Plextor M5 Pro: an other step forward

[Ao1]

Sorry no, but I've seen 30 MB/s plus for 4K RR. This is one of the fastest drives I have played with.

[kaktus1907]

For now M5P's been a let down for me, not much improvements over M3P.. I was expecting ~40MB/s 4K reads, waiting for more reviews for clearer opinion..

[Highendtoys]

Hi Highendtoys. How did you create the files in your TRIM test?

I have to run, I have youth football to coach tonight but you can find the TRIM / fill testing info here:

http://www.tweaktown.com/articles/39...ned/index.html

[Highendtoys]

Also, as I said in the review, this drive has something going on with TRIM but I'm not sure what it is yet. There was an NDA launch and the drive just arrived at 3PM the day before. It's not my best article for sure but I've been working the drive all morning to get some of the other details worked out.

[Ao1]

I have to run, I have youth football to coach tonight but you can find the TRIM / fill testing info here:

http://www.tweaktown.com/articles/39...ned/index.html

So the drive is 75% full. You then delete all the files and re-run the benchmark? If so how long are you waiting between deleting the files and starting the benchmark? If you run it straight away try waiting for two minutes or so. You should find performance has bounced right back to new.

[Highendtoys]

Yes, there is a delay put in between each test and each file load to the drive.

[Ao1]

That's not the M5 Pro.

[johnw]

That's not the M5 Pro.

Right. It appears to have Micron flash and the 9174 controller. Probably the M5S.

[kaktus1907]

Right. It appears to have Micron flash and the 9174 controller. Probably the M5S.

yeah url confirms that

http://www.gdm.or.jp/review/ssd/plextor/m5s/image/08_1024x768.jpg
http://www.gdm.or.jp/review/ssd/plextor/m5s/image/07_1024x768.jpg

[dctokyo]

That's not the M5 Pro.

sorry I posted the wrong links

[Marc HFR]

Also, as I said in the review, this drive has something going on with TRIM but I'm not sure what it is yet. There was an NDA launch and the drive just arrived at 3PM the day before. It's not my best article for sure but I've been working the drive all morning to get some of the other details worked out.

The fact that the drive arrived just before the NDA is your problem because it's your choice to publish the review at the NDA date and not a day or two after ;-)

[Marc HFR]

I don't think there is anything wrong with TRIM on the M5Pro. If you monitor with hIOmon you can see TRIM commands being executed as files get deleted. Also I just created a test file with Iometer that filled the drive. I then ran 4K random writes for half an hour or so. I then deleted the test file, waited a couple of minutes and ran AS SSD.

Attachment 129166

4K reads seem to have taken a bit of a hit, but not bad considering what the drive had just been exposed too.

The drive also uses idle time GC btw.

Hi

I've just run my stress test ( this one : http://www.behardware.com/articles/8...a-6g-ssds.html ) on M5P and it reacts like M3P so i see no problem on TRIM.

[idolclub]

M5 Pro 256GB benchmarks:

[johnw]

I've just run my stress test ( this one : http://www.behardware.com/articles/8...a-6g-ssds.html ) on M5P and it reacts like M3P so i see no problem on TRIM.

It is good to know that you are testing the M5P. behardware's SSD test results are always interesting to see, especially the tests where you fill the SSDs with data and stress them before measuring sequential and random write speeds. I'll be looking forward to seeing the M5P results on behardware!

[Marc HFR]

For the moment i'm going crazy trying to understand Vertex 4 / Agility 4 performance/standard/very-low write performance

[johnw]

For the moment i'm going crazy trying to understand Vertex 4 / Agility 4 performance/standard/very-low write performance

I've seen lots of weird results on the Vertex 4. Even just looking at AS-SSD results with fw1.5 that are posted on the internet, it seems that the sequential read speed can vary from low 400MB/s up to 520MB/s. And most of those are (I assume) with a nearly empty SSD. It seems like the V4 firmware likes to do some really aggressive housekeeping at random times, and this can affect the performance measured.

Even your behardware link to the Vertex 4 testing (with pre-1.5 firmware) shows the Random write "new" speed jumping around a lot.

I guess the V4 is just "twitchy".

[Zaxx]

+1...Ocz still has some work to do regarding V4 FW imo. There's a good chance that the majority of their FW engineers are up to their eyeballs with Ocz's first in-house (and very promising) controller...the Indilinx Barefoot 3. Looking forward to a review that thoroughly compares the marvell based V4 to the M5P.

[Callsign_Vega]

Any more info on the M5P? Plextor made an announcement and there hasn't been anything since as far as availability and reviews besides the TT review. Paper launch?

[UranusFX]

Any more info on the M5P? Plextor made an announcement and there hasn't been anything since as far as availability and reviews besides the TT review. Paper launch?

What else do you need to know specifically?

Personally, Expect trim/power/workload that anand's review.

[Zaxx]

Plextor might be a bit slow with initial production if the 19nm tosh toggle hasn't ramped up sufficiently...just a guess tho.

[dctokyo]

Any more info on the M5P? Plextor made an announcement and there hasn't been anything since as far as availability and reviews besides the TT review. Paper launch?

Paper launch? I think not, you can find them in any computer shop here in Japan.

[Highendtoys]

They are on a boat from Japan to the US most likely. They might of been held up at customs for a few days or any number of things could of happened. The 19nm ramp up hasn't been all that fast and remember that Toshiba cut their side of production 30% too.

[Highendtoys]

Hi

I've just run my stress test ( this one : http://www.behardware.com/articles/8...a-6g-ssds.html ) on M5P and it reacts like M3P so i see no problem on TRIM.

I have data that says different and even know why TRIM isn't running at 100%. I guess you'll have to wait for the follow up article.

[Zaxx]

They are on a boat from Japan to the US most likely. They might of been held up at customs for a few days or any number of things could of happened. The 19nm ramp up hasn't been all that fast and remember that Toshiba cut their side of production 30% too.

Damn...I forgot about the production cut. Looks like consumers won't be seeing lower prices usually associated with a die shrink. I wonder if Intel is gonna pull something similar with it's upcoming 20nm IMFT. That would really suck.

[Highendtoys]

I will know more about 20nm this time next week when I meet with some people closer to the action. When I saw IMFT 20nm at Computex and played with a working drive you could tell that it was a lot slower than 19nm. At Computex 19nm was ready to go and we have proof of that now with the P5 Pro.

I wouldn't be surprised if we don't see 20nm until 2013....but they are still saying 2012 as of right now.