END OF AN ERA, FRACTALFORUMS.COM IS CONTINUED ON FRACTALFORUMS.ORG

it was a great time but no longer maintainable by c.Kleinhuis contact him for any data retrieval,
thanks and see you perhaps in 10 years again

this forum will stay online for reference
News: Follow us on Twitter
 
*
Welcome, Guest. Please login or register. September 18, 2019, 11:08:30 PM


Login with username, password and session length


The All New FractalForums is now in Public Beta Testing! Visit FractalForums.org and check it out!


Pages: [1] 2 3   Go Down
  Print  
Share this topic on DiggShare this topic on FacebookShare this topic on GoogleShare this topic on RedditShare this topic on StumbleUponShare this topic on Twitter
Author Topic: Any recommendations for a good desktop pc for fractals?  (Read 3365 times)
0 Members and 1 Guest are viewing this topic.
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« on: July 10, 2015, 05:44:27 PM »

I'm using an old 32bit computer laptop and I was going to change to a 64bit with two times as much ram but a desktop pc is probalby better and customizable. Are there any recommendations for a good but not too expensive cpu? I don't have a budget yet so price doesn't matter yet but I'm probably not looking for something above 1000 euro's. I prefer to use windows because I'm mainly going to use Kalles Fraktaler and Mandel Machine.
« Last Edit: July 10, 2015, 08:02:58 PM by TheRedshiftRider » Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
hobold
Fractal Bachius
*
Posts: 573


« Reply #1 on: July 10, 2015, 11:52:49 PM »

Kalles fraktaler and Mandel Machine both heavily use floating point computations on the main processor. Graphics card is irrelevant to those programs. A quad core Intel i5 is probably best performance per dollar. The i7 series is a bit faster, but considerably more expensive; while the i3 chips are meant for power constrained machines. AMD is pretty much out of the race, sadly.
Logged
Chillheimer
Global Moderator
Fractal Schemer
******
Posts: 972


Just another fractal being floating by..


chilli.chillheimer chillheimer
WWW
« Reply #2 on: July 11, 2015, 12:28:46 AM »

i've got an i7-3770k and am very happy with it since ... 18(?) months. and it was best bang for the buck.
spend the rest of your budget on ram and especially on ssd harddrive for system an much used stuff. ssd rules and really speeds up,
my windows 8.1 is up within 10 seconds. ssd was the biggest improvement of my upgrades..
and for fractals, check this thread that I'm too lazy to search for now. but my processor scored pretty high and is cheap. I'm happy.
Logged

--- Fractals - add some Chaos to your life and put the world in order. ---
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #3 on: July 11, 2015, 08:15:39 AM »

Kalles fraktaler and Mandel Machine both heavily use floating point computations on the main processor. Graphics card is irrelevant to those programs. A quad core Intel i5 is probably best performance per dollar. The i7 series is a bit faster, but considerably more expensive; while the i3 chips are meant for power constrained machines. AMD is pretty much out of the race, sadly.

i've got an i7-3770k and am very happy with it since ... 18(?) months. and it was best bang for the buck.
spend the rest of your budget on ram and especially on ssd harddrive for system an much used stuff. ssd rules and really speeds up,
my windows 8.1 is up within 10 seconds. ssd was the biggest improvement of my upgrades..
and for fractals, check this thread that I'm too lazy to search for now. but my processor scored pretty high and is cheap. I'm happy.

Thanks. Depending on what else I'm going to use these could be a good choice. I can choose between i5 and i7, I'm not making a choice yet because there probably is a lot more to look for. I think I will also get a graphics card for various other reasons. smiley I'm not sure whether I need to use windows 7 or 10. But I think 10 should be better, I'm not going to use windows 8 because I think it is too much aimed to tablets and a little inconvenient for a desktop pc.
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
PieMan597
Conqueror
*******
Posts: 122



WWW
« Reply #4 on: July 11, 2015, 12:10:04 PM »

For the graghics ncard, I would get the GTX 750. That is what I have, and it works like a charm.

I would also use windows 7, because you can upgrade to 10 for free later.
Logged
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #5 on: July 11, 2015, 01:12:13 PM »

For the graghics ncard, I would get the GTX 750. That is what I have, and it works like a charm.

I would also use windows 7, because you can upgrade to 10 for free later.

Thats only limited and I don't think I will be able to get it all before that expires. But I think 7 is a good choice.
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #6 on: July 12, 2015, 01:27:30 PM »

What about the ammout of ram? I thought 8 would be good. Or should it be more?
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
cKleinhuis
Administrator
Fractal Senior
*******
Posts: 7044


formerly known as 'Trifox'


WWW
« Reply #7 on: July 12, 2015, 01:46:28 PM »

i am running with 16 gigs, and still think it is far to less wink
depending on what you want to do with your computer while rendering, or how many instances you might want to run at the same time,
and most important is that you match the memory speeds with your board speeds, and take the fastest available ram, ram is one of the
most essential things a computer needs, so the faster, the bigger, the better wink
Logged

