Greetings!
This is Autostakkert! v3.1.4 – 64 bit
I have gone through three computers in the last two years, that I use for astronomical image processing. Each one was a little faster than the others and where we stand today is a fairly new
Ivanova2021 win10pro64bit 21H2 Core I5-6500 @3.2GHz, 8GB; 240GB SSD 8GB DDR4 1x8GB (4 slots) sda 240GB SSD lexar sdb 6TB WDBlue data; sdc 4TB external USB3 data backup 1920×1080 HDMI; 6xUSB3; 2xUSB2 $260CDN x2DP,x1VGA BIOS 2021Nov18
And it is still TOO SLOW. I think that maybe I am processing larger images, to higher degrees of quality? So I thought.. where may there be a potential bottleneck? Don’t know… have to run some benchmarks.
So I took a sample image, filename: 2022-09-03-0312_9_kk-L-Jupiter.ser
filesize 5,339,633KB = 5.3GB
8522 frames, 8bpp
file on local C:\data SSD drive
And ran it through autostakkert! v3 four times, increasing the thread options from 1/4 to 4/4.
There is INDEED a difference!
The process was the best 10% using 75 Alignment points, drizzle x1.5 and saved as a .PNG
The CPU% and RAM% did increase and the processing time did drop:
number of threads 1
CPU% 27%
RAM GB 3.8
stacking time 1017 sec
number of threads 2
CPU% 40%
RAM GB 3.9
stacking time 527 sec
number of threads 3
CPU% 55%
RAM GB 5.6
stacking time 363 sec
number of threads 4
CPU% 75%
RAM GB 5.6
stacking time 327 sec
So far so good… use all the threads. Next up.. RAM. at 8GB and it seems that autostakkert uses all of what is available.
Soooo.. we have ordered more RAM and will rerun the above processes with 16GB onboard! DDR4 !