Another implementation for IOTA PearlDiver, which allows to be multi-threaded executed

dcurl use SSE implementation for CPU and OpenCL for GPU, which are referenced from iotaledger/iota.lib.go and iotaledger/ccurl, respectively.

• You need to configure OpenCL platform and device by yourself in src/clcontext.c
• You also need to configure OpenCL Library path in Makefile
• dcurl is limited to use only One GPU currently, and multi-gpu support is in future work.

• Make a directory for build

$ mkdir build

• Running test program

$ make test

• Create a multi-thread environment via python

$ make libdcurl.so

$ python3 ./test/test_multi_pow.py <# of CPU task> <# of GPU task>

Runtime is measured with 100 transaction trytes.

(8, 7, 32), which means 8 pow tasks executed in CPU, 7 pow tasks executed in GPU AT THE SAME TIME, and 32 threads are created to search nonce for each pow task in CPU.

• src/pow_sse.c is modifed from utamaro/curl_trials
• src/pow_cl.c and src/pow_kernel.cl is modifed from iotaledger/ccurl