OneAPI Crystal

This library is the DPC++ porting (and extension) of the Crystal library, originally written in CUDA.

It Implements a collection of block-wide device functions that can be used to implement high performance implementations of SQL queries on CPUs and GPUs.

Usage on Intel® DevCloud

Above you can find a few details on how to run this code on Intel CPUs and GPUs using Intel DevCloud platform, assuming you cloned the repository already

Intel® Xeon® Gold 6128 CPUs

>> cd oneapi_crystal
>> qsub -I -d .

Now you have an interactive shell on a CPU node.

Intel® UHD Graphics P630 (iGPU):

>> cd oneapi_crystal
>> qsub -I -l nodes=2:gpu:ppn=2 -d .

Now you have an interactive shell on a (integrated) GPU node.

Intel® Iris® Xe MAX Graphics (dGPU):

>> cd oneapi_crystal
>> qsub -I -l nodes=1:iris_xe_max:ppn=2 -d .

Now you have an interactive shell on a (discrete) GPU node.

Run the queries

To run the star schema benchmark implementation:

# Generate the test generator / transformer binaries
cd ssb/dbgen
make
cd ../loader
make 
cd ../../

# Generate the test data and transform into columnar layout
# Substitute <SF> with appropriate scale factor (eg: 1)
python util.py ssb <SF> gen
python util.py ssb <SF> transform

configure the benchmark settings

vi queries/ssb_utils.h
# edit BASE_PATH in ssb_utils.h with the location
# of oneapi_crystal 

now build the queries

make queries

and run, say q11

./build/q11

Run the operators

Compile the operators running

make operators

then benchmark, say the hash join like this

./build/join <table_size>