Just wondering what your thoughts on this are, given that at this time it seems as though there won't be any 256MB NV30's.
Doomtrooper said:Interesting you would post that now
Nagorak said:Aren't we getting to the point where, even with full FSAA, almost half the 256 MBs of memory would be completely unused most of the time? I think this is jumping the gun, just a little bit...
Althornin said:Nagorak said:Aren't we getting to the point where, even with full FSAA, almost half the 256 MBs of memory would be completely unused most of the time? I think this is jumping the gun, just a little bit...
6xFSAA at 1600x1200 uses how much memory?
(on a 9700)?
Even 4x uses enough to cause slowdows (AGP texturing) in some maps of UT2K3 at that res.
1600 * 1200 * 4 * 3 * 6 = ~131 MBs?
Tahir said:Yeah I hear from my sources 256MB is gonna be an option on R350.
No, really.
That calculation doesn't seem to count the memory needed for the Z-buffer. With a 32-bit multisampled Z-buffer, it becomesDave H said:So, with triple buffered 6xMSAA, that'd be 1600 * 1200 * 4 * (6 + 2) = ~59 MB. (I think. Can you get away with reusing any of the buffers and still have triple buffering?)
Those numbers sound so excessive as to be fake. I'm excited.MuFu said:Yeah - the 6x mode is apparently about as fast as the current 4x mode on the R300. 256MB RAM will be beneficial in certain situations. "Ultra high-end" AIBs only though.
That calculation doesn't seem to count the memory needed for the Z-buffer. With a 32-bit multisampled Z-buffer, it becomes
1600 * 1200 * ( 6*8 + 2*4 ) = ~103 MBytes.
Dave H said:That calculation doesn't seem to count the memory needed for the Z-buffer. With a 32-bit multisampled Z-buffer, it becomes
1600 * 1200 * ( 6*8 + 2*4 ) = ~103 MBytes.
Thanks for the correction. Aren't most Z-buffers still 24-bit, though? (That'd get you to ~92MB.)
OT but related: how exactly are framebuffer/z-compression working such that they save bandwidth yet don't save memory usage? Conversely, wouldn't hierarchical-Z require more memory space (in the same way that mipmaps take up more memory but reduce bandwidth utilization)?