Release v1.2 (#274)

* release v1.2

* remove debug script

* remove unnecessary changes

* remove cuda error check

* added init_mdj_constnt back

README.md

@@ -53,7 +53,8 @@ Features
 - MP2/DF-MP2 and CCSD (experimental);
 - Polarizability, IR, and NMR shielding (experimental);
 - QM/MM with PBC;
-- CHELPG, ESP, and RESP atomic charge
+- CHELPG, ESP, and RESP atomic charge;
+- Multi-GPU for density fitting (experimental)
 
 Limitations
 --------
@@ -134,22 +135,22 @@ References
 ---------
 ```
 @misc{li2024introducting,
-      title={Introducing GPU-acceleration into the Python-based Simulations of Chemistry Framework}, 
+      title={Introducing GPU-acceleration into the Python-based Simulations of Chemistry Framework},
       author={Rui Li and Qiming Sun and Xing Zhang and Garnet Kin-Lic Chan},
       year={2024},
       eprint={2407.09700},
       archivePrefix={arXiv},
       primaryClass={physics.comp-ph},
-      url={https://arxiv.org/abs/2407.09700}, 
+      url={https://arxiv.org/abs/2407.09700},
 }
 
 @misc{wu2024enhancing,
-      title={Enhancing GPU-acceleration in the Python-based Simulations of Chemistry Framework}, 
+      title={Enhancing GPU-acceleration in the Python-based Simulations of Chemistry Framework},
       author={Xiaojie Wu and Qiming Sun and Zhichen Pu and Tianze Zheng and Wenzhi Ma and Wen Yan and Xia Yu and Zhengxiao Wu and Mian Huo and Xiang Li and Weiluo Ren and Sheng Gong and Yumin Zhang and Weihao Gao},
       year={2024},
       eprint={2404.09452},
       archivePrefix={arXiv},
       primaryClass={physics.comp-ph},
-      url={https://arxiv.org/abs/2404.09452}, 
+      url={https://arxiv.org/abs/2404.09452},
 }
 ```

gpu4pyscf/__init__.py

@@ -1,4 +1,4 @@
-__version__ = '1.1.0'
+__version__ = '1.2.0'
 
 # monkey patch libxc reference due to a bug in nvcc
 from pyscf.dft import libxc

gpu4pyscf/lib/gvhf-md/unrolled_md_j.cu

@@ -1,3 +1,5 @@
+#include <stdio.h>
+#include <cuda_runtime.h>
 #include "gvhf-rys/vhf.cuh"
 #include "gvhf-rys/gamma_inc_unrolled.cu"
 
@@ -5016,6 +5018,7 @@ int md_j_unrolled(RysIntEnvVars *envs, JKMatrix *jk, BoundsInfo *bounds,
     dim3 threads(16, 16);
     dim3 blocks;
     int ijkl = lij*9 + lkl;
+
     switch (ijkl) {
     case 0: // lij=0, lkl=0
         blocks.x = (bounds->npairs_ij + 255) / 256;

gpu4pyscf/scf/j_engine.py

@@ -97,14 +97,14 @@ def get_j(mol, dm, hermi=1, vhfopt=None, omega=None, verbose=None):
     )
 
     libvhf_md.init_mdj_constant(ctypes.c_int(SHM_SIZE))
-
     uniq_l_ctr = vhfopt.uniq_l_ctr
     uniq_l = uniq_l_ctr[:,0]
     l_ctr_bas_loc = vhfopt.l_ctr_offsets
     l_symb = [lib.param.ANGULAR[i] for i in uniq_l]
     n_groups = len(uniq_l_ctr)
     tile_mappings = {}
     workers = gpu_specs['multiProcessorCount']
+
     info = cp.empty(2, dtype=np.uint32)
 
     for i in range(n_groups):
@@ -158,7 +158,7 @@ def get_j(mol, dm, hermi=1, vhfopt=None, omega=None, verbose=None):
                         mol._atm.ctypes, ctypes.c_int(mol.natm),
                         mol._bas.ctypes, ctypes.c_int(mol.nbas), _env.ctypes)
                     if err != 0:
-                        raise RuntimeError(f'RYS_build_jk kernel for {llll} failed')
+                        raise RuntimeError(f'MD_build_j kernel for {llll} failed')
                     if log.verbose >= logger.DEBUG1:
                         ntasks = tile_ij_mapping.size * tile_kl_mapping.size
                         t1, t1p = log.timer_debug1(f'processing {llll}, tasks ~= {ntasks}', *t1), t1

gpu4pyscf/scf/tests/test_scf_j_engine.py

@@ -43,3 +43,4 @@ def test_j_engine():
     ref = get_jk(mol, dm, with_k=False)[0]
     assert abs(lib.fp(vj1) - -2327.4715195591784) < 1e-9
     assert abs(vj1 - ref).max() < 1e-9
+