Re: "They Just Have To Be Good"

Spalls Hurgenson wrote:

On Mon, 31 Mar 2025 19:10:07 -0000 (UTC), candycanearter07
<candycanearter07@candycanearter07.nomail.afraid> wrote:
 

Spalls Hurgenson <spallshurgenson@gmail.com> wrote at 14:28 this Saturday (GMT):

On Fri, 28 Mar 2025 15:50:06 -0000 (UTC), candycanearter07
<candycanearter07@candycanearter07.nomail.afraid> wrote:
>

Zaghadka <zaghadka@hotmail.com> wrote at 15:11 this Thursday (GMT):

On Tue, 25 Mar 2025 22:40:06 -0000 (UTC), in
comp.sys.ibm.pc.games.action, candycanearter07 wrote:
>

Zaghadka <zaghadka@hotmail.com> wrote at 18:30 this Sunday (GMT):

On Sun, 23 Mar 2025 06:10:04 -0000 (UTC), in
comp.sys.ibm.pc.games.action, candycanearter07 wrote:
>

I think the GPU price is more because of the AI bubble rn..

>
Crypto followed by AI. Nvidia fell into Scrooge McDuck swimming gold.

>
Yeah, wasnt NVIDIA the biggest stock for a bit?

>
Still is doing quite well, in fact. Until the bubble bursts.
>
I wish I had gotten in at the ground floor.
>
It might _still_ be beneficial to get in, but the dramatic upside that
existed before crypto found its stride is, IMO, done. I don't have the
"intestinal fortitude" to time the bubble burst window and get out.
>
I think Nvidia's fortunes will end in an equally dramatic, and perhaps
sudden, downside. It's not a "buy" for me because its market valuation is
nearly insane, IMO.*

>
>
And I don't use NVIDIA because of the Linux stuff :)

>
I've never understood Nvidia's reluctance to open-source their
drivers. Especially since increasingly their core users --not gamers,
but crypto & AI bros-- aren't Windows based. Even the much-ballyhoo'd
"we've open sourced a few drivers" from a few years ago was largely
smoke and mirrors. Is so much of Nvidia's functionality built into
their drivers that they dare not release the code?

>
>
Who knows. My bet is they're either trying to hold onto "security
through obfuscation" or are being forced to by a contract..

 The most /convincing/ reason I've heard is that, with open-source
drivers, users would be able to turn consumer-level cards into
workstation-level cards, accessing features normally restricted by
software. Why buy a $3000 workstation card when a $500 card and some
hacked drivers get you the same capabilities?
 Whether this is true or not, I've not the foggiest. But it seems more
likely than problems with patents or trade-secrets (since you could
re-write your drivers to avoid that). It's not like AMD and Intel
haven't faced similar problems.
 I wonder if perhaps Nvidia is like some of those Japanese software
companies (Konami, Sega, Capcom, etc.) who were so scarred by past
events that even the idea of opening up is unthinkable. Capcom etc.
were fiercely embattled by pirates and hackers ripping off their
arcade boards and ROMs in the '80s, and I think a lot of this is why
they've such a negative view of user-mods today. Similarly, if there
ever was any truth to "drivers+consumer cards=workstation capable"
rumor, I can imagine Nvidia being quite loathe to even think of going
down a similar route.
 But as far as I can tell, Nvidia has never given a clear explanation
as to why they won't open-source their drivers, beyond the usual
corp-speak (e.g., "it's not in our interests today").
 

They would be easier to hack for good or evil.  The company could grow lax relying on independent coders to improve their drivers too much. Errors would be pointed out making the company suffer.  And lastly, China would have a whole bunch of grand new video cards.
Open source does very little to help the company making it, and is really only a marketing scheme to show 'enlightenment' to the human condition.
If I had Nvidia stock and I heard they were going open source I would sell it all off.  But with what you report, I'm going to buy a couple shares when April finally gets here.  And I consider myself a savvy investor, with my only weakness in the stock market being volume.
--
Wears a paper bag like ab oxygeb mask
till your mibd starts to gel
'Cause the ball in the can has a crazy beat
The funky dyin brain cell