-
Notifications
You must be signed in to change notification settings - Fork 23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
GPU_OFFSET question #64
Comments
Yes, exactly, the only other thing to be underlined it’s that if algo-specific values for an algo are missing, global values are used in place as a fallback. This way you only need to set algo-specific settings for algos you’d like to use different values from global ones. This was already like that before my changes. But it was not explicated.
Inviato dal mio telefono Windows 10
Da: Payam Nab
Inviato: lunedì 21 maggio 2018 11:29
A: papampi/nvOC_by_fullzero_Community_Release
Cc: Subscribed
Oggetto: [papampi/nvOC_by_fullzero_Community_Release] GPU_OFFSET question(#64)
Regarding the two ALGO_SPECIFIC_with_GPU_OFFSET and GLOBAL_with_GPU_OFFSET modes, does this how they work?
ALGO_SPECIFIC_with_GPU_OFFSET: Set Algo Specific values then add plus/minus values for individual GPUs
And same for Global with offset ...
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.
|
So isn't it better to set all GPU_SPECIFIC values to blank to prevent misuse problems? |
We know, as script writers, that leaving those variables unset will be equivalent to a variable set to 0, but users may get confused easily. I'd rather keep all initialized to 0. |
I would still like to champion GPU discovery and a template per GPU
make/model.
Something simple like
if gpu = 1060, core=-200 me=900, power=76
if gpu = 1070, core=100 mem=700, power=110
etc etc
Can use the settings from buriedone just as a "runs well out of the box"
then users can tweak their own as time goes on, but immediately will be
working to a good standard (The above gets 22mh/250sols on a 3gb 1060 and
29.8mh/410sols on a 1070)
…On Tue, May 22, 2018 at 9:13 AM, Luca Piccirillo ***@***.***> wrote:
We know, as script writers, that leaving those variables unset will be
equivalent to a variable set to 0, but users may get confused easily. I'd
rather keep all initialized to 0.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#64 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AiaWrS8U09rQe-W9xriQrnJWqO034MFwks5t08i4gaJpZM4UGskf>
.
|
If you think that a rig will likely be composed of a fixed set of (hopefully not mixed) gpus you'll quickly find how useless would be a chunk of code which always detect the gpu type and decides "optimal" oc values at runtime. Then you should also consider tthose values are "ggod for all" only as a starting point, then everybody should check each car for stability and eventually tweak configs. You would obtain the same effect by simply suggesting users to visit buriedone checking for what kind of OC their cards support. |
Agree with Luke |
@LuKePicci , should I set salfter values to 0 as well, or just the uppercase oc values? |
All of them, since every policy equally applies to any kind of switchers. |
Done in PR #60 |
Regarding the two ALGO_SPECIFIC_with_GPU_OFFSET and GLOBAL_with_GPU_OFFSET modes, does this how they work?
ALGO_SPECIFIC_with_GPU_OFFSET: Set Algo Specific values then add plus/minus values for individual GPUs
And same for Global with offset ...
The text was updated successfully, but these errors were encountered: