qblueRed42 reacted to NicoD in Better video playback with Vivaldi Browser - for all arm sbc's
I've discovered Vivaldi Browser for arm. A fork of Chromium.
There is a armhf version and a arm64 version.
Youtube playback with this is a lot better. I've tested it on the NanoPi M4.
The same video in Chromium had 2/3 dropped frames. (10 frames/s) 1080p video
With Vivaldi browser you get 1/3 dropped frames. (20 frames/s)
A lot better experience.
Here you can download it.
Here the source where I found it. From Meveric @ Odroid. Also explanation of how to install. No wget, and change filename to the file you've downloaded for gdebi. Or use gdebi package installer.(not tested)
I tried in armhf on the M4 in armhf Lubuntu, worked great. Also tried the arm64 in Armbian Stretch. Also great.
I didn't find any posts about Vivaldi in the Armbian forum. I thought it could be helpful.
qblueRed42 reacted to Tido in Rock Pi S, RK3308 CPU, is it supported by anything?
is aimed at ALLWINNER SoC, not the ones from Rockchip
qblueRed42 reacted to lanefu in List of Stuff
My list of arm boards was too long to make it a tasteful signature so here's some highlights:
My List of geek Stuff
3xOpi Plus2e - hashi cluster servers Opi Prime - main work area (screen, ssh vim, git) OrangePiOne plus - hashi cluster worker Le Potato - tester Atomic PI - desktop Tritium-h5 - tester Opi PC2 - tester Opi3 - tester Espressobin - tester Helios 4 - 2ndary NAS Orange Pi R1 - tuyaconvert Opi Lite - retrorangepi R69 retrorange pi Opi One - root cause of my SBC and Armbian addiction Frankendell i7-2600 - garage computer and build box Dual Xeon E5-2650 v3 - new build server Synology 1815+ Primary NAS Edge Router Lite - Router
Do you want to share highlights of your Arm empire?
qblueRed42 reacted to sfx2000 in Raspberry Pi 4 Released - From $35 USD
Interesting - some casual benchmarking of Pi4 vs Pi3...
For virtualization - the new Broadcom SoC is a good step forward for the Pi Folks...
A lot of this comes from how the new chip does interrupt handling.
qblueRed42 reacted to NicoD in Review : Khadas VIM3
@lanefu Sorry, I forgot to share my benchmark file.
I did find out that 7-zip multicore results are not to be trusted on SBC's with different cluster performance. The VIM3 has a very big difference so it only uses 500% of 600%. The N2 has a smaller difference in performance between A73/A53 so uses 550% in 7-zip multicore scores. (also lack of memory is a problem)
CPU Miner is good, 7zip single core benchmarks are good. And for me Blender is important too.
Reasons why benchmarks can be misleading ---------------------------------------- throttling 32-bit/64-bit Difference in cores A53/A7/A15/A72 distro (ubuntu/debian...) distro version kernel version driver versions compiler version software version/outdated repositories desktop Mate/Xfce/LXDE/... display resolution/headless background processes cpu clockspeed ram clockspeed/latency ram useage/swap/zram process sheduler optimizations for the system/distro crypto engine for encryption Undervoltage config settings Wifi dongle CONFIG_HZ=250 - 1000 or any other 7-zip works a bit better on 32-bit vs 64-bit, it doesn't use all cores at 100% in multi-core scores. The percentage differs with different distro's and boards. So it's not completely exact. 