This particular story started after 2.1, when I noted that Cyberpunk behaved weird at launch already (at launch the game (nearly) freezes up). Few weeks back I send an Process Monitor log to CD Projekt Red in reply to their third failure to help in diagnosing what the issue was, because I realised that might show the issue. Never heard something back.
Skip forward to today. I ran into some issues with Stellaris and send Paradox some details including the DxDiag. Paradox did react pretty timely, but both failed to read something correctly but gave me an invaluable note which I hadn't realized I could see it there myself. Namely, they pointed out a certain BEX-error. BEX-errors are (according to the Internet) Buffer Overflow Exception errors. In other words, the program which causes the error tries to ram too much data in the RAM. This is basically what's also called an 'memory leak' - a program mismanages the calls to the RAM, holds onto RAM which isn't needed anymore and then tries to pile more memory requests on top of it.
For those with sufficient technical knowledge: those are the BEX-related errors in DxDiag:
1) Possibly - the CPU gets up to 100% load while the GPU is setting at <5%
2) Most certainly: if you get an dev-null error basically. What do I mean with that? See the attachment. Normally errors are located at a certain location in the memory, which reads in human-readable language as a set of numbers and letters. 'Dev-null' is when a set of data just gets thrown straight away, which is what basically happens here - there's no related address at all.
Skip forward to today. I ran into some issues with Stellaris and send Paradox some details including the DxDiag. Paradox did react pretty timely, but both failed to read something correctly but gave me an invaluable note which I hadn't realized I could see it there myself. Namely, they pointed out a certain BEX-error. BEX-errors are (according to the Internet) Buffer Overflow Exception errors. In other words, the program which causes the error tries to ram too much data in the RAM. This is basically what's also called an 'memory leak' - a program mismanages the calls to the RAM, holds onto RAM which isn't needed anymore and then tries to pile more memory requests on top of it.
For those with sufficient technical knowledge: those are the BEX-related errors in DxDiag:
And yes, I've sent this also already onwards towards support. Did that before I started typing this. And yes, I've got enough RAM (16 GB 3200), GPU (6600XT) and CPU (5800X3D). Before the 5800X3D I had an Ryzen 5 3600, so at first I thought purely the CPU was the error. Now with the 5800X3D and that BEX-error though.... How can you see if there's an memory leak for you too?
1) Possibly - the CPU gets up to 100% load while the GPU is setting at <5%
2) Most certainly: if you get an dev-null error basically. What do I mean with that? See the attachment. Normally errors are located at a certain location in the memory, which reads in human-readable language as a set of numbers and letters. 'Dev-null' is when a set of data just gets thrown straight away, which is what basically happens here - there's no related address at all.
Attachments
-
REDLauncherError.PNG