Jump to content


Photo

Plextor PX-M3P SSD Review Discussion


  • You cannot start a new topic
  • Please log in to reply
18 replies to this topic

#1 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,406 posts

Posted 28 March 2012 - 12:35 PM

After the popularity of the Plextor M3S review, we finally got our hands on the new PX-M3P, which showed surprisingly strong performance in a wide range of conditions and a boost in performance with firmware tweaks.

Plextor PX-M3P SSD Review

#2 Beenthere

Beenthere

    Member

  • Member
  • 178 posts

Posted 28 March 2012 - 01:00 PM

Thanks for the review.

It looks like a mixed bag with the M3P. The M3P should be the "standard drive" and Plextor should work on a real Pro drive IMO. With most consumers buying the 128 Gig. drives I think they would be a better choice for review but of course the SSD makers don't want consumers to know the significant drop off in sequential write performance of a 128 vs. a 256 GB. SSD drive.

#3 johnw42

johnw42

    Member

  • Member
  • 66 posts

Posted 28 March 2012 - 01:19 PM

Great job on the steady-state portion of the review!

But there were two major problems, the first is by far the worst:

1) Terrible choice of comparison SSDs. All 4 competitors were nearly equivalent Sandforce SSDs! What were you thinking?

The competitors should have been: Intel 520, one other Sandforce SSD (Vertex3, HyperX, or KC100), Samsung 830, Crucial m4, and Intel 320.

2) Your "real world" tests are flawed, and not at all representative of real world usage. For example, look at the "HTPC disk capture" test. Video files are a perfect example of incompressible data. And yet your Sandforce SSDs report rates of 450MB/s, which is close to the easily-compressible sequential write rates you measured with IOMeter of 460MB/s, and nowhere near the incompressible sequential write rates of less than 300MB/s you measured with IOMeter. Clearly your "real world" test is writing UNrealistic highly compressible files instead of real video files.

Edited by johnw42, 28 March 2012 - 01:21 PM.

#4 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,406 posts

Posted 28 March 2012 - 02:29 PM

Great job on the steady-state portion of the review!

But there were two major problems, the first is by far the worst:

1) Terrible choice of comparison SSDs. All 4 competitors were nearly equivalent Sandforce SSDs! What were you thinking?

The competitors should have been: Intel 520, one other Sandforce SSD (Vertex3, HyperX, or KC100), Samsung 830, Crucial m4, and Intel 320.

2) Your "real world" tests are flawed, and not at all representative of real world usage. For example, look at the "HTPC disk capture" test. Video files are a perfect example of incompressible data. And yet your Sandforce SSDs report rates of 450MB/s, which is close to the easily-compressible sequential write rates you measured with IOMeter of 460MB/s, and nowhere near the incompressible sequential write rates of less than 300MB/s you measured with IOMeter. Clearly your "real world" test is writing UNrealistic highly compressible files instead of real video files.


We are ever so close to getting a dynamic chart generation system in progress (actually using it manually now) and with that in place it will be easier to handle comparables. This we used the most current set with the M3S and M3P added on.

On the real-world trace aspect, it is still working with direct LBA addressing, the next iteration can apply specific datatypes (fully incompressible/compressible). With it requiring a full retest of every drive to be meaningful at launch, we will be introducing it on consumer drives when we launch out next testing platform that is in validation now. Direct LBA interaction is still a step forward from generic I/O from a synthetic benchmark... and addressing the newer SSD controllers is a step we already have in place ready to be rolled out.

Our main goal right now is to provide a wide range of benchmarks so that if one area might be weaker than others, another testing segment can be more relevant for a given workload. Its really coming down to what the best stuff we have on hand at a given time during a review. Some of the behind the scenes stuff is growing the lab and improving some of the same tests that are in need of an overhaul. So you can be assured that we are taking note and looking at all new ways of testing B).