7-zip has problems with big-LITTLE architecture for multi-core benchmark. Better to test the clusters seperately. Blender works a lot better on 64-bit than on 32-bit. It uses 100% of the cores. CPU Miner only works on 64-bit. Works better in Ubuntu Bionic than in Debian Stretch. Blender : BMW render @ 1080p 7-zip : Numbers are average of 3 of decompressing only All tests are done with a fan when necessary so no throttling occurs. 64-bit SBC's Khadas VIM3 |SBC bench result |CPU Miner |7-zip s/c A53 |7-zip b/c A73 |7-zip multi avg. of 3 |Blender Ubuntu 18.04.02 http://ix.io/1MFD 13.10kH/s 1577 2311 10578 42m51s Armbian@1.9S.C./1.7B.C. http://ix.io/1NRJ 48m45s Odroid N2 |SBC bench result |CPU Miner |7-zip s/c A53 |7-zip b/c A73 |7-zip multi avg. of 3 |Blender Ubuntu Bionic http://ix.io/1Brv 11.35kH/s 1564 1879 9988 50m28s NanoPC T3+ |SBC bench result |CPU Miner |7-zip s/c |7-zip b/c |7-zip multi avg. of 3 |Blender Armbian Bionic http://ix.io/1iRJ 10.99kH/s 1290 10254 1h10m25s Arbmian Stretch http://ix.io/1qiF 8.55kH/s 1275 10149 1h13m55s Rock Pi 4B |SBC bench result |CPU Miner |7-zip s/c |7-zip b/c |7-zip multi avg. of 3 |Blender Ubuntu http://ix.io/1uVr 9.50kH/s 1242 1818 7802 1h17m22s NanoPi M4 |SBC bench result |CPU Miner |7-zip s/c |7-zip b/c |7-zip multi avg. of 3 |Blender Armbian bionic hz1000 http://ix.io/1nLh 10.23kH/s 1335 2005 8352 1h13m50s CONFIG_HZ=250 http://ix.io/1BLW 10.45kH/s 1335 2007 8320 1h08m28s Armbionic@1.4/1.8 hz250 1253 1828 7821 1h12m52s Armbian bionic nightly http://ix.io/1pDo 10.24kH/s 1329 1990 8292 1h13m28s Armbian stretch desktop http://ix.io/1odF 8.66kH/s 1350 1977 8400 1h14m12s Armbian stretch dsk nightly //ix.io/1pM0 8.80kH/s 1359 1993 8500 1h15m04s Armbian stretch core no fan //ix.io/1pKU 8.80-8.65kH/s 1353 1989 8461 Armbian stretch core //ix.io/1pL9 8.76kH/s 1354 1988 8456 Armbian stretch core nightly //ix.io/1pLf 8.82kH/s 1357 1994 8494 Lubuntu Bionic arm64 http://ix.io/1oGJ 9.24kH/s CPU Miner 1056 1551 6943 1h28m13s Lubuntu Bionic armhf http://ix.io/1pJ1 1111 1769 7705 2h02m54s Lubuntu Xenial armhf http://ix.io/1oCb 989 1507 6339 2h20m51s Khadas Vim2 Max |SBC bench result |CPU Miner |7-zip s/c |7-zip b/c |7-zip multi avg. of 3 |Blender Ubuntu Xenial http://ix.io/1qkA 6.86kH/s 823 1134 6682 1h14m39s 7-zip only 600% of 800% used Odroid C2 |SBC bench result |CPU Miner |7-zip big core |7-zip multi avg. of 3 |Blender Armbian Stretch Core http://ix.io/1pZu 4.65kH/s 1390 5342 Armbian Stretch Core Nightly //ix.io/1pZJ 4.66kH/s 1391 5340 Armbian Stretch Desktop http://ix.io/1q1C 4.65kH/s 1394 5363 Armbian Stretch Desktop NGHT //ix.io/1p02 4.59kH/s 1394 5356 2h38m18s Meveric Stetch No-OC 1337 5223 2h40m00s Meveric Stretch Only RAM OC 1361 5292 Meveric Stretch OC 1548 6049 2h14m17s Ubuntu Mate Bionic http://ix.io/1q2S clocked to 100Mhz 2h35m10s Ubuntu Mate Bionic OC Doesn't work/Clocked to 100Mhz 1607 5960 2h10m21s Rock64 |SBC bench result |CPU