New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
I sometimes run machine learning on very big data. For me, runabove is the best option in terms of cost efficiency.
However, their plan might not meet your needs. They have 6 cores > 8 threads > 176 threads which skips the range of # of cores you want.
BTW, I don't know much about c4d, but if you are 3D rendering, shouldn't you be looking for powerful GPU rather than CPU cores?
Amazon probably has the cheapest instances (they provide GPU containers).
C4D does not have built-in gpu rendering. You have to buy third party solutions for GPU rendering.
GPU rendering has the big advantage that you can basically have an army of cuda cores and shaders which on the other hand are not as intelligent as CPU cores which is why they cannot perform every task reliable.
But quality of GPUs and software have improved much and I am planning to include GPU very soon.
Well, you know the solution, then. No need to look for CPU cores :P
I've used the Vultr 16 and 24 core servers before to run some video processing stuff for a friend, worked out way under his budget actually.
If you haven't already, try using the aws spot instances instead of the normal ones - the gpu/high men/cpu instances are very cheap that way. The only issue I had was with the cpu instances not having enough memory for some scenes with higher texture resolutions/etc
You are brillant dude, I didn't think of this. I need high CPU instances quite often these days and will use it for sure.
Update:
I benchmarked the spot EC2 instances and vultr instances and I have to say that vultr definetely has the best value-for-money ratio.
You get much more cpu power for less money compared to the EC2 Windows instances and you have more than enough free traffic included which is very important when you have to transfer all the rendered images. EC2 traffic is very expensive.
I didn't test the support with EC2. I had a small issue with vultr and my ticket was answered within 3 minutes resolving my issue.
@DaveA
Thank you dude, I will recommend your services to my friends. Keep up the good work!
And no guys I don't work for them, I'm just very satisfied with this service and wanted to publish my experiences.
What render engine do you use?
I've tried some computational stuff on EC2 and Azure (which is comparable to EC2 and yes, you can run Linux VMs) but wasn't impressed with the performance. CPUs were 2Ghz something. A desktop i5 was so much faster, even considering the huge RAM available on the cloud instance. But YMMV.
Side note:
I've not used C4D. I use Poser periodically and by far the fastest rendering on the planet is to buy a big video card and render in it - e.g., with the Reality plug in. With Reality you can do real-time final renders (yes, as good as the stuff your CPU cranks on for hours). But the catch is that everything - models, textures, etc. - has to fit in memory and 4GB video cards are mighty pricey.