what is the average power draw when using the 8 cores and I guess GPU as well as operating temperatures.
I do plan on using a fan but just would like to know power draw when all cores are maxed with limited things connected to USB ports etc and regular HDMI etc.
It all depends, lies, damned lies & benchmarks etc. 
I am running room temperature 18 Centigrade at the moment on a VIM2 Max that I have just a passive heatsink on the CPU.
$ uname -a
Linux dukla 4.9.40 #2 SMP PREEMPT Wed Sep 20 10:03:20 CST 2017 aarch64 aarch64 aarch64 GNU/Linux
chris@dukla:~/bin/Pi3$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 17.10
Release: 17.10
Codename: artful
I ran an old script I had lying around from RPi days (source at the bottom of this post) that loaded all 8 CPUs 100% for about 15 minutes, together with some IO. The max temp I observed was 76 C, I never observed any throttling (I use gkrellm as a realtime monitor). I have a Brennehstuhl monitor on my AC wall output: at idle the power usage is too low to register (power to 1/100W, current to 1/100A) at 238VAC. Under load while running the script I read 9.52W, 0.04A, 238V
First below is final output from the script:
writing: 9
512+0 records in
512+0 records out
536870912 bytes (537 MB, 512 MiB) copied, 2.4971 s, 215 MB/s
writing: 10
512+0 records in
512+0 records out
536870912 bytes (537 MB, 512 MiB) copied, 2.60761 s, 206 MB/s
./octest.sh: line 17: 19107 Terminated nice yes > /dev/null
./octest.sh: line 17: 19109 Terminated nice yes > /dev/null
./octest.sh: line 17: 19111 Terminated nice yes > /dev/null
./octest.sh: line 17: 19113 Terminated nice yes > /dev/null
./octest.sh: line 17: 19115 Terminated nice yes > /dev/null
./octest.sh: line 17: 19117 Terminated nice yes > /dev/null
./octest.sh: line 17: 19119 Terminated nice yes > /dev/null
./octest.sh: line 17: 19121 Terminated nice yes > /dev/null
CPU freq: 1512000
CPU temp: 73000
[176695.357241] x5 : 0000007f41f63f20 x4 : 0000000000000001
[176695.357252] x3 : 00000000e3ffac30 x2 : 0000000000000000
[176695.357262] x1 : 0000007f9a2e4700 x0 : 0000000134cff6f8
[177266.141747] hid-generic 0005:04E8:7021.0010: unknown main item tag 0x0
[177266.143254] input: Bluetooth 3.0 Keyboard as /devices/platform/c11084c0.serial/tty/ttyS1/hci0/hci0:12/0005:04E8:7021.0010/input/input18
[177266.144166] hid-generic 0005:04E8:7021.0010: input: BLUETOOTH HID v1.1b Keyboard [Bluetooth 3.0 Keyboard] on b0:f1:ec:a3:ae:71
[178328.758611] hid-generic 0005:04E8:7021.0011: unknown main item tag 0x0
[178328.760040] input: Bluetooth 3.0 Keyboard as /devices/platform/c11084c0.serial/tty/ttyS1/hci0/hci0:12/0005:04E8:7021.0011/input/input19
[178328.763655] hid-generic 0005:04E8:7021.0011: input: BLUETOOTH HID v1.1b Keyboard [Bluetooth 3.0 Keyboard] on b0:f1:ec:a3:ae:71
Not crashed yet, probably stable.
And now the script itself:
#!/bin/bash
#Simple stress test for system. If it survives this, it's probably stable.
#Free software, GPL2+
echo "Testing overclock stability..."
#Max out all CPU cores. Heats it up, loads the power-supply.
for ((i=0; i<$(nproc --all); i++)); do nice yes >/dev/null & done
#Read the entire SD card 10x. Tests RAM and I/O
for i in `seq 1 1`; do echo reading: $i; sudo dd if=/dev/mmcblk0 of=/dev/null bs=4M; done
#Writes 512 MB test file, 10x.
for i in `seq 1 10`; do echo writing: $i; dd if=/dev/zero of=deleteme.dat bs=1M count=512; sync; done
#Clean up
killall yes
rm deleteme.dat
#Print summary. Anything nasty will appear in dmesg.
echo -n "CPU freq: " ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq
echo -n "CPU temp: " ; cat /sys/class/thermal/thermal_zone0/temp
dmesg | tail
echo "Not crashed yet, probably stable."Did you ever try to run cpuburn-a53? I wouldn’t call ‘nice’ a stress testing tool since it only generates some pretty light load. Better try this instead:
git clone https://github.com/ssvb/cpuburn-arm
gcc cpuburn-a53.S
./a.out
Yup - it does create more load and shows the thermal throttling kicking in at 80C.
I don’t think OP was looking for max stress - more ballparks or averages. But as per my original post there are lies, damned lies & benchmarks, and in that world I figure it is crucial for full disclosure hence my answer was longer than 9.5W, 76C!