Miner |7-zip small core |7-zip multi avg. of 3 |Blender Armbian Stretch 1.5Ghz http://ix.io/1nCj 4.06kH/s 1406 5407 3h00n32s OLD Armbian Stretch 1.3Ghz //ix.io/1iHB 3.80kH/s 1211 4904 Armbian Bionic 1.5Ghz core //ix.io/1qbK 5.00kH/s 1384 5379 Armbian Bionic 1.5Ghz dsk //ix.io/1qcb 4.94kH/s 1379 5326 2h55m56s 32-bit SBC's Odroid XU4 |SBC bench result |7-zip s/c |7-zip b/c |7-zip multi avg. of 3 |Blender Debian Jessie http://ix.io/1q6X 950 1653 8823 1h12m19s Ubuntu Bionic http://ix.io/1qbL 1219 2094 9395 1h44m19s Asus Tinker board |SBC bench result |7-zip big core|7-zip multi avg. of 3 |Blender Tinker OS 9.5 Stretch http://ix.io/1pRN 1983 7536 2h55m00s Raspberry Pi 3B+ |SBC bench result |7-zip small core |7-zip multi avg. of 3 |Blender Raspbian Default no fan http://ix.io/1q10 1471 5027 Raspbian Default http://ix.io/1q1Q 1411 5371 5h47m31s Raspbian OC http://ix.io/1q5J 1591 6141 Ubuntu Mate Xenial http://ix.io/1q65 7-zip didn't work Software versions ----------------- GIMP Blender GTKPerf SysBench SBC-bench M4 : Lubuntu Xenial armhf 2.8.18 2.79b 0.40 0.4.12 0.6.1 Lubuntu Bionic armhf : 2.8.22 2.79b 0.40 1.0.11 LuaJIT 2.1.0-beta3 0.6.1 Armbian Stretch desktop 9.5 : 2.8.18 2.79b 0.40 0.4.12 0.6.1 Armbian Bionic : 2.8.22 2.79b 0.40 1.0.11 LuaJIT 2.1.0-beta3 0.6.1 Tinker : TinkerOS 9.5 Stretch : 2.8.18 2.79b 0.40 0.4.12 0.6.1 Odroid C2 : Armbian Stretch 9.5 : 2.8.18 2.79b 0.40 0.4.12 0.6.1 : Ubuntu Mate Bionic : 2.8.22 2.79b 0.40 1.0.11 LuqJIT 2.1.0-beta3 0.6.1 Doesn't work clocks to 100Mhz Meveric Stretch 9.5 : 2.8.18 2.79b 0.40 0.4.12 Doesn't work Rock64 : Armbian Stretch 9.5: 2.8.18 2.79b 0.40 0.4.12 0.6.1 : Armbian Bionic : 2.8.22 2.79b 0.40 1.0.11 0.6.2 RPi 3b+ : Raspbian Stretch 9.5 : 2.8.18 2.78a 0.40 0.4.12 0.6.1 Ubuntu Mate Xenial : 2.8.16 0.40 0.4.12 0.6.1 Odroid XU4 : Debian Jessie : 2.8.14 2.72b 0.40 0.4.12 0.6.1 7-zip doesn't work : Ubuntu Bionic : 2.8.22 2.79b 0.40 1.0.11 0.6.2 NanoPC T3+ : Armbian Bionic : 2.8.22 2.79b 0.40 1.0.11 0.4.6 Armbian Stretch : 2.79b 0.40 0.4.12 0.6.2 Khadas Vim2 Max : Ubuntu Xenial : 2.8.16 2.76b 0.40 0.4.12 0.6.2 CPU Clocks ---------- Khadas VIM3 : Ubuntu Bionic : 2x1.8Ghz(A53) + 4x2.2Ghz(A73) 64-bit Odroid N2 : Ubuntu Bionic : 2x1.9Ghz(A53) + 4x1.8Ghz(A73) 64-bit NanoPi M4 : Armbian Bionic/Stretch : 2x2Ghz + 4X1.5Ghz 64-bit Lubuntu armhf/ARM64 : 2x1.8Ghz + 4X1.4Ghz armhf 32-bit / ARM64 64-bit Tinker Board : TinkerOS Stretch : 4x1.8Ghz 32-bit Odroid C2 : Armbian Stretch : 4x1.5Ghz 64-bit Ubuntu Mate Bionic : 4x1.5Ghz RAM 912Mhz 64-bit Ubuntu Mate Bionic OC : 4x1.75Ghz + RAM 1104Mhz 64-bit Rock64 : Armbian Stretch : 4x1.5Ghz 64-bit Armbian Bionic : 4x1.5Ghz 64-bit RPi 3B+ : Raspbian Stretch : 4x1.4Ghz no fan 4x1.2Ghz above 60°C 32-bit Raspbian Stretch OC : 4x1.570Ghz over_voltage=4 core_freq=500 sd_freq=510 32-bit Ubuntu Xenial : 4x1.4Ghz 32-bit Odroid XU4 : Debian Stretch : 4x1.4Ghz + 4x1.9Ghz 32-bit : Ubuntu Mate Bionic : 4x1.5Ghz + 4x2Ghz Underclocks when above 75°C 32-bit NanoPC T3+ : Armbian Bionic : 8x1.4Ghz 64-bit Some benchmark tools can give an estimate of the performance. But they are never an exact reflection.
