Question about benchmark time of optimized debug build

The Partridge Family were neither partridges nor a family. Discuss.
Post Reply
WilsonHuang
Posts: 22
Joined: February 13th, 2019, 3:23 am

Question about benchmark time of optimized debug build

Post by WilsonHuang » December 18th, 2019, 2:29 am

The mean benchmark time of surface optimized debug build run on my computer is around 322 ~ 400 ms, which is much slower than chili's (his benchmark time is 54.2862ms). Why there are huge time gap between mine and chili's, and what should I do to improve the benchmark time? Upgrade my PC?

Here are some of my benchmark time results (Debug):
Sample size: 5
Mean: 322.418ms
Min: 303.895ms
Max: 341.248ms

Sample size: 5
Mean: 333.178ms
Min: 302.982ms
Max: 395.975ms

Sample size: 5
Mean: 399.676ms
Min: 397.759ms
Max: 401.475ms

albinopapa
Posts: 3982
Joined: February 28th, 2013, 3:23 am
Location: Oklahoma, United States

Re: Question about benchmark time of optimized debug build

Post by albinopapa » December 18th, 2019, 2:42 am

Was the test code something you followed along with or downloaded from GitHub?
Last I knew chili's proc was an old Intel server CPU, so cache speed/size and such may have given him an advantage in that regard.

Tell ya what, upload a copy of your project to the forum and I'll run it and see what I get using your copy.

I'm running an AMD Ryzen 2400G, so nothing fancy. If I get better than 322ms then it's time to upgrade, if not then there may be something you missed ( if it's follow along code ) during the tutorials.
If you think paging some data from disk into RAM is slow, try paging it into a simian cerebrum over a pair of optical nerves. - gameprogrammingpatterns.com

WilsonHuang
Posts: 22
Joined: February 13th, 2019, 3:23 am

Re: Question about benchmark time of optimized debug build

Post by WilsonHuang » December 18th, 2019, 5:31 am

albinopapa wrote:
December 18th, 2019, 2:42 am
Was the test code something you followed along with or downloaded from GitHub?
Last I knew chili's proc was an old Intel server CPU, so cache speed/size and such may have given him an advantage in that regard.

Tell ya what, upload a copy of your project to the forum and I'll run it and see what I get using your copy.

I'm running an AMD Ryzen 2400G, so nothing fancy. If I get better than 322ms then it's time to upgrade, if not then there may be something you missed ( if it's follow along code ) during the tutorials.
Well, the test code is I downloaded from GitHub. I also follow alone the tutorials, but the benchmark time has no difference (still around 322 ~ 400 ms)
The attachment is my project.
Attachments
Sprite.zip
(748.63 KiB) Downloaded 9 times

albinopapa
Posts: 3982
Joined: February 28th, 2013, 3:23 am
Location: Oklahoma, United States

Re: Question about benchmark time of optimized debug build

Post by albinopapa » December 18th, 2019, 5:04 pm

Code: Select all

Sample size: 5	
32 run average Mean:	107.699156
32 run average Min:		102.358531
32 run average Max:		114.564938
If you think paging some data from disk into RAM is slow, try paging it into a simian cerebrum over a pair of optical nerves. - gameprogrammingpatterns.com

albinopapa
Posts: 3982
Joined: February 28th, 2013, 3:23 am
Location: Oklahoma, United States

Re: Question about benchmark time of optimized debug build

Post by albinopapa » December 18th, 2019, 8:09 pm

I wouldn't fret too much. At this point in the tutorials, you are just learning to program. If you have a slower computer, it will make it more interesting when you make games or other applications. If it runs well on a slower computer then it will run really well on a faster computer.

You don't need a high end computer to learn to code. Sure it make the results more enjoyable, but you can worry about that later. For now, just focus on the learning aspect. Also, I'm sure your computer would work just fine once compiled in Release mode for most of the tutorials.
If you think paging some data from disk into RAM is slow, try paging it into a simian cerebrum over a pair of optical nerves. - gameprogrammingpatterns.com

Post Reply