufacturing defect. If you look at the failure curve you are near the bottom of the initial high rate of failure due to infant mortality. Then as the device nears the end of its anticipated useful lifetime, the rate starts to climb again. This is also referred to as a "bathtub" curve.

MTBF is a guideline and it's arrived at by analyzing a much larger sample size than one. On any single device basis it is an essentially meaningless statistic. There's also MTTF (Mean time to Failure) which is another interesting statistic when talking about many drives.

 

The final destination disk was the 3.5” FAT-X hard drive residing inside my modified Xbox. It is not a very user friendly accessible location and removing it results in a non-operational Xbox until its returned.
Admittedly, my method was not optimal. However, as MTRON rated the write endurance at greater than 50GB per day for 140 years, I did not think more of it. At this stage I cannot say why the hard drive failed so it might have nothing to do with wear endurance.

My point, which I could have expressed more clearly, was that the failure of the disk was not likely due to user error.

My usual luck :-(

 

I don't know, for the cost of an external hard drive (< $100 for 500 GB) that would have been a good investment.

At this point, more likely a manufacturing defect. I sympathize with you on the failure -- been there done that. Just had a machine's motherboard crash two days into production in our data center. It's not fun, but it is what it is.

I get your point -- I feel the same about myself and then I do something "less than optimal"* and realize I make mistakes too. My problem is the mistakes I make are usually bigger mistakes when I do make them.


* A sanitized way of saying stupid

I don't disagree with your being unhappy about it. If the situations were reversed I probably would be too. It's much easier for me to be calm about it when it's not my drive