qblueRed42 reacted to NicoD in Review : Khadas VIM3
I've just finished reviewing the Khadas VIM3.
It's an amazing SBC with unmatched performance. It's got the Amlogic A311D SoC. Comparable with the S922X of the Odroid N2.
But the 4x A73 cores are clocked at 2.2Ghz and the 2 x A53 cores to 1.8Ghz. So it performs a lot better.
It also comes with an NPU, MIPI-DSI and MIPI-CSI. And can be powered with USB type-c from 5V up to 20V.
Here my review video.
Here all my gathered information.
Khadas VIM3 ----------- Ubuntu server - fan over fins ----------------------------- Blender BMW no fan thermal pad : 59m with copper shim : 58m18s with 5V fan : 45m01s with 5V fan over CPU : 43m28s Ubuntu XFCE - fan over CPU, not over fins ----------------------------------------- Blender no fan : 55m22s Blender 3V fan : 43m03s Blender 5V fan : 42m51s Blender with fan : 48m45s Armbian Bionic Small cores @1.9Ghz Big cores @1.7Ghz Blender with fan : 46m29s Armbian Disco Small cores @1.9Ghz Big cores @1.7Ghz Blender with fan : 46m56s Armbian Buster Small cores @1.9Ghz Big cores @1.7Ghz Kdenlive 16m bench 3V lxde : 1h09m (using 1GB zram + 4GB swap file. Not enough ram with 2GB) Video playback -------------- Ubuntu ------ Up to 4K 30fps with MPV. Very slight screen tearing. Anything lower is perfect Youtube playback with Firefox up to 1440p 30fps perfect KODI doesn't work (yet) No VPU acceleration for now. All done by CPU. LibreElec --------- Up to 4K 30fps with MPV. Very slight screen tearing. Anything lower is perfect Wifi doesn't work. power consumption ----------------- Idle no fan : 0.27A 5V (governor set to interactive) Maxed out with fan : 1.4A 5V (fan is 0.15A) Temperatures (fan over CPU, not over fins) ------------------------------------------ Idle No fan : 44°C Maxed out No fan : 75°C heavy throttling. 1Ghz small cores / 1.8Ghz big cores Idle 3V fan : 34°C Maxed out 3V fan : 73°C very light throttle. Takes 5min to reach 70°c. +10min to start throttling Idle 5V fan : 32°C Maxed 5V : 70°C no throttling Transfer speeds --------------- eMMC 16GB : 60 MB/s write : 165.9 MB/s read SSD over USB3 : 380.3 MB/s write : 286.1 MB/s read SSD over USB2 : 40.2 MB/s Samsung EVO plus with on-board sd-reader : 10.2MB/s write 22.1MB/s read with USB3 sd-reader : 31.7MB/s write 88.6MB/s read Ethernet internet speed : 53.9 Mbps 11.5 Mbps Wifi internet speed : 43.7 Mbps 11.2 Mbps Issue's ------- When using a 2.4Ghz dongle for mouse in the right USB3 port, wifi connection is slow. When using a USB3 device in the USB3 port wifi stops working. Use left USB2 port for those. Ubuntu server boots/shuts down slower than XFCE version. Shut down stop job for ifup for eth0 (not in use) 1m30s Boot hangs a while. Can't find the reason. After -> Scanning for Btrfs filesystems. Done. about 30s Can't install Tensorflow khadas@Khadas:~/tensorflow$ pip3 install tensorflow-1.8.0-cp35-none-linux_aarch64.whl tensorflow-1.8.0-cp35-none-linux_aarch64.whl is not a supported wheel on this platform. Android : Can't connect wifi. I can type wifi password, but can't go further. No connect button. Tips ---- Use interactive governor for better thermals and power consumption with low use/idle. Plus/Minus ---------- + Best single core performance + best multi-core performance for an ARM SBC + Nice case + PSU + USB-c cable included + remote + eMMC + Amazing low power consumption / best performance per watt + Can be powered with 5V up to 20V + NPU - Wifi issue with USB3/No 5Ghz wifi. Could have solved the USB3 interference - Placing of the buttons would be better on the back. They often get pressed when plugging a USB device. You do get used to it. - No VPU and X11 drivers (yet) - Slow On-board SD-card reader
