Jump to content
g__day

Pondering a workstation build to process astro images + occassionally play games

Recommended Posts

So it's consistent across all its operations? From the moment you press go and look at CPU usage you only see one core loaded throughout? Will try it on my sons i5 and i7 cores on WIn 7 and WIn10 - but from memory yours is the first ever PC I have heard only using one core. Will also ask on several astronomy forums and see what they can suggest.

 

Cheers Dasa

Share this post


Link to post
Share on other sites

yeah it never loads more than one core

it couldnt be due to the image im using?

Share this post


Link to post
Share on other sites

Possibly most images would have the star fields move by a few pixels each shot given tracking isn't perfect or the mounts axis might not be orthogonal to the Earth's axis.

 

You could try ten frames and if you photoshoped then to shift the image around slightly but keep the same picture size (width and breadth) it would give DSS a bit more work to do.

Share this post


Link to post
Share on other sites

Interesting development last week - GITHub posted a 20.4MB 64 bit version of Deep Sky Stacker v4.1.0 and its much, much faster - I just stacked 20 images of M20 and all up it took about 3 minutes.

 

https://github.com/LucCoiffier/DSS/releases/download/4.1.0-Release/DeepSkyStacker64Installer.zip

 

Have to play with it some more - but it looks pretty much the same - just the stacking is a lot faster, not sure about the registering but stacking was wow the first time around!

Share this post


Link to post
Share on other sites

 

i wouldnt bother with aio there expensive more prone to failure and loud for the level of performance they provide

Can you point me to some failure rate info?

 

The main reason I haven't looked at this myself is because if liquid cooling fails it can do more damage.

 

As for noise & performance, from what I've read it's only the very top end air solutions that stack up and, as I've experienced, they are very dependent on mobo layout.

 

Just a little update on this

Mates pc with a h80 ~7 years old just exploded at the cpu block all over his sound card the pc was making some horrible sounds through the speakers when he tried to turn it on and when he realized why he ripped the power out till it dried and it seems to have survived

A few days later i see somebody on whirlpool forums just upgraded to a 1080ti only to have there h80i leak all over it a few months outside of its 5y warranty they were not so lucky most there pc is dead including the new $1400 gpu

 

I have seen a few do this inside of warranty in the past and on a case by case basis corsair may cover the cost of the parts that died

Edited by Dasa

Share this post


Link to post
Share on other sites

 

 

i wouldnt bother with aio there expensive more prone to failure and loud for the level of performance they provide

Can you point me to some failure rate info?

 

The main reason I haven't looked at this myself is because if liquid cooling fails it can do more damage.

 

As for noise & performance, from what I've read it's only the very top end air solutions that stack up and, as I've experienced, they are very dependent on mobo layout.

 

Just a little update on this

Mates pc with a h80 ~7 years old just exploded at the cpu block all over his sound card the pc was making some horrible sounds through the speakers when he tried to turn it on and when he realized why he ripped the power out till it dried and it seems to have survived

A few days later i see somebody on whirlpool forums just upgraded to a 1080ti only to have there h80i leak all over it a few months outside of its 5y warranty they were not so lucky most there pc is dead including the new $1400 gpu

 

I have seen a few do this inside of warranty in the past and on a case by case basis corsair may cover the cost of the parts that died

 

 

I guess it doesn't help if the card has any form of dust (which of course it does) but wouldn't the water be demineralised? I woulda hoped so :(

Share this post


Link to post
Share on other sites

I guess it doesn't help if the card has any form of dust (which of course it does) but wouldn't the water be demineralised? I woulda hoped so :(

It would be to start with so if it leaked when it was brand new it may not do much harm but over time water absorbs minerals from the metals in the loop and becomes conductive

Share this post


Link to post
Share on other sites

IMO relying on "pure water" to be nonconductive is fantasy-land wishful thinking. Even if it wasn't conductive such an explosion would mean it'd immediately pick up dust and contaminants from the air and surface of parts it landed on.

I've got a water cooled system sitting on the carpet here - it's not been powered up in a long time but I still regard it as a bit of a time-bomb.

Share this post


Link to post
Share on other sites

This was an itneresting re-read, now that I'm a little more familiar witht he process behind astro stacking, and hardware has moved on in 3 years.  eg, a Xeon ES-2695 @ 2.4GHz used to be a beat at the top, but now just loses out to a 'lowly' ryzen 2700x, and both are smashed by the ryzen 9 3900x and threadripper 1950x, let alone the new 3950x.

 

@Master_Scythe have you looked back over the current rankings on http://pixinsight.com/benchmark/ to see how your predictions played out?

 

@g__day how are you finding your workstation?  And if you don't mind my asking, what kind of workflow are you using?  I grabbed DSS the other week to have a play with stacking shots taken with my phone camera, rather than relying on Google's 'night sight/astrophotography' algorithms, but then I got distracted with dead PC/new PC mess, so I haven't had a chance to dig into it further.  When the night is clear I enjoy spending the 40-60 minutes outside until my phone dies trying to take decent photos, so I'm probably in the boat of starting to question if I want to save for a motorbike, or a proper tracking mount for/and a digital camera, ahem.


 

Using 10 copies of the image Dasa posted above and doing what you said with suggested defaults, registering is using 6 logical cores, but they're not pegged at 100%. ~1:20 per image to register.  Resmon says deepskystacker.exe is using 3 threads across 6 cpus *I think*.  Actual stacking after registering (which took around 15 mins) was ridiculously fast, less than 5-10s? And seemed to use all cores but again not pinning them (I can't say if that'd go up using a proper set of lights/darks/flats, or if there'd be some thermal throttling of my CPU by that point if it dragged out over 30s).  Luminance, fiddle with every slider, hit apply, and less than 5 seconds, all cores used to maybe 60%, and done.  With a single light with 10 copies, no darks or flats or variations, it didn't seem to be using my SSDs with swap, and I didn't see memory usage spike, but I'm not sure if RAM speed was a limiting factor for registering?  Tested using DSS 4.2.3 FWIW (system is a Ryzen 3700x, 32GB of DDR4 @ 3200, and swap set to use my two Samsung EVO SSDs hanging off SATA ports).

 

FWIW, this was from registering.  Stacking happened too fast to line up windows:
 

image.png

Share this post


Link to post
Share on other sites

Nich I haven't played with it in quite a while - will revisit it all.  Stepping to the 64 bit version of DSS and running it on this big PC was many times faster than the older quad core PC running 32 bit software!

  • Thanks 1

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×