---

divide and conquer - iterate and rule - chaos is No random!
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #8 on: July 12, 2015, 01:53:06 PM »

i am running with 16 gigs, and still think it is far to less wink
depending on what you want to do with your computer while rendering, or how many instances you might want to run at the same time,
and most important is that you match the memory speeds with your board speeds, and take the fastest available ram, ram is one of the
most essential things a computer needs, so the faster, the bigger, the better wink

It will be for rendering mostly, I don't have to worry about Kalles Fraktaler and Mandel Machine not being fast enough. smiley But is it possible to use between 8 and 16 ram, for instance 12gb of ram?
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #9 on: July 12, 2015, 02:15:09 PM »

Btw, I've been thinking to also install the google neural networks program, and it require quite a lot of computation power so I thought 12 would be good. But is there a sort of quality-scale I can choose from?
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
Chillheimer
Global Moderator
Fractal Schemer
******
Posts: 972


Just another fractal being floating by..


chilli.chillheimer chillheimer
WWW
« Reply #10 on: July 12, 2015, 07:06:46 PM »

regarding deepdream:
i believe the most important thing is ram. I've allocated 12gb to the virtualbox to achieve 2560*1440 resolution.
if you want it to compute faster (large images with 10*4 iterations take around 20 minutes here) you need a nvidia graphic-card that supports cuda. problem is, for the same resolution you would also need 12gb graphics ram... (yes, it is possible to use 12gb, but pay attention to dual-channel ram which (if i remember correctly) is much faster.
when I look at my task manager, it doesn't use all cpus and its going up and down all the time. I don't think the amount of cores is not so important.
« Last Edit: July 12, 2015, 10:22:22 PM by Chillheimer » Logged

--- Fractals - add some Chaos to your life and put the world in order. ---
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #11 on: July 12, 2015, 08:00:00 PM »

regarding deepdream:
i believe the most important thing is ram. I've allocated 12gb to the virtualbox to achieve 2560*1440 resolution.
if you want it to compute faster (large images with 10*4 iterations take around 20 minutes here) you need a nvidia graphic-card that supports cuda. problem is, for the same resolution you would also need 12gb graphics ram... (yes, it is possible to use 12gb, but pay attention to dual-channel ram which (if i remember correctly) is much faster.
when I look at my task manager, it doesn't use all cpus and its going up and down all the time. I don't think the amount of cores is so important.
Ok. Thanks for the information.
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
3dickulus
Global Moderator
Fractal Senior
******
Posts: 1558



WWW
« Reply #12 on: July 12, 2015, 08:41:35 PM »

re:deep dream

I have 6G system mem and a 2G nVidia card and I can render 2880x3840 images
Using cnn-vis.py https://github.com/jcjohnson/cnn-vis with cuDNN

in cnn-vis the gfx card mem is used for calculating, not holding image data, the image data goes to a png file,

the --batch-size option controls how much ram is used on the GPU,
I get this output no matter what the image dimensions are when --batch-size=100 ...
Code:
I0712 11:20:44.759096  8298 net.cpp:213] Network initialization done.
I0712 11:20:44.759105  8298 net.cpp:214] Memory required for data: 625125600
I can push it past 128 but at 128 and less seems to perform better smiley
Logged

Resistance is fertile...
You will be illuminated!

                            #B^] https://en.wikibooks.org/wiki/Fractals/fragmentarium
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #13 on: July 12, 2015, 09:57:27 PM »

re:deep dream

I have 6G system mem and a 2G nVidia card and I can render 2880x3840 images
Using cnn-vis.py https://github.com/jcjohnson/cnn-vis with cuDNN

in cnn-vis the gfx card mem is used for calculating, not holding image data, the image data goes to a png file,

the --batch-size option controls how much ram is used on the GPU,
I get this output no matter what the image dimensions are when --batch-size=100 ...
Code:
I0712 11:20:44.759096  8298 net.cpp:213] Network initialization done.
I0712 11:20:44.759105  8298 net.cpp:214] Memory required for data: 625125600
I can push it past 128 but at 128 and less seems to perform better smiley


Hm... Memory limiting. Good to know, thank you.
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
TheRedshiftRider
Fractalist Chemist
Global Moderator
Fractal Iambus
******
Posts: 854



WWW
« Reply #14 on: July 13, 2015, 12:17:26 AM »

i've got an i7-3770k and am very happy with it since ... 18(?) months. and it was best bang for the buck.
spend the rest of your budget on ram and especially on ssd harddrive for system an much used stuff. ssd rules and really speeds up,
my windows 8.1 is up within 10 seconds. ssd was the biggest improvement of my upgrades..
and for fractals, check this thread that I'm too lazy to search for now. but my processor scored pretty high and is cheap. I'm happy.
I think Im going for the INTEL CORE I7 4770. Im not conpletely sure yet but the price/performance ratio is great. Though it is still expensive in here, I may have to go to gemany for better prices. smiley
Logged

Motivation is like a salt, once it has been dissolved it can react with things it comes into contact with to form something interesting. nerd
Pages: [1] 2 3   Go Down
  Print  
 
Jump to:  


Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Valid XHTML 1.0! Valid CSS! Dilber MC Theme by HarzeM
Page created in 0.31 seconds with 27 queries. (Pretty URLs adds 0.014s, 2q)