M1 NVMe compatibility reports

zenith
Posts: 1
Joined: Sat Apr 30, 2022 3:54 am
languages_spoken: english
ODROIDs: M1
Has thanked: 0
Been thanked: 1 time
Contact:

Re: M1 NVMe compatibility reports

Post by zenith »

1. Brand : PNY
2. Model : XLR8 CS3030 2TB
3. Capacity : 2TB
4. NVMe controller and firmware version : Phison PS5012-E12
5. Additional comments : performance (fio and iozone3 benchmark) or any other notes
Tested on Ubuntu 20.04.4 with kernel: `Linux server 4.19.219-odroid-arm64 #1 SMP Wed, 11 May 2022 13:19:00 +0000 aarch64 aarch64 aarch64 GNU/Linux`

$ sudo fio --direct=1 --rw=read --bs=1024k --ioengine=libaio --iodepth=64 --size=4G --runtime=10 --numjobs=4 --group_reporting --name=fio-rand-read-sequential --eta-newline=1 --filename=/dev/nvme0n1 --readonly | grep READ

Code: Select all

FIO: READ: bw=1572MiB/s (1648MB/s), 1572MiB/s-1572MiB/s (1648MB/s-1648MB/s), io=15.6GiB (16.7GB), run=10147-10147msec
$ sudo iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2

Code: Select all

Iozone: Performance Test of File I/O
	        Version $Revision: 3.489 $
		Compiled for 64 bit mode.
		Build: linux 

	Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
	             Al Slater, Scott Rhine, Mike Wisner, Ken Goss
	             Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
	             Randy Dunlap, Mark Montague, Dan Million, Gavin Brebner,
	             Jean-Marc Zucconi, Jeff Blomberg, Benny Halevy, Dave Boone,
	             Erik Habbinga, Kris Strecker, Walter Wong, Joshua Root,
	             Fabrice Bacchella, Zhenghua Xue, Qin Li, Darren Sawyer,
	             Vangel Bojaxhi, Ben England, Vikentsi Lapa,
	             Alexey Skidanov, Sudhir Kumar.

	Run began: Thu May 12 14:40:53 2022

	Include fsync in write timing
	O_DIRECT feature enabled
	Auto Mode
	File size set to 102400 kB
	Record Size 4 kB
	Record Size 16 kB
	Record Size 512 kB
	Record Size 1024 kB
	Record Size 16384 kB
	Command line used: iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
	Output is in kBytes/sec
	Time Resolution = 0.000001 seconds.
	Processor cache size set to 1024 kBytes.
	Processor cache line size set to 32 bytes.
	File stride size set to 17 * record size.
                                                              random    random     bkwd    record    stride                                    
              kB  reclen    write  rewrite    read    reread    read     write     read   rewrite      read   fwrite frewrite    fread  freread
          102400       4    39809    85742   100184   100179    38249    97105                                                                
          102400      16   134263   303597   316060   260833    73160   299379                                                                
          102400     512   991335  1075799   769769   780755   779122  1042951                                                                
          102400    1024  1138841  1186298  1072487  1092522   983424  1181640                                                                
          102400   16384  1425745  1436479  1306003  1342527  1320509  1423933                                                                
6. Price and where to buy : ~ 190 euro in (Dutch) webshops, https://tweakers.net/pricewatch/1391498 ... 0-2tb.html
These users thanked the author zenith for the post:
tobetter (Thu May 12, 2022 9:53 pm)

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: M1 NVMe compatibility reports

Post by mctom »

I'm always very happy to see new users contributing to the community straight away. :) Thank you!

