Jump to content

What about the new ATI HD4850 ?


Soner
 Share

911 posts in this topic

Recommended Posts

This is fantastic news!

Do macs support dual monitors very well?

I use Ultramon on Windows XP and it's really quite nice, can OSx still handle 2 screens well? I hope so.

 

I have an ATI4870, Intel P45 chipset, Intel E8500 C2D CPU, SATA drives, onboard audio (Realtek?) and Marvel Gigabit ethernet, fingers crossed.

Link to comment
Share on other sites

not as amazing as could be

 

1zme9fs.png

 

glview isnt best opengl test tho

btw, enabling 4xAA doesnt drop performance (jsut 15-20 points drop)

 

currently works only on snow 10a222 build

 

and has seriours problems with Hz for display, e.g. if i set 60, it sets 80..

 

4kirus.png

Link to comment
Share on other sites

not as amazing as could be

 

1zme9fs.png

 

glview isnt best opengl test tho

btw, enabling 4xAA doesnt drop performance (jsut 15-20 points drop)

 

currently works only on snow 10a222 build

 

and has seriours problems with Hz for display, e.g. if i set 60, it sets 80..

 

4kirus.png

 

Sweeet! but atleast it works some what hehehe

Link to comment
Share on other sites

not as amazing as could be

 

1zme9fs.png

 

glview isnt best opengl test tho

btw, enabling 4xAA doesnt drop performance (jsut 15-20 points drop)

 

currently works only on snow 10a222 build

 

and has seriours problems with Hz for display, e.g. if i set 60, it sets 80..

 

Netkas: ... not as amazing as could be

 

No, for such very fast+new GPU i would ever use an FSAA *4 or better FSAA*8 to compare it with others.

 

You(& we :( ) will see much more the REAL! GPU-Hardwareperformance differences and the benchvalues are much less CPU / Driver dependend - also reason why GPU Tests in magazines always use FSAA*4 , sometimes even FSAA*16 for X2 cards to see real gpu perf diffs and not the CPU/Driver Limitation.

WORST CASE for GPU Test is XBENCH: for newer cards (> 2001) nearly only an CPU/L2-Cache/Systembus test running the Xbench OpenGL-"Bench".

 

NETKAS, would it be possible to post an FSSA*8 GLView Bench ?

 

My ATI 3850, i measured now gets (FSAA*8, 1600x1200 (i have 4:3 CRT, x1000 would be about 15% better)

Running FSAA*4, like you (you said lossing about 25 points= around 2400 points FSAA*4) gets only around 490 points!!!

Your ATI 4870 is in that case (FSAA*4) 4 times FASTER than my (good!) 3850 GPU - thats a LOT!

 

My 3850 looses by FSAA*8 a lot of performace: from around 1900 (FSAA*0) over 490(FSAA*4) to only around 380 (FSAA*8). Slower GPUs like ATI 2600 XT, NV 7600GT will loose much more using FSAA*8 !

I guess/bet: Will be BIG!!! difference to ATI 4870 FSAA*8 (Pre Snow, Pre patched!!)

Bild_127.jpg

Link to comment
Share on other sites

cod4 mac version, 1680x1050 4XAA 16AF, all maxed, benchmark gives 34fps ( in game its mostly 40-50 fps) , while on nvidia same config gives slideshow, cuz of aa :blink:

 

now glview, on 1680x1050 it has different behavior with aa.

 

153350y.png

a4u6gl.png

2iw2g5i.png

2qnpe1l.png

Link to comment
Share on other sites

cod4 mac version, 1680x1050 4XAA 16AF, all maxed, benchmark gives 34fps ( in game its mostly 40-50 fps) , while on nvidia same config gives slideshow, cuz of aa :blink:

 

HAHAHAHAAH ROFL :D

Link to comment
Share on other sites

cod4 mac version, 1680x1050 4XAA 16AF, all maxed, benchmark gives 34fps ( in game its mostly 40-50 fps) , while on nvidia same config gives slideshow, cuz of aa :blink:

 

now glview, on 1680x1050 it has different behavior with aa.

 

153350y.png

a4u6gl.png

2iw2g5i.png

2qnpe1l.png

 

 

Give me snow leo!!! :D

Link to comment
Share on other sites

 Share

×
×
  • Create New...