academe-01

Hi @filipecosta90, my tests were made with with rdb disabled (save "") and I always pin redis to specific cores/numa nodes. Sorry I didn't mention it before. Also I have...

@stuartarchibald if I cast final result GPU won’t utilize full performance of operations. There is a huge penalty of INT64 vs INT32/INT16/… how can I overcome it ?

@AskAlexSharov Let's better discuss just pure MDBX operations. Using *intergration mdbx_to_mdbx* (which is broken btw https://github.com/ledgerwatch/erigon/issues/5708#issuecomment-1284562455 as it's simply completely overwrites dst db because of *os.RemoveAll(to)*) Considering it's just a...

@AskAlexSharov 1. thanks 2. already tried, didn't help much 3. it's not an option, i have that much free ram only in test env, in production i have only ~256...

@AskAlexSharov Alex, out of curiosity, Can you run mdbx_to_mdbx on your dataset and share at least some output strings? I would like to see a performance in real life, maybe...

> 4\. it must not use CPU, I don't know why you see it. you can add --pprof flag - it will print in logs command for cpu profiling: something...

@polakowo any chance for you to look into this ? I've tried and unfortunately simple change "cdef" not "cdef api" didn't work. some of ta-lib indicators are not implemented in...

@mrjbq7 I did, however I'm ending up with Segmentation fault _func.pxi: ``` cdef public api np.ndarray ADX_api(np.ndarray high, np.ndarray low, np.ndarray close, int timeperiod): return ADX(high, low, close, timeperiod) ```...

> My recommendation is to use `ConnectedComponents`, even though it is not parallel, the algorithm is quite fast. I need to process few billions of CC, unfortunately single threaded ```ConnectedComponents```...