End of year Xorg status rant
gurenchan at gmail.com
Sun Jan 1 06:53:30 UTC 2017
In the past nvidia had closed sourced drivers in my opinion because they
used public papers on gpu rendering techniques, built it into hardware and
sorta duck taped it all together, amd sorta did the same thing as well so
they were both guilty but nvidia got the lead and stayed ahead.
Those reasons are going to be irrelevant in the vulkan dx12 years.
What do I mean by that? nvidia always had a major gpu binary release around
the time of major software release because nvidia would take the code from
the game devs try it on their machines and would basically go in and
rewrite shaders, fix edge case bugs, etc to make sure the next big game
always ran on nvidia hardware, amd didn't have these resources so they were
always sorta catching up.
The product releases, people are praising nvidia for their awesome
performance and amd get's the shaft in the public's eye.
This had a lot to do with how previous generations opengl/ directx had huge
driver layers that did all kinds of things to help with memory management,
in vulkan, dx12 land that's no longer the case because there is no driver
layer to speak of. So it's my opinion that both chip makers will be more
open source friendly because now if a developer writes crap programs it
will just kill the card.
There's no place; the driver implementation; for chip makers to go and do
their magic. That responsibility falls on the developers shoulders and I
think this is a great thing.
Let both companies make the best hardware and have some base standard api
to talk to the hardware, everything else is game.
The reason vulkan took so long to release was because of this battle,
nvidia understood what vulkan would mean; the ______ amount of dollars they
invested in the driver team, outreach to make sure you get help from nvidia
on your game, all those resources would be gone, while that's not great for
nvidia it's good for the consumer and developers.
Make the hardware and let people use it, stop selling gimped hardware and
gating things such that you can charge a premium for non premium cards.
I just migrated to a new laptop with a gtx1070 but i will also be getting a
laptop with a vega card when those are available. I don't care who wins, if
either "team" wins we all lose. They should worry about creating great
products and good documentation and let the people do what they want with
the hardware that they bought.
That's my post new years rant.
On Sun, Jan 1, 2017 at 11:42 AM, Matthew Macy <mmacy at nextbsd.org> wrote:
> ---- On Sat, 31 Dec 2016 18:47:23 -0800 Thomas Mueller <
> mueller6722 at twc.com> wrote ----
> > excerpt from O. Hartmann:
> > > I think we face a political problem, not so much a man-power-driven
> one. nVidia provides
> > > a BLOB, this BLOB works well even with the most recent hardware of
> theirs, but it lacks
> > > in support for OpenCL and their own CUDA acceleration framework. I
> never understood why.
> > > I asked nVidia - and they told me, that there is no request from the
> community ... so
> > > far. That is a claim and I can not hold something against it, since
> it seems obvious
> > > that I'm, with some other single people, are the only one compared to
> millions of others
> > > - de facto Null so to speak.
> > > And AMD? Well, 2006 or 2008 the company claimed to support the
> opensource community
> > > better than before, but that left in history to be a insubstancial
> claim. Their hardware
> > > might be a great deal even for GPGPU purposes with OpenCL, but this
> is Linux only as far
> > > as I can tell.
> > I too noticed that nVidia and AMD were not open-source-friendly with
> their graphics; remember Linus Torvalds' comments and cusses regarding
> > That's why I chose Intel on my last computer hardware purchases; I use
> Intel on-CPU graphics.
> Yes, we can leave it to AMD to snatch defeat to defeat from the jaws of
> Nonetheless, companies management turns over and their policies change.
> AMD is *currently* very much committed to bringing their open source Linux
> driver up to par with what they have on Windows. They are also committed to
> a fully open source high performance GPGPU compute stack. They should be
> rewarded for what their *current* policies are if we want to encourage
> that. If they again back off from that stance - then yes, leave them to
> their own devices.
> freebsd-x11 at freebsd.org mailing list
> To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org"
More information about the freebsd-x11