quadro vbo issue

I thought that changing immediate mode into retained mode make OpenGL SW significantly faster, but actually, it only affected to geforce, not quadro. I want to know the reason of this difference.

I’m using both quadro FX 570 and geforce 650 TI graphic cards.

My original OpenGL SW was using old immediate mode style rendering like below.

glEnableClientState(GL_COLOR_ARRAY);
glEnableClientState(GL_VERTEX_ARRAY);

glColorPointer(3, GL_FLOAT, 0, colors);
glVertexPointer(3, GL_FLOAT, 0, vertices);


glDrawElements(GL_TRIANGLES, (VERTEX_NUM-2)*3, GL_UNSIGNED_INT, indices);
	
glDisableClientState(GL_VERTEX_ARRAY);
glDisableClientState(GL_COLOR_ARRAY);

and I changed it using VBO like below ( buffer has made before )

glBindBufferARB(GL_ARRAY_BUFFER_ARB, vboId);
glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, vboId2);
	
glEnableClientState(GL_COLOR_ARRAY);
glEnableClientState(GL_VERTEX_ARRAY);
glEnableClientState(GL_INDEX_ARRAY);

glColorPointer(3, GL_FLOAT, 0, (void*)(sizeof(float) * VERTEX_NUM * 3));
glVertexPointer(3, GL_FLOAT, 0, 0);
glIndexPointer(GL_UNSIGNED_INT, 0, 0);

glDrawElements(GL_TRIANGLE_STRIP, (VERTEX_NUM-2) * 3, GL_UNSIGNED_INT, 0);

glDisableClientState(GL_VERTEX_ARRAY);
glDisableClientState(GL_COLOR_ARRAY);
glDisableClientState(GL_INDEX_ARRAY);

On the Geforce 650 Ti, FPS has increased about three times, rendering exactly same scene, but on the Quadro FX 570, FPS doesn’t changed at all.

May I have the answer for this situation?

Most likely because your application was CPU limited before and now it’s less of that but limited by another bottleneck on the Quadro FX 570.
Mind that the Quadro FX 570 is using a three(!) GPU generations older architecture (G84GL vs. GK106).
If you look at the raw hardware differences in clocks, memory bandwidth, and shader cores you’ll get an idea of the HW performance delta between the two.
(BTW, do not forget to disable vsync in the control panel when doing OpenGL benchmarks or you might be limited by the monitor refresh rate.)