Specifying plan classes

You can specify plan classes using an XML configuration file with the format

<plan_classes>
   <plan_class>
      ... specification of a plan class
   </plan_class>
   ... more plan class specifications
</plan_classes>

The specification of a plan class has the following format. All elements except <name> are optional.

<name>X</name>

the name of the plan class

<virtualbox/>

VirtualBox application

<is64bit/>

64-bit application

<min_cpus>x</min_cpus>

requires at least X CPUs (may be fractional)

<max_threads>x</max_threads>

use at most this many CPUs

<avg_ncpus>x</avg_ncpus>

average # CPUs used. Use for non-compute-intensive apps; for others it's calculated for you.

<projected_flops_scale>x</projected_flops_scale>

in estimating this app's FLOPS on a given host, multiply peak device speeds by this (0 < x <= 1). For GPU apps, use 0.2 or so

<os_regex>regex</os_regex>

use only hosts whose operating system name matches the given regular expression

<cpu_feature>x</cpu_feature>

a required CPU feature (such as sse3). You can include more than one.

The following lets you use a project preference to decide whether to use the app version:

<project_prefs_tag>x</project_prefs_tag>

the name of the tag

<project_prefs_regex>x</project_prefs_regex>

the contents must match this regular expression

Fields for GPU apps:

<gpu_type>X</gpu_type>

the GPU type (nvidia, amd, or intel)

<cpu_frac>x</cpu_frac>

the fraction of total FLOPs that are done by the CPU. This is used to calculate CPU usage and estimated FLOPS.

<min_gpu_ram_mb>x</min_gpu_ram_mb>

The minimum amount of GPU RAM. This is needed because older clients report total RAM but not available RAM.

<gpu_ram_used_mb>x</gpu_ram_used_mb>

require this much available GPU RAM

<gpu_peak_flops_scale>x</gpu_peak_flops_scale>

scale GPU peak speed by this

<ngpus>x</ngpus>

how many GPUs will be used

<min_driver_version>x</min_driver_version>

minimum display driver version

<max_driver_version>x</max_driver_version>

maximum display driver version

<cuda/>

CUDA application (NVIDIA)

<cal/>

CAL application (AMD)

<opencl/>

OpenCL application

<gpu_utilization_tag>x</gpu_utilization_tag>

you can use a project-specific preferences to let users scale the # of GPUs used. This is the tag name.

Fields for NVIDIA GPU apps:

<min_nvidia_compcap>x</min_nvidia_compcap>

minimum compute capability

<max_nvidia_compcap>x</max_nvidia_compcap>

maximum compute capability

Fields for CUDA apps

<min_cuda_version>x</min_cuda_version>

minimum CUDA version

<max_cuda_version>x</max_cuda_version>

maximum CUDA version

Fields for OpenCL apps

<min_opencl_version>x</min_opencl_version>

minimum OpenCL version

<max_opencl_version>x</max_opencl_version>

maximum OpenCL version

Fields for VirtualBox apps

<min_vbox_version>x</min_vbox_version>

minimum VirtualBox version

<max_vbox_version>x</max_vbox_version>

maximum VirtualBox version