RTX4090 Loops and Thrust

[slot]

hello - when running

Code:

hashcat -b -m 610

on a RTX4090 i get 

Code:

Loops: 1024 Thrust: 256 Vec: 1

A friend gets,

Code:

Loops: 512 Thrust: 512 Vec: 1 

and seems to get better performance.

So,
1 Is a lower loops score better?
2 If 1 is yes, how can I get this score?
3 Is high Thrust score better
4 If 3 is yes how I can get that score?

I am running Cuda 12.3.68, but only the drivers. Would installing the SDK make a difference?

Thank you


OpenCL API (OpenCL 3.0 CUDA 12.3.68) - Platform #1 [NVIDIA Corporation]
=======================================================================
* Device #1: NVIDIA GeForce RTX 4090, 23808/24217 MB (6054 MB allocatable), 128MCU

Benchmark relevant options:
===========================
* --optimized-kernel-enable

------------------------------------------
* Hash-Mode 610 (BLAKE2b-512($pass.$salt))
------------------------------------------

Speed.#1.........: 12966.1 MH/s (82.70ms) @ Accel:32 Loops:1024 Thr:256 Vec:1