AMD Phenom II 975 doesn't support sse3/4?

[abdo123]

And soo yeah you would die without having last word and showing yourself as the smartest guy under the sun(even if it would be night). :X

PS. PCSX2 doesn't "require" SSSE3 nor SSE4.1 if we talk about facts and truth.

we all know that Squall likes to be under the spot light

[Squall Leonhart]

And soo yeah you would die without having last word and showing yourself as the smartest guy under the sun(even if it would be night). :X

Got a problem?

PS. PCSX2 doesn't "require" SSSE3 nor SSE4.1 if we talk about facts and truth.

It does unless you want color bugs in some games with ATI gfx
http://forums.pcsx2.net/Thread-purple-wi...ure-glitch

[miseru99]

Nope no problem here, just stating the fact that you see problems everywhere, oh wait, you just did it again.

And I dunno how I can explain it to you on myself, heh maybe check this.

Edit: I don't like "feeding" Hulks nor other green creatures;P, soo reply with an edit - you DO NOT require those instructions to run PCSX2 - conditions to use that word here are fake and fail.:3

[Squall Leonhart]

application of the word Require is relative to the conditions that present.

[xstyla]

It does unless you want color bugs in some games with ATI gfx
http://forums.pcsx2.net/Thread-purple-wi...ure-glitch

RARE and rather small bug. I could fix that in gsdx if I'd have the game and would care about it. but this's all not the case.

bragging with intel is so lame. ;P

[nosisab Ken Keleh]

There is a bit of misunderstanding about the denormal feature. It's right Intel implemented in hardware the ability to calculate denormal numbers yet before the IEEE-754. But this is not the same as the DAZ instruction (Denormals Are Zero) which is part of SSE2 instruction set.

The point is calculating denormals is performance consuming and most of times meaningless to the needed purpose, so making them zero is a better approach and so, either Intel as AMD are the same in this aspect regarding PCSX2 which make them zero at once.

In time (to clarify): Intel would have advantage if the emulator indeed did calculate denormals instead deeming them too small to be meaningful, making them zero then.

PS: I'm not sure how the rounding mods work or if they are related to denormals someway, if so, for those few games which need it, Intel has an edge.

[Squall Leonhart]

There is a bit of misunderstanding about the denormal feature. It's right Intel implemented in hardware the ability to calculate denormal numbers yet before the IEEE-754. But this is not the same as the DAZ instruction (Denormals Are Zero) which is part of SSE2 instruction set.

The point is calculating denormals is performance consuming and most of times meaningless to the needed purpose, so making them zero is a better approach and so, either Intel as AMD are the same in this aspect regarding PCSX2 which make them zero at once.

In time (to clarify): Intel would have advantage if the emulator indeed did calculate denormals instead deeming them too small to be meaningful, making them zero then.

did you somehow miss

It has to do with the fact that AMD chose not to support "Denormals Are Zero" optimizations. Intel chose to support this feature of the IEEE 754 floating point standard, which is a mode the PS2 itself utilizes; and it yields significant speedups in many games.

[nosisab Ken Keleh]

I did not miss it, just the information is misguided. Intel implemented "calculating denormals" in hardware, not Denormals are Zero, which is a SSE2 instruction (shared by AMD also).

[Squall Leonhart]

Yes, because im sure Jake gets information wrong frequently.

[nosisab Ken Keleh]

Sorry, I did not understand the last comment. Making a mistake in a hurry post is something everyone is subject and does not diminish the person or his/her work.