Re: Weird figures with OpenGL 3.2 and Intel HD3000

Hej Ingemar, Hi Justin,
thanks for your answers!

> 400000 triangles is a lot, but when 15000 takes the same time, there is 
> another bottleneck somewhere. What happens for very small models? Still 10 
> ms? How about drawing several? Maybe the overhead is somewhere else?

I have to try.

> I can try the same program on a 9400 if you like, to see if it is the same 
> for other GPUs.

When I run the very same code on my late 2008 old MacBook, with the integrated 
NVidia 9400M chip (which is not supposed to be a racing horse), the same call 
to glDrawElements takes *20 µs* on average; that’s much less that the customary 
call to glClear (GL_COLOR_BUFFER_BIT) that consumes around 90 µs.
(BTW, I made a mistake in my first message. On the HD3000, it takes 10 ms, not 
10 µs).

>> You are rendering quite a bit of geometry, do you have another
>> benchmark to test it against to see how your Mac performs with that?
>> My experience with intel graphics is not great and It wouldn't
>> surprise me if you are maxing it out.

Justin, while I can conceive Intel HD3000 performance to be on par or slightly 
inferior to a 3+ years old NVidia 9400M, here the difference is about three 
orders of magnitude, which makes me think there is a hidden bug somewhere in 
the driver (?). It is as if none of the two VBO (position, indices) and none of 
the texture were actually cached in the GPU memory, and all had to be 
transferred anew at each frame.

Very strange indeed.

Anyhow,
thanks Justin,
tack så mycket Ingemar,

Vincent

 _______________________________________________
Do not post admin requests to the list. They will be ignored.
Mac-opengl mailing list      ([EMAIL PROTECTED])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/mac-opengl/alexiscircle%40gmail.com

This email sent to [EMAIL PROTECTED]