Apex Legends Crash no error - PC (apex_crash.txt)

@scrawny633 it looks like an overclocking issue, since the instruction address that failed to execute doesn't match the instruction address that it wanted to execute.

00007FF60000201B -- this is the address that it complained about executing in the EXCEPTION_ACCESS_VIOLATION line

00007FF68D4F22FF -- this is the address that it should be executing from the RIP register

If you know hex, you can see that some of the bits that are 1 in RIP are 0 in the address it complained about.

We had common crashes like this due to overclocking (on Intel only) before we changed the instructions used to do a particular calculation. I recommend dropping your OC by 0.1 GHz and see if Apex is stable.

Stress tests can prove that your overclocking is unstable, but they can't prove that your overclocking is stable. In other words, they show you don't have problems in their tests over the time period the test runs, but they don't prove that you won't have problems under ANY workload over ANY timeframe.

This is a fundamental limitation of science, by the way. Experiments can prove a theory is WRONG if the results don't match predictions, but the fact that they do match doesn't prove that a theory is RIGHT. For example, Newtonian physics were experimentally verified for centuries, until experiments proved them wrong in the late 1800s / early 1900s (or so), and then we had to develop relativistic physics and quantum mechanics.

So, it's good to run stress tests to test your overclocking, but the fact that the tests are stable doesn't mean that you won't have problems due to overclocking. The program that fails doesn't even have to be particularly stressful; it just has to hit on a sequence of instructions and memory reads such that one stage of the CPU pipeline somewhere doesn't finish before the next stage starts.