#5 Octoploid

Octoploid

    Member

  • Member
  • 5 posts

Posted 29 March 2012 - 04:50 AM

Is there a chance to transform a M3S into a M3P by somehow forcing the M3P firmware onto a M3S?

#6 Brian

Brian

    SR Admin

  • Admin
  • 5,135 posts

Posted 29 March 2012 - 07:46 AM

There is not a firmware upgrade path for the M3S to a M3P.

Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#7 UranusFX

UranusFX

    Member

  • Member
  • 1 posts

Posted 29 March 2012 - 06:26 PM

Had been waiting for a review, very interesting!

But M3S and M3P have been swapped in the data of Power Consulption?

#8 Werner Juvik

Werner Juvik

    Member

  • Member
  • 1 posts

Posted 30 March 2012 - 02:50 AM

We are ever so close to getting a dynamic chart generation system in progress (actually using it manually now) and with that in place it will be easier to handle comparables. This we used the most current set with the M3S and M3P added on.

On the real-world trace aspect, it is still working with direct LBA addressing, the next iteration can apply specific datatypes (fully incompressible/compressible). With it requiring a full retest of every drive to be meaningful at launch, we will be introducing it on consumer drives when we launch out next testing platform that is in validation now. Direct LBA interaction is still a step forward from generic I/O from a synthetic benchmark... and addressing the newer SSD controllers is a step we already have in place ready to be rolled out.

Our main goal right now is to provide a wide range of benchmarks so that if one area might be weaker than others, another testing segment can be more relevant for a given workload. Its really coming down to what the best stuff we have on hand at a given time during a review. Some of the behind the scenes stuff is growing the lab and improving some of the same tests that are in need of an overhaul. So you can be assured that we are taking note and looking at all new ways of testing B).


Thanks for a fine review.

Is this a drive thatīs particularly suitable for video- and raw photography editing working in Macbook Pro environment, or should I be looking elsewhere? There are many nice reviews out there, but I do miss clear guidance from the experts on which drive suits best the different tasks. If reviews could include recommendations intended for those of us with no time or ability to consume all quirks of tech around, it would be a tremendous help in navigating for a high performance work platform. I do think the typical video and photog guy, as you now know concerns me, concentrate on the task, and halt other processes like games. Most likely they would by far prefer an SSD performing better with video editing than games or other applications. I do see you have attempted to bring it down to real life, and maybe itīs asking too much to get recommendations for specific professions. But it would be greatly appreciated and infinitely valuable if it could be done.

Edited by Werner Juvik, 30 March 2012 - 02:51 AM.

#9 johnw42

johnw42

    Member

  • Member
  • 66 posts

Posted 30 March 2012 - 10:35 PM

Is this a drive thatīs particularly suitable for video- and raw photography editing working in Macbook Pro environment, or should I be looking elsewhere?


In my limited experience, an important quality for video editing is high sequential write speed for incompressible data.

I think the Plextor M3P has the highest incompressible data sequential write speed of any consumer SSD, so in that sense it should be a good choice for video editing.

I don't have any experience with SSDs in a Mac environment, so I cannot comment on that aspect of your question.

#10 gandhi

gandhi

    Member

  • Member
  • 11 posts

Posted 04 April 2012 - 02:24 PM

Thanks for the review and in particular for adding steady-state figures in many of the recent reviews - this is very much appreciated and not found elsewhere. This is great help when trying to decide what drive to use in scenarios where TRIM is not an option like ESXi, for example.

Still curious if one could cross-flash M3P firmware on M3 ...

#11 Beenthere

Beenthere

    Member

  • Member
  • 178 posts

Posted 04 April 2012 - 02:55 PM

I believe it was noted in another thread that the reported Plextor M3 drive performance is NOT as good on Macs so it's worth contacting Plextor to get the stated Mac performance data, I would think.

#12 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,406 posts

Posted 04 April 2012 - 03:29 PM

We are currently looking to add a MacBook Pro to the test lab for these types of tests, so we wont have to rely on manufacturer info to supplement reviews. This probably won't happen until the next refresh, but expect to see that scope added to our reviews.

