Best practice configs / AIDA64 configuration

[onliner2021]

Hello,

first of all: CoreCycler is a great tool. Thanks a lot for putting int all that time and effort!
I wouldn't know how to tune my CPU without it.

I have a question: Can you give me pointers on setting up a good config file for AIDA64 integration?
I'm trying to tune a 64 core CPU on a per core negative curve optimizer value basis with CoreCycler and would like to include AIDA testing. This is what I have so far, but I have no idea if this is best practice. Especially what mode is recommended.

So far I've had all 64 cores stable with the following yCruncher-Old AllTests config which took around 11 hours to test:

[General]
maxIterations = 1
stressTestProgram = YCRUNCHER_OLD
runtimePerCore = auto
coreTestOrder = Sequential
numberOfThreads = 2
restartTestProgramForEachCore = 1

[yCruncher]
mode = 19-ZN2 ~ Kagari
tests = BBP, BKT, SFT, FFT, N32, N64, HNT, VST, C17

I was already celebrating and then I ran another pass with just BBP also in yCruncherOld:

[General]
maxIterations = 20
stressTestProgram = YCRUNCHER_OLD
runtimePerCore = 10s
coreTestOrder = Sequential
numberOfThreads = 2
restartTestProgramForEachCore = 1

[yCruncher]
mode = 19-ZN2 ~ Kagari
tests = BBP

And boy, was I in for a surprise. Everything that looked stable up to this point turned out to be unstable (32 cores with checksum errors and 3 cores that even caused a reboot when only BBP was run on them). I found that especially strange, since BBP was also included in the first 11h run, but seems to have caused no errors there.

So I'm running that now once again with just BBP and reducing each core once again, until I get a stable result... (this is still ongoing right now).

Had I known this before, than the BBP only test (limited to only 10seconds, after trying out how long it would take to trigger the checksum error) would have been the very first I had run (and not tuning each Core for months now to get it stable with yCruncher old). All of the other default tests (prime95, low load scenario, avx2,... which also run very long didn't show up errors either).

I don't want to make the same mistake twice: I'd therefore like to also try AIDA64. Maybe it has yet another test that shows early errors/instability.

Is there some best practice (config / mode) that you recommend? Maybe one for quick errors (like my BBP example) and one for a long final run.

So far I've created the following config, but it's just out of the blue, without any experience running AIDA64 tests:

[General]
maxIterations = 10
stressTestProgram = AIDA64
runtimePerCore = 300s
suspendPeriodically = 1
coreTestOrder = Default
skipCoreOnError = 1
numberOfThreads = 2

[Aida64]

# The test modes for Aida64
# Note: "RAM" consumes basically all of the available memory and makes the computer pretty slow
#       You can change the amount of RAM being used / tested with the "maxMempory" setting below
# CACHE: Starts Aida64 with the "Cache" stress test
# CPU:   Starts Aida64 with the "CPU" stress test
# FPU:   Starts Aida64 with the "FPU" stress test
# RAM:   Starts Aida64 with the "Memory" stress test
# You can also combine multiple stress tests like so: CACHE,CPU,FPU
#
# Default: CACHE
mode = FPU,CACHE


# Use AVX for Aida64
# This enables or disables the usage of AVX instructions during Aida64's stress tests
#
# Default: 0
useAVX = 0


# The maximum memory allocation for Aida64
# Sets the maximum memory usage during the "RAM" stress test in percent
# Note: Setting this too high can cause your Windows to slow down to a crawl!
#
# Default: 90
maxMemory = 90

Thanks in advance!