On 1/12/24 22:34, Maarten Verberne wrote:
I've already settled on that - Ryzen 7950X. Enblend and other
graphics applications I use really benefit from multiple cores and
threads, so the more cores, the better.
i'm still dreaming of replacing my 3600 with a 5900 or 5950...but alas
i do not have the resources, let alone to go to a new platform. a 7950
would probably outpace the 3600 by a factor 2 or more.
I don't have resources for it yet, either. But if my Dell laptop gives
up the ghost (the Thunderbolt/USB-C port died last year), the
replacement dollars go into the desktop. That currently has a
motherboard running an Intel Pentium 4, so a motherboard replacement/new
memory/new cooling system is inevitable.
Since I want performance, Intel's continued love affair with
efficiency cores doesn't appeal to me. My experience with Intel
performance processors has been that while Intel loves to chant their
peak clock speeds, their processors can only hit that speed with a
single core is enabled. What's the point of that when the software
happily and rapidly uses multiple cores?
since the whole process of stitching is more a linear process, it is
mostly a single thread process. that does not mean it won't run on
multiple cores, they just wait on each other.
so in my experience with hugin, the best single thread machine is the
quickest.
Not in my experience. Stitching starts, 16 threads fire up, and checking
in htop shows none of them waiting for others.
I understand that enblend isn't always compiled with multiprocessor
support, maybe that's the difference?
My working method is interactive. I don't do nearly as many images as
you do!
it was an experiment for me, but i'm not sure i will continue this,
the fact that, by now, i have a dedicated pc running for some 3 month
a year just to stitch the images is a bit much.
Yeah, might be a bit much. But might be more cost effective than
alternatives?
I think nona uses OpenGL, which NVidia doesn't really support. NVidia
wants to lock customers into their platform; the antithesis of
OpenGL. Nona plus the on-board Intel UHD-630 works fine.
True, however nvidea does state their rtx30xx series does support
openGL, but not without a heavy penalty. when i saw my uhd630 run
circles around the dedicated rtx while the rtx was running at some 100
watt i knew all i needed to know.
Yes, Nvidia isn't on my list. If I was using Adobe under Windows on this
machine, I suppose the resident RTX would get used.
I understand AMD's GPUs are more power efficient than the RTX3000 series
GPUs.
Another question to think about.
Multicore CPU: yes, many cores/threads. Start three processes, each
gets a core/thread.
Is the same true about GPUs? Or does a GPU handle input from only one
source at a time? So if script 1 fires off Nona on the GPU, what
happens when script 2 and script 3 try to run nona on the GPU at the
same time?
It stays linear, so they will wait for each other to finish and
sometimes one of the script gets ahead. the same is true for the gpu.
but with only one cmd script my cpu will only run at some 20-30% (all
cores) so that is where multiple scrips come in handy.
Ah. Interesting.
Maarten, in my experience, replacing your HDD drives with SSDs would
make a big difference. Even connected via SATA cables, SSD is faster.
NVME drives (if your motherboard supports them) would be even faster.
If your motherboard doesn't support NVME, you might invest in a
4-port PCIe expansion card that adds NVME connections, and replace
your HDDs with NVME SSDs on the card. I think it would massively
increase read and write speeds.
I do have place for one extra m.2 and have enough sata ports left and
am aware of the advantage of ssd, but i just can't afford that..at
least not this year.
my solution to use one hdd per running script and one 'master' for the
originals is the best i can currently do with the means i have.
if i had a method to seperate the nona tiff and the location where
enblend writes the final image, i might be able to gain a bit...but
wouldn't know how or if that can be done.
Hmm, I'd think that since you're doing this through scripts, you'd have
control over where final images are written. I've never done that, but
might be worth asking about.
but if i decide to continue this year, there will come a time that i
get one or more ssd drives for this. (i would need something like 4TB
just to get started)
I survive on a mere 2TB M2 drive, but don't do as much heavy lifting as you.
Go for as much processor performance and memory you can. Hugin spends
nearly all of its processing time running on the CPU and using memory.
CPU and GPU are more important, in my experience Hugin isn't that
memory intensive
Hmm, I've had Hugin (particularly enblend) consume more than the 64GB
RAM in my laptop when stitching. Probably depends on the sizes of the
source images and the final image. Perhaps the image format, too?
I don't think GPU matters at all, as you pointed out about Intel onboard
GPUs outrunning the fancy GPUs. If the GPU supports OpenGL (without
throwing you out of house and home with its electric bill!), then any
basic GPU is good. :)
--
David W. Jones
[email protected]
wandering the landscape of god
http://dancingtreefrog.com
My password is the last 8 digits of π.
--
A list of frequently asked questions is available at:
http://wiki.panotools.org/Hugin_FAQ
---
You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/hugin-ptx/1eba9407-b2ea-4e48-b479-b6a05692d103%40gmail.com.
