6950 Poor Vector Performance

[Spydell]

Just run a 2d performance test and noticed that my 6950 has really bad performance in the vector tests and wondered why it performs so badly in
Solid vectors
transparent vectors
and complex vectors
Also not scoring so well in the fonts and text test.
Is this an amd thing or both nvidia and amd cards that perform so badly in this area.
Or is it related to just my card or just 6950's in general.
Can anyone else grab a hold of this test and see how there cards square up.

 

[Geckovich]

What are you comparing your results to? If you've no point of reference with results from any other card how do you know you're getting low performance?

More importantly, does it even matter? If the card is running things OK then nothing to worry about. A £20 card will have adequate 2D performance, let alone a £200 card

 

[Gammawolf]

I would give you results if the program would function on my system but it does not.

 

[paradigm]

Certainly worse than mine, but vector performance still sucks here:

 

[kae2]

I seem to remember reading that modern cards perform worse due to their design (optimised for everything else other than 2D work) as the design of a 2D card is drastically different. I think it also said that an old 2D card would outperform a modern one in certain scenarios.

Rather foggy recollection of something I must have read years ago though, could be talking complete twoddle!

 

[Spydell]

What are you comparing your results to? If you've no point of reference with results from any other card how do you know you're getting low performance?

More importantly, does it even matter? If the card is running things OK then nothing to worry about. A £20 card will have adequate 2D performance, let alone a £200 card

Im comparing my results to the chart you see above you.
solid vectors 1.60 out of a possible 10
transparent vectors 1.72 out of a possible 10
complex vectors 109.1 out of a possible 200
Does it matter? probably not but that still doesn't stop me wondering why a £200+ card presumably scores so low in those area's.
Also in case you didn't notice i asked others if they could post there results as another comparison.

 

[Spydell]

Certainly worse than mine, but vector performance still sucks here:

Thanks.
Just noticed that your bars go higher than mine, mine goes to 400 where as yours goes to 600 and so on.

 

[paradigm]

One simple change (Aeroglass OFF) makes a huge difference:

 

[Geckovich]

Im comparing my results to the chart you see above you.
solid vectors 1.60 out of a possible 10
transparent vectors 1.72 out of a possible 10
complex vectors 109.1 out of a possible 200
Does it matter? probably not but that still doesn't stop me wondering why a £200+ card presumably scores so low in those area's.
Also in case you didn't notice i asked others if they could post there results as another comparison.

I wasn't having a go, it just seemed that you were assuming your performance was low based on the size of the graph.

My point being that low on the chart may not mean low performance.

 

[Spydell]

One simple change (Aeroglass OFF) makes a huge difference:

Wow that's good to know,
i wonder how this would compare when disabling areo while running games?
Do you think it would have an impact on performance?

 

[paradigm]

Wow that's good to know,
i wonder how this would compare when disabling areo while running games?
Do you think it would have an impact on performance?

In theory, Aero on/off should make no difference to 3D performance, as the card will change from 2D clocks to 3D clocks to game with.

The problem with Aeroglass is that the card is still in 2D mode, even though it needs DX9 shaders to achieve the transparency effects.

 

[Spydell]

I wasn't having a go, it just seemed that you were assuming your performance was low based on the size of the graph.

My point being that low on the chart may not mean low performance.

Ok,
but generally speaking if i was to run a benchmark and the graph was from
0 to 10
and i scored 1
I would automatically think i was getting low performance regardless of what card i had, but even more so if i had a top card.
Catch my drift.

 

[Geckovich]

Ok,
but generally speaking if i was to run a benchmark and the graph was from
0 to 10
and i scored 1
I would automatically think i was getting low performance regardless of what card i had, but even more so if i had a top card.
Catch my drift.

Yep, I entirely see where you're coming from.

If I'm honest I think we can pretty much discount this test as being accurate, I just ran it and these are my results:

Completely different to yours and Paradigm's.

So that's 3 people with the same GPU all getting different scores.

 

[Spydell]

Yep, I entirely see where you're coming from.

If I'm honest I think we can pretty much discount this test as being accurate, I just ran it and these are my results:

Completely different to yours and Paradigm's.

So that's 3 people with the same GPU all getting different scores.

Ok great thanks for putting in the time to post your results.