3caf0e8b09
print Ritz values before and after the shift; rollback to simple last lanczos vector rebuild; print formatting is changed
2018-02-02 18:43:37 -05:00
bc5ba39278
relocate the printing block of converged eigenvalues
2018-01-31 12:40:33 -05:00
fbe1209f7e
count converged eigenvalues not assuming candidates are sorted
2018-01-31 12:10:24 -05:00
dc6f637e70
change GparityDomainWallFermion to ZMobius and add command line options to read boundary phase and omega
2018-01-27 08:21:27 -05:00
44b218a595
force hermicity to the block alpha and force diagonal of the block beta to be real
2017-12-29 23:26:17 -05:00
3cb8cb7282
'typename' is added to compile with AVX512 using GCC7.2.0; a semicolon was missing in Grid_avx512.h and the bug is fixed. Option SKL is added to configure script for skylake processor specific AVX512 operations. Code can be compiled with --enable-simd=SKL using GCC 7.2.0, but Test_simd fails. AVX512 support for complex double type with non-intel compilers makes this error; it needs a review.
2017-12-23 14:54:07 -05:00
89c4e9b168
first complete version of IRBL; requires practical test and clean up
2017-12-21 23:13:39 -05:00
fe406e230d
block with a single vector case is working with IRBL
2017-12-18 11:26:42 -05:00
5139eaf491
block Lanczos construction is added.
2017-12-03 23:55:22 -05:00
3a754fcd51
Merge branch 'develop' of https://github.com/paboyle/Grid into feature/Lanczos
2017-10-27 17:34:35 -04:00
034de160bf
Staggered updates : Schur fixed and added a unit test for Test_staggered_cg_schur.cc giving stronger check
2017-10-26 20:58:46 +01:00
9fe6ac71ea
Starting reorg of Blocked lanczos
2017-10-11 10:12:07 +01:00
ef61b549e6
Merge branch 'feature/Lanczos' into ckelly_develop4
2017-10-10 13:41:43 -04:00
bf58557fb1
Block compressed Lanczos
2017-10-10 14:15:11 +01:00
a1d80282ec
cb factorise
2017-10-10 13:49:31 +01:00
f0e084a88c
Schur staggered
2017-10-10 10:00:43 +01:00
4f8b6f26b4
Merge branch 'develop' into feature/dwf-multirhs
2017-10-02 11:41:49 +01:00
771a1b8e79
Merge pull request #128 from paboyle/feature/CG-reliable-update
...
Feature/cg reliable update
2017-09-21 10:12:03 +01:00
b83b2b1415
Stability improvement to BCG. Force m_rr hermitian beyond rounding.
2017-09-04 14:09:47 +01:00
3006663b9c
Schur solver for staggered type (hermition Mpc) opertors
2017-08-31 21:32:01 -04:00
d9cd4f0273
Staggered multinode block cg debugged. Missing global sum.
...
Code stalls and resumes on KNL at cambridge. Curious.
CG iterations 23ms each, then 3200 ms pauses. Mean bandwidth reports
as 200MB/s. Comms dominant in the report. However, the time behaviour suggests it
is *bursty*.... Could be swap to disk?
2017-08-23 15:07:18 +01:00
0145685f96
Added Staggered Type Preconditioned operator
2017-08-18 01:44:31 -04:00
e73e4b4002
Minor changes fixes
2017-08-11 01:35:25 -04:00
caa6605b43
Still tweaking memory saving routines in Lanczos
2017-08-07 00:01:04 -04:00
9939b267d2
Added switching to fallback linear operator in reliable update CG, and added recalculation of b parameter on update.
2017-07-31 13:39:44 -04:00
191fbf85fc
Added ImplicitlyRestartedLanczosCJ to Algorithms.h
2017-07-28 15:33:59 -04:00
9f280b82c4
Added mixed-precision CG with reliable updates
2017-07-25 11:30:41 -04:00
93650f3a61
Adding back (temporarily) dense matrix routines until Lanczos is fininalized
2017-07-24 21:49:25 -04:00
cab4b4d063
Deleting old include file references
2017-07-24 20:51:31 -04:00
cf4b30b2dd
re-adding ImplcitlyRestartedLanczos
2017-07-24 20:40:25 -04:00
c51d0b4078
Merge branch 'develop' of https://github.com/paboyle/Grid into feature/Lanczos
2017-07-24 20:35:29 -04:00
e504260f3d
Able to run a test job splitting into multiple MPI subdomains.
2017-06-22 18:53:11 +01:00
e8b95bd35b
Clean up finished. Could shrink Lanczos to around 400 lines at a push
2017-06-21 02:50:09 +01:00
7e35286860
Simplified lanczos, added Eigen diagonalisation.
...
Curious if we can deprecate dependencly on BLAS.
Will see when we get 48^3 running on our BG/Q port
2017-06-21 02:26:03 +01:00
0486ff8e79
Improved the lancos
2017-06-20 18:46:01 +01:00
e9cc21900f
Block solver complete for staggered. Now stable on mass 0.003 and
...
gives 8x (!) speed up on Haswell laptop vs. standard CG for 8 RHS solves.
166 iterations vs. 537 iterations so algorithmic gain + 2x in flop rate gain.
Better than a slap in the face with a wet kipper.
2017-06-20 12:37:41 +01:00
cfe3cd76d1
Block solver improvements
2017-06-19 14:04:21 +01:00
2f4cbeb4d5
Minor changes
2017-06-12 18:25:18 -04:00
fb7c4fb815
Recovering lapack interface without array allocation
2017-06-07 00:00:59 -04:00
00bb71e5af
Checking in before reworking lapack interface
2017-06-06 16:26:41 -04:00
cfed2c1ea0
Broken Lanczos. Going back to an older verion temporarily.
2017-06-06 12:14:45 -04:00
b1b15f0b70
Further fixes from multidimensional array
2017-06-05 23:13:41 -04:00
927c7ae3ed
changed allocation for LAPACK temporaries, to avoid crashing with some compilers (reported by Christoph)
2017-05-25 21:43:53 -04:00
05d04ceff8
Adding SimpleLanczos
2017-05-25 12:30:47 -04:00
5c479ce663
Merge branch 'develop' of https://github.com/paboyle/Grid into feature/Lanczos
2017-05-24 18:58:53 -04:00
4bf9d65bf8
Checking in memory saving version of Lanczos
2017-05-24 18:57:32 -04:00
3a056c4dff
Re-adding Bisection for SimpleLanczos
2017-05-22 18:23:03 -04:00
b0ba651654
Turning off the final sort for now
2017-05-19 10:49:09 -04:00
25d4c175c3
Cleaning up Lanczos
2017-05-18 18:33:47 -04:00
2439999ec8
Warning elimination; drop to -O2 on G++ bad versions
2017-05-06 14:44:49 +01:00
