If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
My apologies for misunderstanding. I was reading a thread, (http://www.nvnews.net/vbulletin/showthread.php?t=72858) to be exact, where this is discussion about a lack of 2d acceleration for certain features. AA was also mentioned. Many people seem to think that cards from two generations prior are faster than even the latest nvidia cards in 2d.
What you said about how it is toolkit specific makes perfect sense to me, but I guess at the lowest level I thought it all boiled down to driver implementation - as in the way that the toolkit ends up sending information to the driver. It would seem that acceleration of additional features would improve performance in various areas of 2d including fonts.
Sorry if I derailed the thread.
I was referring to the thread you mentioned, not to you specifically. The driver implementation if a linux kernel object, which does nothing else than operate on the hardware level. I'm sure nVidia could somehow come up with a hack for X to accelerate certain parts, and that would possibly satisfy the starter of that thread, but it is far from an ideal solution - in my opinion, and I hope it'll never come to that. Perhaps, in the future, if the entire X server depends on the GPU, it will be an attractive solution.
Okay, a bit more information to throw out there today... Most of the information in this article is based solely on NVIDIA's past history, so accept these thoughts as you wish.
If you measure the time from NVIDIA's first public Linux display driver release (May 17, 2001; 1.0-1251) to their most recent non-legacy driver (May 22, 2006; 1.0-8762), and graph it, you will get something like:
Pardon the version numbers on the bottom getting cut off. This visual pretty much reiterates the long/short release cycle of a few months and then a matter of weeks for a follow-up release. If you base their release cycle on all of the drivers in the past, the average number of days between releases is 70. Coincidently, today is 70 days since the 1.0-8762 driver release. However, if you simply count the time between the major driver releases in the 1.0-7XXX and 1.0-8XXX, there is on average 92 days for the major/longer driver release. With that said, the resulting date based upon NVIDIA's past would push this next driver release to August 22, 2006.
Based upon this NVIDIA release history, as well as a few other tid-bits of information learned from various people in the industry, NVIDIA's inaugural 1.0-9XXX driver release will likely occur within a week of August 23, 2006. This estimated date hasn't been officially confirmed by NVIDIA, so accept this information at your discretion.
The Visual Computing System is compatible with Linux (32-bit and 64-bit), and with that it would be logical to bring the new display drivers. NVIDIA is stating this VCS will begin shipping in September, while The Inquirer is reporting 3 weeks from now.
With this Quadro Plex now, hopefully they will improve the quality of Linux SLI...
I'm still trying to get the August 23rd (+/- a week) Linux driver date confirmed with sources close to NVIDIA. Hopefully some more information will come out during Siggraph.