#13 mike2h

mike2h

    Member

  • Member
  • 258 posts

Posted 04 April 2012 - 05:28 PM

In my limited experience, an important quality for video editing is high sequential write speed for incompressible data.

I think the Plextor M3P has the highest incompressible data sequential write speed of any consumer SSD, so in that sense it should be a good choice for video editing.

I don't have any experience with SSDs in a Mac environment, so I cannot comment on that aspect of your question.

does this apply to games? i dont know which category(compres/incompres) games fall into.

Edited by mike2h, 05 April 2012 - 12:35 AM.

#14 Surlias

Surlias

    Member

  • Member
  • 4 posts

Posted 06 April 2012 - 05:34 PM

After the popularity of the Plextor M3S review, we finally got our hands on the new PX-M3P, which showed surprisingly strong performance in a wide range of conditions and a boost in performance with firmware tweaks.

Plextor PX-M3P SSD Review


Am I correct in my understanding that as of the time of publication of this review, you have declared the PX-M3P the best client SSD on the market? Better even than the Intel 520?

If so, what does this mean in terms of the user experience for someone who primarily uses their computer for productivity and gaming? Is it true that I would never notice a difference between, say, the M3S, the M3P, or the Intel 520 in such a setting?

Edited by Surlias, 06 April 2012 - 07:03 PM.

#15 mike2h

mike2h

    Member

  • Member
  • 258 posts

Posted 07 April 2012 - 12:44 AM

Am I correct in my understanding that as of the time of publication of this review, you have declared the PX-M3P the best client SSD on the market? Better even than the Intel 520?

If so, what does this mean in terms of the user experience for someone who primarily uses their computer for productivity and gaming? Is it true that I would never notice a difference between, say, the M3S, the M3P, or the Intel 520 in such a setting?

not sure where u got the 'declaration' from but i would bet the answer to your question would be yes.

#16 Brian

Brian

    SR Admin

  • Admin
  • 5,135 posts

Posted 07 April 2012 - 08:16 AM

It's very good, but certain workloads benefit from the SF drives. We don't ever call one thing the definitive and only ;)

Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#17 lunadesign

lunadesign

    Member

  • Member
  • 139 posts

Posted 10 May 2012 - 01:50 PM

How would this SSD compare to the Crucial m4?

Also, I'm having a hard time understanding the steady state performance. Is your steady state scenario on the "torture test" end of scale or is it a more moderate usage pattern? Does GC get a chance to kick in or does the drive have to be completely idle for GC to kick in? (I've never quite completely understood that.)

I'm curious because I'd like to use this SSD (or maybe an m4 since they're cheaper and easier to find) in an entry-to-mid ESXi server as a single drive or possibly two in RAID 1. In either case, I wouldn't have the benefit of TRIM. I saw a review on Xbit that showed how the M3S GC recovered performance nicely after 30 mins so I'd assume the M3P would be a bit better. However, I also saw a review on Tom's Hardware that claimed that random writes will eventually trash performance on Marvell based SSDs like this one.

My ESXi server's going to have a number of OSes on 24/7 but not heavily used so I'm trying to understand if the lack of TRIM means that an SSD will degrade over time to the point of being worse than, say, a Velociraptor.

#18 lunadesign

lunadesign

    Member

  • Member
  • 139 posts

Posted 13 June 2012 - 02:24 AM

I just noticed a typo in the article - in the spec section at the top, the random write spec for the 256 GB model is missing a digit. According to the Plextor site, its 68,000.

BTW, why would the 512 GB have substantially lower random read/write specs? I thought the biggest drives always had the best specs...

#19 Brian

Brian

    SR Admin

  • Admin
  • 5,135 posts

Posted 13 June 2012 - 02:41 PM

It's about NAND die layout, lanes, etc.

http://www.storagere...t_family_review

Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users