S922X Ugoos AM6B Device Tree - Performance/Efficiency - Testing Needed

Help & Support Ugoos
284

My old AM6plus s922xj without last firmware 0.5.4n (only normal 0.5.4 firmware for device before 2021 - notB, not updated to nightly CE and used inovative DTB from MasterKeyxda)

CoreELEC (official): 21.1.1-Omega (Amlogic-ng.arm)
Machine model: UGOOS AM6
CoreELEC dt-id: g12b_s922x_ugoos_am6
Amlogic dt-id: g12b_AM6_b4g
Linux version: 4.9.269 (portisch@ubuntu) #1 Mon Aug 26 17:08:18 CEST 2024
Kodi compiled: 2024-08-26 17:21:12 +0200

CoreELEC:~ # cat /sys/class/aml_ddr/freq
752 MHz

and after updated to nightly CE:

CoreELEC (official): 21.1.1-Omega_nightly_20240928 (Amlogic-ng.arm)
Machine model: UGOOS AM6
CoreELEC dt-id: g12b_s922x_ugoos_am6
Amlogic dt-id: g12b_AM6_b4g
Linux version: 4.9.269 (docker@5628672ce231) #1 Sat Sep 28 06:01:52 IDT 2024
Kodi compiled: 2024-09-28 05:42:50 +0200

CoreELEC:~ # cat /sys/class/aml_ddr/freq
752 MHz

CoreELEC:~ # lscpu
Architecture: aarch64
Byte Order: Little Endian
CPU(s): 6
On-line CPU(s) list: 0-5
Vendor ID: ARM
Model name: Cortex-A53
Model: 4
Thread(s) per core: 1
Core(s) per socket: 2
Socket(s): 1
Stepping: r0p4
CPU(s) scaling MHz: 37%
CPU max MHz: 1800.0000
CPU min MHz: 500.0000
BogoMIPS: 48.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32
Model name: Cortex-A73
Model: 2
Thread(s) per core: 1
Core(s) per socket: 4
Socket(s): 1
Stepping: r0p2
CPU(s) scaling MHz: 63%
CPU max MHz: 2208.0000
CPU min MHz: 500.0000
BogoMIPS: 48.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32
Caches (sum of all):
L1d: 192 KiB (6 instances)
L1i: 320 KiB (6 instances)
L2: 1.3 MiB (2 instances)

285

It would be better ask for somebody with Minix u22xj, if checked physically (or have photos) under fan - on minix is uncovering of cpu easy and safe. Ugoos used very stupid holder for cpu fan, so i dont hate courage for next step for uncover cpu.

2024-09-06-20-08-27-161
2024-09-06-20-08-27-1611920Ɨ3413 426 KB

2024-09-06-19-53-08-827
2024-09-06-19-53-08-8271920Ɨ1080 206 KB

1 Like
286

my version is
21.1-Omega_nightly_20240527 and is installed on an external SSD

is it best to install on the internal for these benefits that have been added or will the external still have the improvements

any problems with the latest over the version i have as i have not upgraded in a while ?

thanks

287

Cache Optimization for S922 is already included in the latest nightlies.
No difference if ext or int. install

1 Like
288

I have installed yesterdayā€™s nightly (latest available) on my Ugoos AM6b+.
Canā€™t tell if it feels snappier or not, to be honest. Video library update seemed considerably faster and probably even boot time was.

Anything specific I could/should test (as per discussion title)?

Also, while lscpu states that MAXMHZ is 1800 for A53 and 2208 for A73, in dispinfo I get

max_freq_a73='1800'
max_freq_a53='1896'

Is this normal?

289

Nope. Have you set these values in Config.ini?
Which nightly are you currently running?

