Welcome Guest ( Log In | Register )

Bump Topic Topic Closed RSS Feed

Outline · [ Standard ] · Linear+

 [HELP] Which brand produced the best M.2 SSD, Price vs Performance vs Setup vs Waranty

views
     
TSDellMalaysia
post Aug 11 2016, 10:04 AM, updated 10y ago

Kindly look at my previous successful threads
*****
Senior Member
735 posts

Joined: Sep 2013
From: IYD ~ In Your Dream



[HELP] Which brand produced the best M.2 SSD?

I already google-ed around and saw plenty chart and graft but different site provide different benchmark result.

If in terms of "Price vs Performance vs Setup vs Warranty", which brand & model is the best?

Please put aside Samsung 950 Pro M.2 SSD because most laptop motherboard not supporting it yet.

Currently I'm using 500GB 850 EVO M.2 SSD (I thought this must be the fastest) on my i7+DDR3 setup laptop.

My friend bought Plextor M.2 SSD and as usual my hand very itchy so I ran benchmark software on my friend i3+DDR4 setup desktop.

I was shock when she got slightly better result than mine. Honestly and seriously I cannot brain at all. Hmmm...


----------


My questions:-
1) Which brand produced the best M.2 SSD?
2) Is DDR3 RAM vs DDR4 RAM give extra credit to the M.2 SSD benchmark?
3) Software booster like RapidMode / PlexTurbo will help the M.2 SSD benchmark?
4) Higher Clock of Processor also will help to get as extra benchmark point?
5) Desktop motherboard vs laptop motherboard, are they on a different league?
6) Share your opinion as M.2 SSD user, how do you fully utilize it?

chaw1979
post Aug 11 2016, 10:19 AM

Getting Started
**
Junior Member
198 posts

Joined: Apr 2006
From: KL



1) currently in the market got M.2 PCIE and M.2 SATA, best one must go for M.2 PCIE SSD...top read speed can reach 2GB/s, write speed 800k/s. For M.2 PCIE SSD i will go for Samsung 950 PRO. i am using it in my Dell XPS13, so far not problem.

2) RAM will not affect SSD performance. or you planning create RAMDisk.

4) CPU will not affect SSD performance benchmark also.

5) Most of desktop motherboard quality definitely better than laptop motherboard. Cos Desktop motherboard got more layers and bigger size.

6) My SSD just use as normal working and some gaming....but fastest SSD first image is reduce bootup time.
horns
post Aug 11 2016, 11:30 AM

\m/
*******
Senior Member
3,038 posts

Joined: Nov 2009
what's the exact model of your friend's plextor ssd? (if it's a pcie type, it will outrun your sata ssd). i can see you are doing casual benchmark (laptop vs. desktop). external factors, like component differences between these units, will definitely affect results also.

1. as of now, samsung imo (pcie, nvme type). however other brands will come in soon.

2. and 4. technically speaking ram and cpu will impact ssd benchmark, e.g. a faster cpu will process things faster (affects round trips of data from and to ssd), and more ram will reduce writes to page file. however, how significant they are to benchmark, not sure, because most people do benchmark of different subjects on the same machine, attempting to rule these out / minimize their impacts to results.

3. yes ram-cache affects results. that's why normally we disable it to get raw speed. note that ram-cache software are not designed equally. that's why they perform differently also. for instance, samsung's rapid mode is relatively much slower than, say, primocache.

Attached Image

the ram in my aw18 is a typical 1600mhz 9-9-9-24 ddr3l only. i believe 2133mhz ddr4 might produce better results in benchmark

5. desktop mobos should be performing better.

6. i use it for work mainly. (e.g. frequent making of different virtual machines for development and tests)

all in all, benchmark is just a reference. the main thing is still about your experience as user. in real world client work loads, i find performance differences are not significant between sata ahci ssd and pcie nvme ssd (especially when you apply ram cache for both). the numbers look good in benchmark, but you actually gain nothing practical. (maybe a relatively faster boot time)

This post has been edited by horns: Aug 11 2016, 03:29 PM
TSDellMalaysia
post Aug 11 2016, 03:55 PM

Kindly look at my previous successful threads
*****
Senior Member
735 posts

Joined: Sep 2013
From: IYD ~ In Your Dream



QUOTE(chaw1979 @ Aug 11 2016, 10:19 AM)
1) currently in the market got M.2 PCIE and M.2 SATA, best one must go for M.2 PCIE SSD...top read speed can reach 2GB/s, write speed 800k/s. For M.2 PCIE SSD i will go for Samsung 950 PRO. i am using it in my Dell XPS13, so far not problem.

