Phun physics benchmark
3 posts • Page 1 of 1
Phun physics benchmark
by niinja2 » Wed Feb 29, 2012 9:06 pm
I have decided to determine what kind of CPU ( amd or intel ) is better for running phun, specifically colliding large number of particles (gas). And i don't want to only to determine who has the "best" system, but to determine:
1 How big is the difference between different CPU families
2 How much cpu, ram and NB frequency influences it.
Results that were obtained by running phun from USB are not valid.
Benchmark and post your results here or on the places where you downloaded benchmarks from:
1 Score (must)
2 Cpu type and speed(must)
3 Version of phun/algodoo(must)
4 Ram speed
5 NB speed
6 Operating system
The problem until now is that phun does not have any built in benchmark and any other benchmark has a scale of scores that is either nonlinear or nobody knows how to exactly interpret results, because 400% vs 200% result in some benchmark does not have to mean double the performance.
So far i have created 2 benchmarks. The benchmarks are nothing special when considering how "fun" they are, nor should they be, they are benchmarks. So i have spend more time tweaking them and making them accurate and reliable representation of cpu power, rather then making them fun or fancy.
First one measures how fast can you run 400 particles at frame rate of 40 FPS.
It scales almost perfectly linear.
I will later discuss and interpret the results in this benchmark, but for now 2 times higher score means 2 times stronger machine.
My second benchmark measures how many particles can you run at 40 FPS. It adds more particles as the sim goes on.
Its scaling is not linear, it suffers diminishing returns in the following manner
From these pictures you can roughly see what the results you get actually mean. But this will be discussed later on below.
1 How big is the difference between different CPU families
2 How much cpu, ram and NB frequency influences it.
Results that were obtained by running phun from USB are not valid.
Benchmark and post your results here or on the places where you downloaded benchmarks from:
1 Score (must)
2 Cpu type and speed(must)
3 Version of phun/algodoo(must)
4 Ram speed
5 NB speed
6 Operating system
The problem until now is that phun does not have any built in benchmark and any other benchmark has a scale of scores that is either nonlinear or nobody knows how to exactly interpret results, because 400% vs 200% result in some benchmark does not have to mean double the performance.
So far i have created 2 benchmarks. The benchmarks are nothing special when considering how "fun" they are, nor should they be, they are benchmarks. So i have spend more time tweaking them and making them accurate and reliable representation of cpu power, rather then making them fun or fancy.
First one measures how fast can you run 400 particles at frame rate of 40 FPS.
It scales almost perfectly linear.
I will later discuss and interpret the results in this benchmark, but for now 2 times higher score means 2 times stronger machine.
My second benchmark measures how many particles can you run at 40 FPS. It adds more particles as the sim goes on.
Its scaling is not linear, it suffers diminishing returns in the following manner
From these pictures you can roughly see what the results you get actually mean. But this will be discussed later on below.
Last edited by niinja2 on Wed Feb 29, 2012 9:19 pm, edited 1 time in total.
- niinja2
- Posts: 3
- Joined: Wed Feb 29, 2012 6:43 pm
Re: Phun physics benchmark
by niinja2 » Wed Feb 29, 2012 9:08 pm
Reserved for benchmark data interpretation!
Last edited by niinja2 on Wed Feb 29, 2012 9:11 pm, edited 1 time in total.
- niinja2
- Posts: 3
- Joined: Wed Feb 29, 2012 6:43 pm
Re: Phun physics benchmark
by niinja2 » Wed Feb 29, 2012 9:10 pm
Reserved for CPU family analysis and conclusions!
- niinja2
- Posts: 3
- Joined: Wed Feb 29, 2012 6:43 pm
3 posts • Page 1 of 1
Who is online
Users browsing this forum: No registered users and 8 guests
