Desktop hardware configuration Linux Mint Cinnamon & Blender

Hello,

I am completely new to Linux. I wanted to buy a desktop computer mainly to run Blender. I was leaning towards Linux Mint Cinnamon.

My local computer assembler guy has never installed or ran Mint or Blender. So he wanted to know if anyone can confirm if the following desktop hardware configuration would work:

Intel G4560 7th gen 3.5Ghz
Motherboard Chipset Intel H110- HDMI
Video Graphics Card Nvidia GeForce GT1030 2Gb DDR5
RAM 4GB DDR4

I know this is a simple configuration but it’s what fits my budget. More expensive choices do not, so it’s useless to tell me to go buy other more expensive components.

The main question for me is if this hardware configuration will run Linux Mint Cinnamon and Blender without problems, if anyone has ever done it before…

thanks and I hope this is a good forum to ask this

That system should run fine, though you won’t be able to render large scenes which is limited by gpu and system memory.
One thing to keep in mind is that the gt1030 doesn’t support hardware video ENcoding.

Thanks. If things go well, I’ll upgrade later.

Before you go nuts afterward, three working procedures to have cuda and a current driver:
[url]https://devtalk.nvidia.com/default/topic/1032507/cuda-setup-and-installation/nvidia-driver-and-cuda-installation-sequence-/post/5252949/#5252949[/url]
[url]https://devtalk.nvidia.com/default/topic/1031213/linux/problem-installing-nvidia-390-42-driver-on-ubuntu-16-04/post/5247202/#5247202[/url]
You should never use the .run installer unless you know what you’re doing.

Fun times ahead!

I was trying to understand what this means: “gt1030 doesn’t support hardware video ENcoding.” I found a couple of pages like wikipedia on “hardware video ENcoding” but it’s still not clear what this would mean for working with Blender. It means the CPU will do most of the rendering and calculations?

Your comment did prompt me, however, to take a look at the gt1050 cuda information and price, and I should be able to afford it at this time. From the little I understand, for Blender, it’s really good to have this CUDA thing working, right?

HW gpu video encoding has nothing to do with blender, it’s useful for stream casting etc, e.g. chromecast, steam link, video chat. Just wanted to let you know before you wonder why it doesn’t work afterwards. Of course , those still support hw decoding (for watching video/streams).
And yes, you want to have CUDA working for blender, it lets you offload render computation to the gpu which might be faster than the cpu. It’s the main reason for having a nvidia gpu with blender.
When thinking about an upgrade to the gpu, imho you shouldn’t look at the 1050 since it also only has 2GB, limiting size of scenes. E.g. the 1050ti has 4GB. For starters, the 1030 is allright.
Beware: due to shortage of GDDR memory new models are on the market with DDR4 memory which are slower. So always read the data sheet.

You’ll run out of memory very, VERY quickly with only 4GB system RAM installed. Consider 8GB an absolute minimum; 16GB highly recommended.

Maybe some numbers:
i5-4430 CPU @ 3.00GHz
GT1030
8GB System memory
Blender 2.79

Demo fishy cat
Total sysmem used: 3.6GB
CPU render time 1:05 min
GPU render time 2:11 min
Total GPU mem used: 920MB

Demo classroom
Total sysmem used: 1.8GB
CPU render time 0:47 h
GPU render time ~3h
Total GPU mem used: 820MB

So the GT1030 doesn’t yield anything regarding render time, any recent quad-core is faster.

Ok, just the defaults were bad. Optimized run, fishy cat GPU 1:08min, classroom GPU 35:37min.