196.75 drivers fail it

Tom Lake wrote:

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only* those
3 Blizzard games. Bit too wierd for coincidence, no ? Thankfully I
play none of them, and the drivers are blazing fast. Keepers for me,
fix or not.

My .75 revision kept the GPU temp at 151 C. After reverting to the earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

I'm sure you mean Fahrenheit...


--
Norman
Registered Linux user #461062

Ian D wrote:
"Tom Lake" wrote in message
...
"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only* those 3
Blizzard games. Bit too wierd for coincidence, no ? Thankfully I play
none of them, and the drivers are blazing fast. Keepers for me, fix or
not.
My .75 revision kept the GPU temp at 151 C. After reverting to the earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

Version 196.75 was designed to increase sales of replacement
nVidia GPUs. Heh, heh. Actually, a bug in 196.75 turned the
fans off resulting in GPU failure. Who says software can't
destroy hardware?


How difficult would it have been, to automate the fan control
in a failsafe way ? That to me is the crime, not the bad drivers.
Who designs hardware, in a way that allows it to burn up ?

1) There probably aren't any video cards, where the fan should
stop rotating. The PWM control register simply should not
support a zero setting! Have the PWM range span 7 volts
to 12 volts, for a 12 volt fan. That way, no matter how
brain dead the software, your fan gets at least 7 volts.
(Assuming 7 volts is enough to keep the particular video
card fan spinning.) There should be no reports of a video
card, where the software stopped the fan from spinning. And
that can be guaranteed in hardware.

2) (Insert crafty scheme, where PWM output goes to 100%, if
a critical temperature is passed. This gives an audible
warning to the user, that something is amiss. Effectively,
this is a hardware bypass for bad software, such as missing
closed loop feedback fan control. If the hardware detects a
151C GPU, it should force the fan to 100% setting.)

3) Processors have had THERMTRIP for several generations. How
hard would it be, for a 151C GPU to assert a THERMTRIP signal,
which is connected to the disable pins on the voltage regulators
on the video card ? That is the "lights out" approach to
protecting your video card.

Paul

"Norman Peelman" wrote in message
g.com...
Tom Lake wrote:

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only* those 3
Blizzard games. Bit too wierd for coincidence, no ? Thankfully I play
none of them, and the drivers are blazing fast. Keepers for me, fix or
not.

My .75 revision kept the GPU temp at 151 C. After reverting to the
earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

I'm sure you mean Fahrenheit...

No, I have the temp measured in Celsius. Those are the actual numbers.
nVidia chips run quite hot.

Tom Lake

In article , "Ian D" wrote:

"Tom Lake" wrote in message
...

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only* those 3
Blizzard games. Bit too wierd for coincidence, no ? Thankfully I play
none of them, and the drivers are blazing fast. Keepers for me, fix or
not.

My .75 revision kept the GPU temp at 151 C. After reverting to the earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

Version 196.75 was designed to increase sales of replacement
nVidia GPUs. Heh, heh. Actually, a bug in 196.75 turned the
fans off resulting in GPU failure.



Who says software can't destroy hardware?





Toyota?????

On 03/10/10 21:32, Norman Peelman wrote:
Tom Lake wrote:

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only*
those 3 Blizzard games. Bit too wierd for coincidence, no ?
Thankfully I play none of them, and the drivers are blazing fast.
Keepers for me, fix or not.

My .75 revision kept the GPU temp at 151 C. After reverting to the
earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

I'm sure you mean Fahrenheit...

40 degrees F? Doubt it. In my fridge, sure, but my desktop computer, not
without some serious cooling from a freezer.

On 2010-03-12, Tony Harding wrote:
On 03/10/10 21:32, Norman Peelman wrote:
Tom Lake wrote:

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only*
those 3 Blizzard games. Bit too wierd for coincidence, no ?
Thankfully I play none of them, and the drivers are blazing fast.
Keepers for me, fix or not.

My .75 revision kept the GPU temp at 151 C. After reverting to the
earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

I'm sure you mean Fahrenheit...

40 degrees F? Doubt it. In my fridge, sure, but my desktop computer, not
without some serious cooling from a freezer.

"my GPU temp is now 40 C" != "my GPU temp is now 40, see?"
lol

Tony Harding wrote:
On 03/10/10 21:32, Norman Peelman wrote:
Tom Lake wrote:

"Dirk Dreidoppel" wrote in message
...
Am 08.03.2010 01:10, schrieb Carl:
Why oh why is there so many dodgy drivers from Nvidia? It's a PITA!


The most interesting question right now would be why it's *only*
those 3 Blizzard games. Bit too wierd for coincidence, no ?
Thankfully I play none of them, and the drivers are blazing fast.
Keepers for me, fix or not.

My .75 revision kept the GPU temp at 151 C. After reverting to the
earlier
version, my GPU temp is now 40 C. I don't play any games at all.

Tom L

I'm sure you mean Fahrenheit...

40 degrees F? Doubt it. In my fridge, sure, but my desktop computer, not
without some serious cooling from a freezer.

Um... 151 C. = 304 F. Unless 151 C. was really a typo???


--
Norman
Registered Linux user #461062

Hi!

Um... 151 C. = 304 F. Unless 151 C. was really a typo???

I'd suggest (based on prior experience) that survival of a semiconductor
device at such temperatures can happen for a while.

A customer of mine had a 700MHz AMD Athlon processor on a cheap no-name
motherboard. At some point both the CPU and PSU fans quit, but the system
kept right on running. When I sat down to work on it, I could not help but
comment on the very strong "hot" smell.

Their response was that it had been smelling hot for some time. And sure
enough, the case was hot to the touch...okay, let's pop the cover and see.
No fan was running, and somehow the system didn't melt despite several
months of this. I replaced both fans and the system continued working. Maybe
it's still working? I don't know. Haven't seen that computer in years.

William

"William R. Walsh" m
wrote in message ...
Um... 151 C. = 304 F. Unless 151 C. was really a typo???

I'd suggest (based on prior experience) that survival of a semiconductor
device at such temperatures can happen for a while.

Sure, semiconductor devices don't run on water, so the 100C barrier is
purely psychological. Often the death occurs due to different rates of
thermal expansion between certain close-fitting components, for example
between the silicon chip die and polymer chip package.

A customer of mine had a 700MHz AMD Athlon processor on a cheap no-name
motherboard. At some point both the CPU and PSU fans quit, but the system
kept right on running. When I sat down to work on it, I could not help but
comment on the very strong "hot" smell.

Interesting, so the CPU heatsink relied on natural convection and reached an
equilibrium point?

A couple of years ago, when I bought my first Radeon X1900XT, the default
fan schedule was geared toward low-noise. I hit 110C in some games. There
was no "hot" smell, but I could hear the card creak and crackle afterwards
as it cooled down. Pretty cavalier way of running a $500 video card!

Don't most nVidia cards have a hard-wired shutdown threshold set at 124 C?

--
"War is the continuation of politics by other means.
It can therefore be said that politics is war without
bloodshed while war is politics with bloodshed."

On 2010-03-15, First of One wrote:

Don't most nVidia cards have a hard-wired shutdown threshold set at 124 C?

In my Nvidia control panel, the "Slowdown Threshold" is listed as 105*C
for my 9600GT.