ORCA 5.0.2发布

wzkchem5

https://orcaforum.kofo.mpg.de/viewtopic.php?f=54&t=8429

Dear ORCA Users,

the ORCA development team is proud to announce the availability
of the second bugfix release in the ORCA 5.0 version series,
namely ORCA 5.0.2!

This is a pure bugfix release, addressing most of the bugs mentioned
in the Known Bugs section of ORCA 5.0.1.

This ORCA release is especially relevant as it fixes a race-condition in
a combination of shared matrices and parallel linear algebra,
leading to unreliable results and undefined run-time behaviour the more
cores have been used.
Due to this bug, all users are thoroughly encouraged to upgrade their
ORCA version to 5.0.2 immediately!

Among the bugs fixed are:
orca_plot with spin-densities from CASSCF and CASSCF/NEVPT2
CrystalPrep
WriteHess in ESD
Reverse ISC
SOC with TDA false
Windows version crashing in diagonalization
Dipole derivatives for frequencies
Wrong harmonic frequencies from density functionals without exact exchange
Quite a few bugs in %compound
Crash in orca_euler
MTR occasionally crashing or producing mtr.dcdcas.*.dat*
Center of inversion is not recognized and point group is too low
Wrong QROs
Wrong results with M06 and other hyb_*_x_* functionals from LibXC
!KeepDens crashes CASSCF+SOC
NEVPT2/CASPT2 memory leak
!UseSym crashes CASSCF/MRCI SOC calculations
CASSCF problem with RIJK conv
MRCI-SOC/SSC multipolar calculations
SOC-CD calculation in CASSCF, MRCI, and LFT
Crash during point group recognition
SOC-CASSCF absorbence calculation
RIJK conv in CASSCF/NEVPT2 memory/batching
ICE NRoots > 1 without starting CFGs
NEVPT2 D4Step=Efficient
ICE with CPCM

A full list of bugs fixed with additional information can be looked up at
Known Bugs section of ORCA 5.0.1

For the first time, this release also provides binaries for both MacOS architectures, Intel & Arm64.

In addition, a new site has been set up to allow all ORCA users to share their compound scripts:
GitHub: ORCAQuantumChemistry/Compound Scripts


Enjoy this new release of ORCA!


The ORCA Development Team

牧生

庆祝新版发布

ABQTrap

请问 “This ORCA release is especially relevant as it fixes a race-condition in
a combination of shared matrices and parallel linear algebra,
leading to unreliable results and undefined run-time behaviour the more
cores have been used.” 这段话具体是什么意思呢？
之前用ORCA 5.0.1 DFT计算（没有用M06，wB97M）的结果可靠吗？数据结果可以用吗？

biogon

linux的预编译版本能不能出个arm64的

wzkchem5

ABQTrap 发表于 2021-12-9 01:25
请问 “This ORCA release is especially relevant as it fixes a race-condition in
a combination of sh ...

race condition一般要么导致程序崩掉，要么产生肉眼可见的极其夸张的问题，比如能量差10000个Hartree之类的。如果不是这种情况，应该没问题

wzkchem5

biogon 发表于 2021-12-9 10:55
linux的预编译版本能不能出个arm64的

可以在orca论坛的feature request版问一下

ABQTrap

wzkchem5 发表于 2021-12-9 19:20
race condition一般要么导致程序崩掉，要么产生肉眼可见的极其夸张的问题，比如能量差10000个Hartree之类 ...

谢谢wzkchem5的解答！

浩文

您好，请问您一下，我注册orca论坛，但是说我需要输入有效的邮箱地址，我用了qq邮箱和学校的邮箱都不行，请问需要什么样的邮箱才行呀？

wzkchem5

浩文 发表于 2022-2-10 05:12
您好，请问您一下，我注册orca论坛，但是说我需要输入有效的邮箱地址，我用了qq邮箱和学校的邮箱都不行，请 ...

后缀是edu或者edu.cn的邮箱应该都行吧
另外检查一下@是不是打成全角符号了

浩文

wzkchem5 发表于 2022-2-10 15:39
后缀是edu或者edu.cn的邮箱应该都行吧
另外检查一下@是不是打成全角符号了

您好，我用的edu.cn的邮箱，并且检查了@的全角半角，还是显示The solution you provided was incorrect
Please note that you will need to enter a valid email address before your account is activated. You will receive an email at the address you provide that contains an account activation link.请问有什么解决办法吗

wzkchem5

浩文 发表于 2022-2-13 11:24
您好，我用的edu.cn的邮箱，并且检查了@的全角半角，还是显示The solution you provided was incorrect
...

把你的详细操作流程逐步截图，然后发邮件给Frank Wennmohs（frank.wennmohs@kofo.mpg.de）问怎么处理。ORCA论坛的技术问题是他负责

zhaojq

现在官网也上不去，哪里可以下载啊？

413

zhaojq 发表于 2022-8-22 22:45
现在官网也上不去，哪里可以下载啊？

官网可以上去呀
可能要用梯子