2) RAM will not affect SSD performance. or you planning create RAMDisk.

4) CPU will not affect SSD performance benchmark also.

5) Most of desktop motherboard quality definitely better than laptop motherboard. Cos Desktop motherboard got more layers and bigger size.

6) My SSD just use as normal working and some gaming....but fastest SSD first image is reduce bootup time.
*
Is the one for laptop considered as PCIE? its a long M.2 SSD not the short one.

QUOTE(horns @ Aug 11 2016, 11:30 AM)
what's the exact model of your friend's plextor ssd? (if it's a pcie type, it will outrun your sata ssd). i can see you are doing casual benchmark (laptop vs. desktop). external factors, like component differences between these units, will definitely affect results also.

1. as of now, samsung imo (pcie, nvme type). however other brands will come in soon.

2. and 4. technically speaking ram and cpu will impact ssd benchmark, e.g. a faster cpu will process things faster (affects round trips of data from and to ssd), and more ram will reduce writes to page file. however, how significant they are to benchmark, not sure, because most people do benchmark of different subjects on the same machine, attempting to rule these out / minimize their impacts to results.

3. yes ram-cache affects results. that's why normally we disable it to get raw speed. note that ram-cache software are not designed equally. that's why they perform differently also. for instance, samsung's rapid mode is relatively much slower than, say, primocache.

Attached Image

the ram in my aw18 is a typical 1600mhz 9-9-9-24 ddr3l only. i believe 2133mhz ddr4 might produce better results in benchmark

5. desktop mobos should be performing better.

6. i use it for work mainly. (e.g. frequent making of different virtual machines for development and tests)

all in all, benchmark is just a reference. the main thing is still about your experience as user. in real world client work loads, i find performance differences are not significant between sata ahci ssd and pcie nvme ssd (especially when you apply ram cache for both). the numbers look good in benchmark, but you actually gain nothing practical. (maybe a relatively faster boot time)
*
Thanks for your explanation. Primocache is easy to setup? I downloaded it but i have no idea how to configure it.
horns
post Aug 11 2016, 04:15 PM

\m/
*******
Senior Member
3,038 posts

Joined: Nov 2009
QUOTE(DellMalaysia @ Aug 11 2016, 03:55 PM)
Is the one for laptop considered as PCIE? its a long M.2 SSD not the short one.

Thanks for your explanation. Primocache is easy to setup? I downloaded it but i have no idea how to configure it.
*
selective recent models of laptops already have m.2 slot for pcie ssd for some time.

--

yes in general primocache is very easy to set up. unless you play with their advanced features, it's as easy as:
1. create cache tasks;
2. select target disks that you wanna cache;
3. choose 'read' profile;
4. define size of level-1 cache; you can choose to enable defer writes (and define latency, ie. the time to retain in cache before writing into disk, if necessary), or not; then press the start button to initialize and use the cache.

http://www.romexsoftware.com/en-us/primo-c...eate-cache.html


TSDellMalaysia
post Aug 11 2016, 04:18 PM

Kindly look at my previous successful threads
*****
Senior Member
735 posts

Joined: Sep 2013
From: IYD ~ In Your Dream



QUOTE(horns @ Aug 11 2016, 04:15 PM)
selective recent models of laptops already have m.2 slot for pcie ssd for some time.

--

yes in general primocache is very easy to set up. unless you play with their advanced features, it's as easy as:
1. create cache tasks;
2. select target disks that you wanna cache;
3. choose 'read' profile;
4. define size of level-1 cache; you can choose to enable defer writes (and define latency, ie. the time to retain in cache before writing into disk, if necessary), or not; then press the start button to initialize and use the cache.

http://www.romexsoftware.com/en-us/primo-c...eate-cache.html
*
I will follow your step and do the benchmark again. Anyway, mine is like this:-

user posted image

this one not PCIE right?
horns
post Aug 11 2016, 04:25 PM

\m/
*******
Senior Member
3,038 posts

Joined: Nov 2009
QUOTE(DellMalaysia @ Aug 11 2016, 04:18 PM)
I will follow your step and do the benchmark again. Anyway, mine is like this:-

user posted image

this one not PCIE right?
*
ok

no samsung 850 evo is not a pcie ssd. (you can see in the picture it say it's a Serial ATA 6Gb/s on the box) i have 850evo m.2 ssd also.

http://www.samsung.com/semiconductor/minis..._Sheet_Rev2.pdf

This post has been edited by horns: Aug 11 2016, 04:26 PM

Topic ClosedOptions
 

Change to:
| Lo-Fi Version
0.0171sec    1.11    6 queries    GZIP Disabled
Time is now: 20th December 2025 - 05:40 AM