290

am6b+

cat /sys/class/aml_ddr/freq
752 MHz
291

EDIT
Since Iā€™m not familiar with the syntax used by dispinfo, I might be wrong but maybe those two entries are effectively remmed out and do not act on the machine? See below.
/EDIT

Hereā€™s my dispinfo: https://paste.coreelec.org/DittoIvory
21.1.1-Omega_nightly_20240928

Here lscpu:


Architecture:           aarch64
  Byte Order:           Little Endian
CPU(s):                 6
  On-line CPU(s) list:  0-5
Vendor ID:              ARM
  Model name:           Cortex-A53
    Model:              4
    Thread(s) per core: 1
    Core(s) per socket: 2
    Socket(s):          1
    Stepping:           r0p4
    CPU(s) scaling MHz: 100%
    CPU max MHz:        1800.0000
    CPU min MHz:        500.0000
    BogoMIPS:           48.00
    Flags:              fp asimd evtstrm aes pmull sha1 sha2 crc32
  Model name:           Cortex-A73
    Model:              2
    Thread(s) per core: 1
    Core(s) per socket: 4
    Socket(s):          1
    Stepping:           r0p2
    CPU(s) scaling MHz: 100%
    CPU max MHz:        2208.0000
    CPU min MHz:        500.0000
    BogoMIPS:           48.00
    Flags:              fp asimd evtstrm aes pmull sha1 sha2 crc32
Caches (sum of all):    
  L1d:                  192 KiB (6 instances)
  L1i:                  320 KiB (6 instances)
  L2:                   1.3 MiB (2 instances)

And lscpu --output-all -e:

BOGOMIPS CPU CORE SOCKET CLUSTER NODE BOOK DRAWER L1d:L1i:L2 POLARIZATION ADDRESS CONFIGURED ONLINE       MHZ SCALMHZ%    MAXMHZ   MINMHZ MODELNAME
   48.00   0    0      0       -    -    -      - 0:0:0      -            -       -             yes 1800.0000     100% 1800.0000 500.0000 Cortex-A53
   48.00   1    1      0       -    -    -      - 1:1:0      -            -       -             yes 1800.0000     100% 1800.0000 500.0000 Cortex-A53
   48.00   2    0      0       -    -    -      - 2:2:1      -            -       -             yes 2208.0000     100% 2208.0000 500.0000 Cortex-A73
   48.00   3    1      0       -    -    -      - 3:3:1      -            -       -             yes 2208.0000     100% 2208.0000 500.0000 Cortex-A73
   48.00   4    2      0       -    -    -      - 4:4:1      -            -       -             yes 2208.0000     100% 2208.0000 500.0000 Cortex-A73
   48.00   5    3      0       -    -    -      - 5:5:1      -            -       -             yes 2208.0000     100% 2208.0000 500.0000 Cortex-A73
292

Yes, lines starting with # are comments, not effective. Ignore it.

1 Like
293

Personally I want to cry :sob:. Ugoos may not have had a choice of DDR because the old ones were EOL. Heck even the DDR4 on your and my ugoos is EOL now so they had to find new DDR and then release a new firmware.

294

Were you able to quantify any differences with lmbench, with & without the L2 cache changes made to the DTB?

295

Newer boxes not showing much higher speeds

Vim4 1008
AM8 950

296

No lmbench for Ram did not show any improvement. It uses memcpy which is already written in assembly for armv8.

Lmbench used 100% CPU of 1 core. So no expectations of improvement.

Do you know of any benchmark names for web browsing and UI navigation? I recall there was something that was popular around MacBook M1. Maybe we can use that one.

297

I donā€™t know of anything for UI navigation, I donā€™t know if there is too much variation in UIā€™s to make a general tool to test that. Cachebench is another cache specific benchmarking tool that might be useful.

298

Is cache bench available on entware?

299

I donā€™t know, I was looking for a compiled arm binary, but hadnā€™t had any luck.

300

I see Youtubers, use this web browsing test all the time. not sure if it could be of help
https://browserbench.org/Speedometer2.0/

301

My box came with 054n
Same results as yours
752 Mhz

Maybe that is for 1 gig if Ram avg
Multiply X 4
You get over 3000 Mhz

Ha ha ha

Never need Ram to be faster than CPU max 2208

302

of course DDR4 speed is not at that low speed lowest DDR4 is 2133MHz, so no way these run under that.

303

DDR is double data rate no? So itā€™s base clock x2.