qblueRed42 reacted to sfx2000 in SBC recommendations for a wireless router
espressoBIN is a decent choice - good performance at a decent price, and a community here on Armbian that can support it
The ClearFogs are a step up, even though their on an older ARMv7A Armada - the 8K is newer and is ARMv8, but not sure where the BSP support is there.
There's been a trio of QC-Atheros IPQ-40xx devices recently announced, and they have integrated 802.11ac dual band radios - but most of them are somewhat tight - e.g. to get best performance out of them, one has to use the QSDK, which is built on an older version of OpenWRT - there is support for some IPQ40xx in OpenWRT 18.06 and Master, but I'm not sure it's stable enough yet for daily-driver use on the WiFi (ath10k drivers are under heavy development, and there's issues with the Switch outside of typical setup - see DSA on this chip over on OpenWRT forums...)
I was starting to develop a fairly open board based on AR9331/AR9531 - not ARM based, as these are MIPS cores, but the SOC is very well documented, and strong OpenWRT support, but that got set aside when work stepped up and my free time was reduced quite a bit.
qblueRed42 reacted to lanefu in Experience working with Linaro or have Linaro contacts?
I'd really like to build a relationship with Armbian and Linaro... even a small one. Although most of the SBC manufactures we support aren't a member of Linaro... Armbian is a huge win for Arm as a whole, and impacts influencers who can choose to use ARM products in enterprise settings etc.
qblueRed42 reacted to sfx2000 in Raspberry Pi 4 Released - From $35 USD
I'll reserve opinions for now, other than it seems like a good step forward for the RPi team, and an appreciation for the amount of effort to basically tailor a chip for their product.
But it seems like a nice device, and Pi has good community support, and they offer both HW and SW that works for their audience - so that's all good.
Quick thoughts though...
1) 4 big ARM cores - 28nm, so some thermal challenges as we know from other SoC vendors on this node
2) VC6 - which is a big step from the BCM chips that Pi has used in the past
3) Real USB3 - early reports is there is opportunity for Pi folks to improve some perf there
4) Real 1Gb Ethernet - same here - driver work is likely going to improve performance
5) Early reports are that the USB-C for power is also USB2 OTG - need to confirm
Big step for them on the HW side, and moving to Buster for Raspbian is a bit bold, considering that Buster at the moment is still in testing upstream - Rasbian is still 32-bit, so I get that, and that's not a complaint, just an observation.
That being said - It's still in many ways the same story - there's Open, and then there is Kinda Open - Pi's prior to Pi4B were in that Kinda Open realm with the odd bootloader and sequence there, along with it having a close source OS underlying it all running on the VC4 host.