I edited my original post so nobody has to look for instructions for too long.
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

sledge42
Posts: 13
Joined: Sun Sep 13, 2020 10:46 pm
languages_spoken: english, german
ODROIDs: XU4Q, C2, M1
Has thanked: 3 times
Been thanked: 0
Contact:

Re: M1 NVMe compatibility reports

Post by sledge42 »

1. Brand : Samsung
2. Model : 980 MZ-V8V1T0BW (Samsung 980 - 1000 GB - M.2 - 3500 MB/s)
3. Capacity : 1 TB
4. NVMe controller and firmware version : (if available)
5. Additional comments : performance (fio and iozone3 benchmark) or any other notes
6. Price and where to buy : CHF 99 @ https://superstore.foletti.com/de/produ ... 2-3500-mbs

Code: Select all

root@server:/home/odroid# sudo fio --direct=1 --rw=read --bs=1024k --ioengine=libaio --iodepth=64 --size=4G --runtime=10 --numjobs=4 --group_reporting --name=fio-rand-read-sequential --eta-newline=1 --readonly --filename=/dev/nvme0n1 | grep READ
   READ: bw=1568MiB/s (1644MB/s), 1568MiB/s-1568MiB/s (1644MB/s-1644MB/s), io=15.4GiB (16.5GB), run=10041-10041msec

Code: Select all

root@server:/home/odroid# sudo iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
        Iozone: Performance Test of File I/O
                Version $Revision: 3.489 $
                Compiled for 64 bit mode.
                Build: linux

        Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
                     Al Slater, Scott Rhine, Mike Wisner, Ken Goss
                     Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
                     Randy Dunlap, Mark Montague, Dan Million, Gavin Brebner,
                     Jean-Marc Zucconi, Jeff Blomberg, Benny Halevy, Dave Boone,
                     Erik Habbinga, Kris Strecker, Walter Wong, Joshua Root,
                     Fabrice Bacchella, Zhenghua Xue, Qin Li, Darren Sawyer,
                     Vangel Bojaxhi, Ben England, Vikentsi Lapa,
                     Alexey Skidanov, Sudhir Kumar.

        Run began: Sat May 14 13:36:26 2022

        Include fsync in write timing
        O_DIRECT feature enabled
        Auto Mode
        File size set to 102400 kB
        Record Size 4 kB
        Record Size 16 kB
        Record Size 512 kB
        Record Size 1024 kB
        Record Size 16384 kB
        Command line used: iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
        Output is in kBytes/sec
        Time Resolution = 0.000001 seconds.
        Processor cache size set to 1024 kBytes.
        Processor cache line size set to 32 bytes.
        File stride size set to 17 * record size.
                                                              random    random     bkwd    record    stride
              kB  reclen    write  rewrite    read    reread    read     write     read   rewrite      read   fwrite frewrite    fread  freread
          102400       4    39855    85339    94160    92610    40167    89127
          102400      16   121876   259635   291026   287282   163463   272542
          102400     512  1005944  1079257   864032   869048   867546  1084930
          102400    1024  1116710  1117195  1024194  1025355  1017779  1206865
          102400   16384  1414549  1443161  1375922  1370648  1376533  1438559

iozone test complete.

drolid
Posts: 3
Joined: Fri May 06, 2022 3:52 am
languages_spoken: english
ODROIDs: M1
Has thanked: 2 times
Been thanked: 0
Contact:

Re: M1 NVMe compatibility reports

Post by drolid »

1. Brand : Samsung
2. Model : Samsung SSD 980 PRO 2TB (MZ-V8P2T0)
3. Capacity : 2 TB
4. Firmware: 3B2QGXA7

Code: Select all

root@server:~# fio --direct=1 --rw=read --bs=1024k --ioengine=libaio --iodepth=64 --size=4G --runtime=10 --numjobs=4 --group_reporting --name=fio-rand-read-sequential --eta-newline=1 --readonly --filename=/dev/nvme0n1 | grep READ
   READ: bw=1565MiB/s (1641MB/s), 1565MiB/s-1565MiB/s (1641MB/s-1641MB/s), io=15.4GiB (16.5GB), run=10079-10079msec

Code: Select all

root@server:/mnt/nvme# iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
        Iozone: Performance Test of File I/O
                Version $Revision: 3.489 $
                Compiled for 64 bit mode.
                Build: linux

        Contributors:William Norcott, Don Capps, Isom Crawford, Kirby Collins
                     Al Slater, Scott Rhine, Mike Wisner, Ken Goss
                     Steve Landherr, Brad Smith, Mark Kelly, Dr. Alain CYR,
                     Randy Dunlap, Mark Montague, Dan Million, Gavin Brebner,
                     Jean-Marc Zucconi, Jeff Blomberg, Benny Halevy, Dave Boone,
                     Erik Habbinga, Kris Strecker, Walter Wong, Joshua Root,
                     Fabrice Bacchella, Zhenghua Xue, Qin Li, Darren Sawyer,
                     Vangel Bojaxhi, Ben England, Vikentsi Lapa,
                     Alexey Skidanov, Sudhir Kumar.

