• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

SSE4 questio

B

BinarySynapse

[H]F Junkie
2FA
Joined
Feb 6, 2006
Messages
13,984
I was always under the impression that each SSE revision had to be specifically implemented in software before the new instructions could provide a performance boost (meaning, SSE2 programs wouldn't benefit from SSE3 until the developer recompiled to actually use the new instructions in SSE3).

Is this not the case, or is it just not the case for SSE4? The reason I asked is because I see a bunch of people saying the Q9xxx series will kick the Q6xxxx series because is has SSE4 (among other reasons).
 
ryan_975 said:
I was always under the impression that each SSE revision had to be specifically implemented in software before the new instructions could provide a performance boost (meaning, SSE2 programs wouldn't benefit from SSE3 until the developer recompiled to actually use the new instructions in SSE3).

Is this not the case, or is it just not the case for SSE4? The reason I asked is because I see a bunch of people saying the Q9xxx series will kick the Q6xxxx series because is has SSE4 (among other reasons).
Click to expand...

You are correct, it does need to be implemented in software. There are a few programs out now that support it, but I think people overestimate SSE4 to the point that they make it sound like it will triple the power of the processor. Of course it will help, but it's not the second coming either.
 
u r right , sse4 has to be implemented in software b4 the software can take advantage of it . my guess would be someone heard from some misinformed fellow that sse4 is the reason y q9 quads r better than q6 , and more people heard from them , so thats y u hear them saying that sse4 is the reason behind q9's success .

but its not so , its because of better architecture of the q9's quads . so even the application with sse2 see a 10% increase on q9 quads . last time i checked only one software supported sse4 that was divx codec (that too experimental) .another was tmpgenc express (i'm not sure) .

as more software's implement sse4 , the performance difference will be even greater .
 
wolf2009 said:
u r right , sse4 has to be implemented in software b4 the software can take advantage of it . my guess would be someone heard from some misinformed fellow that sse4 is the reason y q9 quads r better than q6 , and more people heard from them , so thats y u hear them saying that sse4 is the reason behind q9's success .

but its not so , its because of better architecture of the q9's quads . so even the application with sse2 see a 10% increase on q9 quads . last time i checked only one software supported sse4 that was divx codec (that too experimental) .another was tmpgenc express (i'm not sure) .

as more software's implement sse4 , the performance difference will be even greater .
Click to expand...

better architecture? Isn't it the same plus a few tweaks.
 
its different transistors , high-k transistors . and a few tweaks .
 
some of the SSE instructions are used by the OS, too. So when an update to an OS comes out sometimes those added instruction can help out everything.
 
You must log in or register to reply here.
Back
Top