Busy with jobby-job stuff at the moment, so I'll let other run into the fire on Pi4B.
qblueRed42 reacted to chwe in Raspberry Pi 4 Released - From $35 USD
I couldn't less agree on this one..
It is a TV box SoC with schematics.. It might not matter if you only consume and someone else provides the images.. But it matters if you're on the other side. You (mostly) know what you get, whereas TV boxes sometimes change depending on what is cheaper at the moment (e.g. different wifi module, NAND instead of eMMC etc.).
well.. IMO it was RPi who had to respond.. all those RK3399 based boards crushed the RPi fully..
I assume we'll see more specialized SBCs in the future. e.g. https://www.cnx-software.com/2019/06/19/rock-pi-s-tiny-sbc-rockchip-rk3308-processor/
some NAS ones etc.
Indeed it is the first RPi since v2 which doesn't look annoying. They try their first step away from VC4 which might be interesting.. I don't get this (claimed) full backwards compatibility of images. IMO it was a good opportunity to make a cut (keep, pin-compatibility etc. but craft an image which is either fully armhf - or arm64). having a VC4 and a VC6 branch of raspian is something the majority of their believers would probably accept.
I will likely buy one (will be the first one I bought after the Pi2 IIRC), I've a project where the A72 cores might shine, and if thermals are somehow controllable passively, 4xA72 for 35$ looks good to me (don't need really much ram for this project, it's really about numbers crunshing). They did a couple of things right, e.g. SPI NOR, USB-C instead of microUSB (btw it's a 'dumb' one right? so no PD), even the double HDMI was IMO a smart move (now it's somehow unique to its competitors). And from the software side I'm quite sure things will mature over time (looking forward to the PXE, my project would really benefit from a proper PXE implementation), and we'll soon get some deeper insights from people experienced in kernelcode.. E.g.
But I don't think that 'chinese' (whatever that means) have to respond. For different use-cased several companies have a good line-up, and the user group they target is often different.
It was a needed, and from the first glace well crafted update of the (IMO) complete failure called RPi3b+. I was surprised by the 4xA72 core they've chosen and the 1,2 and 4 GB ram was smart too. Let's face it, the average 'rich' user will always opt for the 4GB variant no matter he needs it or not and I assume the profit margin will be slightly higher with this variant.
qblueRed42 reacted to chwe in SBC recommendations for a wireless router
but sure, I wouldn't go with the R2 as my main router.. things are just not 100% clear yet.. and network is still a bit.. mhmm.. flunky might describe it best. There's no development cause nobody is really interested in pushing stuff here..
me neither.. you can learn things by pure pain.. pain and time..
I would separate NAS from router jobs..
If the price is not your main driving force, you might look into solidrun devices. https://www.solid-run.com/product-category/sbcs-fanless-pcs/
qblueRed42 reacted to chwe in Very Small Platforms - Rockchip 3308 and Allwinner V3s
remember those dirt cheap android sticks you could buy years ago? The MK802.. I knew that I bought one years ago.. but didn't know for a long time that they're equipped with a AW A10... http://linux-sunxi.org/Rikomagic_mk802
Well, seems that they're supported in mainline (u-boot and kernel).
__ __ _ _____ ___ ____ | \/ | |/ ( _ ) / _ \___ \ | |\/| | ' // _ \| | | |__) | | | | | . \ (_) | |_| / __/ |_| |_|_|\_\___/ \___/_____| Welcome to Debian Stretch with Armbian Linux 5.1.7-sunxi System load: 2.26 1.12 0.43 Up time: 2 min Memory usage: 4 % of 999MB IP: CPU temp: 11°C Usage of /: 4% of 30G ... root@mk802:~# armbianmonitor -m Stop monitoring using [ctrl]-[c] Time CPU load %cpu %sys %usr %nice %io %irq CPU 21:44:11: 1008MHz 1.57 82% 30% 20% 0% 30% 0% 11.6°C 21:44:16: 1008MHz 1.68 19% 19% 0% 0% 0% 0% 10.9°C 21:44:21: 1008MHz 1.55 13% 13% 0% 0% 0% 0% 11.4°C 21:44:27: 1008MHz 1.42 22% 21% 0% 0% 0% 0% 11.2°C 21:44:32: 1008MHz 1.31 17% 14% 1% 0% 0% 0% 11.6°C 21:44:37: 1008MHz 1.20 21% 20% 0% 0% 0% 0% 11.0°C 21:44:43: 1008MHz 1.11 13% 13% 0% 0% 0% 0% 8.5°C well the thermal is a bit sloppy.. in fact the SoC was roughly 60°C at this time...