        Run began: Sat May 14 20:00:29 2022

        Include fsync in write timing
        O_DIRECT feature enabled
        Auto Mode
        File size set to 102400 kB
        Record Size 4 kB
        Record Size 16 kB
        Record Size 512 kB
        Record Size 1024 kB
        Record Size 16384 kB
        Command line used: iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
        Output is in kBytes/sec
        Time Resolution = 0.000001 seconds.
        Processor cache size set to 1024 kBytes.
        Processor cache line size set to 32 bytes.
        File stride size set to 17 * record size.
                                                              random    random     bkwd    record    stride
              kB  reclen    write  rewrite    read    reread    read     write     read   rewrite      read   fwrite frewrite    fread  freread
          102400       4    41022    91175    97803    99937    55352    89838
          102400      16   150675   281334   297232   302496   154780   277238
          102400     512   929294  1011768   921153   982205   923512  1031333
          102400    1024  1082385  1144263  1026875  1096172  1064803  1043281
          102400   16384  1337636  1361892  1262653  1381199  1362779  1358975

Code: Select all

root@server:~# time ( dd if=/dev/nvme0n1 of=/dev/zero bs=1M count=20000 status=progress )
20465057792 bytes (20 GB, 19 GiB) copied, 31 s, 660 MB/s
20000+0 records in
20000+0 records out
20971520000 bytes (21 GB, 20 GiB) copied, 31.792 s, 660 MB/s

real    0m31.800s

tkaiser
Posts: 752
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1
Has thanked: 0
Been thanked: 20 times
Contact:

Re: M1 NVMe compatibility reports

Post by tkaiser »

drolid wrote:
Sun May 15, 2022 5:10 am

Code: Select all

   READ: bw=1565MiB/s (1641MB/s), 1565MiB/s-1565MiB/s (1641MB/s-1641MB/s), io=15.4GiB (16.5GB), run=10079-10079msec
          102400   16384  1337636  1361892  1262653  1381199  1362779  1358975
20971520000 bytes (21 GB, 20 GiB) copied, 31.792 s, 660 MB/s
3 'benchmarks' fired up, 3 totally different numbers generated (1640 MB/s, 1360 MB/s and 660 MB/s). Which is the right one? Is there even a right one? Do these numbers have any meaning?

drolid
Posts: 3
Joined: Fri May 06, 2022 3:52 am
languages_spoken: english
ODROIDs: M1
Has thanked: 2 times
Been thanked: 0
Contact:

Re: M1 NVMe compatibility reports

Post by drolid »

tkaiser wrote:
Fri May 20, 2022 3:38 pm
drolid wrote:
Sun May 15, 2022 5:10 am

Code: Select all

   READ: bw=1565MiB/s (1641MB/s), 1565MiB/s-1565MiB/s (1641MB/s-1641MB/s), io=15.4GiB (16.5GB), run=10079-10079msec
          102400   16384  1337636  1361892  1262653  1381199  1362779  1358975
20971520000 bytes (21 GB, 20 GiB) copied, 31.792 s, 660 MB/s
3 'benchmarks' fired up, 3 totally different numbers generated (1640 MB/s, 1360 MB/s and 660 MB/s). Which is the right one? Is there even a right one? Do these numbers have any meaning?
For the given workload, they are all "the right one's" :D

1 dd ~ 700MB/s , 3 dd's ~ 1130

Image

To have this kind of performance on a 100$ SBC with 8GB RAM is insane.

Post Reply

Return to “Hardware and peripherals”

Who is online

Users browsing this forum: No registered users and 1 guest