LOTUS cracking on 4x R9 290 with New-Multiforcer

Discussion and support for the CUDA Multiforcers (Windows and Linux)
  • Ads

LOTUS cracking on 4x R9 290 with New-Multiforcer

Postby rvn » Mon Mar 10, 2014 2:24 pm

Hi,
i have a test setup of the svn version of New-Multiforcer and wanted to share a few details if anbody wants to recreate it.

As a base system i use a debian7 minimal install with a dummy xdm to run amdconfig
I use the following drivers / SDKs:

ADL_SDK 6.0
AMD-APP-SDK-v2.9-lnx64
catalyst 13.12.
od6config for clocking and fan selection

To build the source from svn I had to install the cuda sdk (from debian sources) and had to add -DCL_USE_DEPRECATED_OPENCL_1_1_APIS as a new compiler flag (i think one #include was missing that threw an error during the compilation but can easily be resolved by googling).

The initial performance was poor (12Mio/card) but increased to ~60 Mio /card when using --vectorwidth=1.

--bfi_int crashes when i set it as an option.
rvn
 
Posts: 1
Joined: Mon Mar 10, 2014 2:16 pm

Return to CUDA Multiforcers

Who is online

Users browsing this forum: No registered users and 2 guests

cron