Ok thanks.
Must i test all freq for RAM ?
I set at 1030 yesterday by misstake, and i stay at 1730, instead 1830 with 1025Mh for RAM.....

Here is my problem

1. every time i close the miner WATTMAN always reset to default crash wattman
2. every restart must toggle on off HBCC to get real speed

- use ver 0.5 of cast AMD
- use blockchain driver aug 23
- use win 10

i wanna test the miner but

[06:07:34] Initializing GPU, loading kernel ...
Blockchain Compute Driver detected.
GPU0: gfx901 | 64 Compute Units
[06:07:35] Connecting to ssl://europe.monero.miningpoolhub.com:20580 ...
[06:07:35] SOCKET ERROR - Connection Error: GetAddrInfo: H¶te inconnu.
[06:07:35] Connecting to Pool failed. Retrying in 20 secs ...

i try with removing ssl://  but no luck  miningpoolhub not supported ?

1020/1025 for mem clock seems to be a sweet spot, probably some resonating effect of memory and cpu clock.

The range between 1025 - 1050 seems not to be very effective.

Memory clocks at 1050 and above are more faster then 1025, but depends on the the card if it can handle it.

nobody reply to me

who to correctly enter the pool name ?

cast_xmr-vega -S ssl://europe.monero.miningpoolhub.com:20580

or
cast_xmr-vega -S europe.monero.miningpoolhub.com:20580   not working

Does the pool require login? password? could be the issue

It's my understanding that 16GB is not enough to drive HBCC in 4 GPUs, according to other threads I've read.

I have 32 GB in mine, but have not gotten around to setting up all four GPUs. I will do that sometime this weekend and try to report back.

i have 4gb for 4 vega...

set your pagefile to 70go !

yes it s pool with login & password (password is not important i think)