and soldering UART to test points isn't as fun.. but it works
and now imagine this board with sata instead of HDMI wired out.. (well maybe with the A20 instead of A10)..
qblueRed42 reacted to gounthar in Very Small Platforms - Rockchip 3308 and Allwinner V3s
qblueRed42 reacted to yam1 in Good Armbian compatible boards
If you can live with hdmi being not working in the mean time, I would recommend OPI One Plus or Lite2. They are just faster H5's and they work with
little heat sinks. I used them to compile the esp32 cross tool chain which took like two hours and they just worked. Before this I also had high hopes
for k1 plus but it crashed like 20 minutes into the compilation - this was using the bigger stock heat sink without fan - my PCs are already making
enough noise for me. Heat is the biggest problem for any serious computing using these boards. Note that CPU throttling worked amazingly well on these
H6 boards, the temperature never got higher than 78C and the CPUs were throttling between 1.5/73 to 1.8/78. They are also cheaper, for the lack of
USB's, I use a $2 hub, SPI lcd works in 4.20, but not 5.1.
qblueRed42 reacted to guidol in Good Armbian compatible boards
a good start - I think - would be the NanoPi K1 Plus
- 64bit SoC (well supported H5-CPU - also cool running)
- 2GB of RAM
- HDMI (but no 4K)
Wth 1GB of RAM there is also the OrangePi PC2 - but no WiFi
qblueRed42 reacted to sfx2000 in Announcement : Odroid N2
Normally with big.LITTLE, tasks migrate from the LITTLE core to the big core - much depends on the kernel scheduler in use, and whether that scheduler is aware of ARM MP in a big.LITTLE context - in any event, one does need to consider performance per core, and how to schedule tasks. Done right, it works pretty well, done wrong and one can get into contention conflicts between cache, main memory, and arbitration on who has control of the CCI at any given time.
nVidia also has the 4 Cortex-A53's along with the 4 Cortex-A57's, but they don't expose the A53's directly to the kernel - management is done below the kernel in the hardware, which might make scheduling easier in the linux kernel.
Worth reading - it's a few years old - https://www.arm.com/files/pdf/big_LITTLE_Technology_the_Futue_of_Mobile.pdf
More recently, there's been a lot of work around energy aware scheduling, and ARM is one of the big players there - https://developer.arm.com/tools-and-software/open-source-software/linux-kernel/energy-aware-scheduling
It's in Kernel 5.0, but perhaps some folks might consider back-porting it if they have the spare time to do so.
Anyways - more cores doesn't mean faster - there's a limit there - and this is more of a systems engineering issue, along with SW... and this goes back to 1967 with Amdahl's presentation...
Much like Shannon with Information Theory and Wireless Communication - one can only put so many bits into a medium before it becomes error limited.
Anyways - it's an interesting time where we're getting the big ARM cores in boards that are $100USD or lower - but it also starts to point out that we need to have better solutions to power these larger cores...
qblueRed42 reacted to sfx2000 in Announcement : Odroid N2
Would be interesting to see how it competes with Jetson Nano with a good power supply...
Jetson has a fairly large heatsink, but it's generally cool to the touch right now with 2.5A over MicroUSB - 5 amp Power Supply is scheduled to arrive tomorrow, and then we can turn up the Jetson clocks and see what it can really do...
(Jetson Nano under full load can easily consume 20w when doing GPU compute tasks - not too much different than Google Coral)