Newest Nvidia driver 378.49 causes memory leak

[bcrawley57]

Reverting to 376.33 resolved my VAS issues. Back to enjoying flying again.

[Martyson]

Reverting to 376.33 resolved my VAS issues. Back to enjoying flying again.

What do you think is causing your VAS issue using a newer Nvidia driver?

Have you contacted Nvidia?

[tonaz]

i just did my test.
same exact setup for both tests.
JF757 at KJFK, some addons.

378.78: VAS under 300MB, almost OOT
376.33: VAS between 1GB and 950MB

Meh... -.-''

[Martyson]

i just did my test.
same exact setup for both tests.
JF757 at KJFK, some addons.

378.78: VAS under 300MB, almost OOT
376.33: VAS between 1GB and 950MB

Meh... -.-''

What was your test .. just start P3D with aircraft JF757 parked at a JFK gate?

[tonaz]

nope, tookoff and went around NYC with a circle FMC route.
Right after take off i was instantly 400mb (anyways from a low 690), then 300 after first lap. Did some laps then i stopped the test after 45 minutes. I was barely OOT with 285MB.
Then, reinstalled 376.33 and did same laps, 3 or 4 for about 30 minutes, and VAS was always more than 980MB.
Is this a good way to test?
Seems enough to me to say 378.78 got leak.
Why you ask?

[Martyson]

"Right after take off i was instantly 400mb."

Just to verify .. is this 400MB remaining (before an OOM) form the initial 4GB?
If so,this is not a leak.

IMO, that is a very bad problem on your setup which is showing up when you are using the newer Nvidia drivers.

Have you posted on the Nvidia forum?

[tonaz]

no please. nobody from NV answer there.
i don't need an utopic answer from NV, i was just reporting.

no leak? so how to explain 400mb right after start with 378.78 and stable 980mb right after start with 376.33?

bad setup?! doesn't make sense to me.

[Martyson]

no please. nobody from NV answer there.
i don't need an utopic answer from NV, i was just reporting.

no leak? so how to explain 400mb right after start with 378.78 and stable 980mb right after start with 376.33?

bad setup?! doesn't make sense to me.

If everyone using P3D and a newer Nvidia driver were getting 400MB VAS remaining (very near OOM) right after takeoff there would be many pilots joining this thread.

That's is why I asked if you contacted Nvidia.

Did you see my VAS remaining on my last test flight (previous page of these posts)?
That is a small leak during a longer test flight.

[marcon8122]

I've asked nvidia but we have do to all too many people with this iussue not reporting. If you not report they don't check

[Martyson]

I've asked nvidia but we have do to all too many people with this iussue not reporting. If you not report they don't check

I do not understand your statement.


Are you seeing the same problem Tonaz reported (post above)?

[marcon8122]

Yes i've the same and i know many people with this vas leak iussue i've already reported to nvidia without reply. I think they need more reports to start an investigation to fix the iussue.

[tonaz]

Cmon we all know they do not check anyways if it is not about a AAA+ latest videogame...
Vaughan a lot, A LOT, of people have this problem, more than you think, and actually a lot of people suggest to stay 367.33, even the great p3d gurus.

[N987PL]

I'm assuming that this error may also be related to this latest NVIDEA driver (378.78)????

http://www.prepar3d.com/forum/viewtopic ... 2&t=123941

[tonaz]

Could be... Do you use fspassengers?

[Martyson]

Cmon we all know they do not check anyways if it is not about a AAA+ latest videogame...
Vaughan a lot, A LOT, of people have this problem, more than you think, and actually a lot of people suggest to stay 367.33, even the great p3d gurus.

Your 400MB after takeoff is not a slow leak on a long haul flight.
You have a severe problem with your setup, IMO.


Have you uninstalled of all of your add-ons and P3D followed by a full reninstall of P3D and then your default KJFK scenery test in a default P3D aircraft using new Nvidia driver?

No addons installed while testing Nvidia and P3D and reporting results here in the forum.

debug steps here:
http://www.prepar3d.com/forum/viewtopic ... 2&t=123246