Endonuclease PvuII (1PVI) DNA - GATTACAGATTACA
CAP - Catabolite gene Activating Protein (1BER)
DNA - GATTACAGATTACAGATTACA Endonuclease PvuII bound to palindromic DNA recognition site CAGCTG (1PVI) DNA - GATTACAGATTACAGATTACA TBP - TATA box  Binding Protein (1C9B)
CAP - Catabolite gene Activating Protein (1BER)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
TBP - TATA box  Binding Protein (1C9B)
 

YASARA for MacOS

With release of the first ARM-based Macs in November 2020, Apple finalized a development that has been obvious during the years before: their plan to give up on compute-intensive scientific workloads, and focus on highly efficient mobile solutions instead.

Our goal to bring you the highest possible performance can only be reached by abundant use of the CPU's SIMD instruction set (SSE/AVX), which makes it impossible to simply recompile YASARA for ARM-based Macs. Instead, a considerable amount of porting work is involved to support the ARM equivalent called 'Neon'.

We evaluated the option to simply fade out YASARA for Mac, but decided against it. The scenario where a YASARA user does the hard work on a Linux/Windows machine or on a remote cluster, but wants to look at the results on a MacBook while travelling is just too common to ignore it. In addition, Apple is working on CPUs with more cores, that may one day provide state-of-the-art performance when delivered in a large-enough case to be combined with the best GPUs from AMD or nVIDIA.

As a consequence, we decided to stick with our approach from 2007, when we set the YASARA price tag for MacOS 20% higher to finance the unfortunate combination of a small user base and many technical difficulties below the shiny surface of MacOS. The port of YASARA to ARM/Neon was released in September 2023, and yielded a 50% performance jump. The CPU-only DHFR molecular dynamics benchmark running on an Apple MacBook Air from 2020 with 4 performance and 4 efficiency cores improved from 75 ns/day (using Rosetta-2 emulation) to 125 ns/day (using native ARM Neon code).

As before, you get an application that is specifically tailored to MacOS, including special support for the MacBook track pad gestures, the Magic Mouse and multi-monitor configurations. By default, YASARA is delivered as a universal binary that runs optimally on Intel- and ARM-based Macs, but you can save disk space by choosing your Processor on the download pages.

In addition, there are two ways to avoid the price premium:

  1. As soon as your summed up contribution to YASARA development exceeds 1000 EUR (academic) or 5000 EUR (commercial), no more MacOS price premium will be asked.
  2. If you help grow the YASARA MacOS user base by pointing your colleagues to this site, the price premium will be removed entirely as soon as there are enough users. The current ratio Windows:Linux:MacOS is 55:33:12 (in 2020, was 51:37:12 in 2016, 49:36:14 in 2014, 47:39:14 in 2011, 51:37:12 in 2009, and 57:41:2 in 2008), and MacOS should reach up to either Linux or Windows.

MacOS
12 YASARAs in parallel

The screenshot above shows 12 YASARA instances running on a MacBook Pro with Retina display. You can choose between four scaling factors for the user interface.  1× yields plenty of space as shown above, 2× yields the normal MacOS size, 3× and 4× are good for extreme resolution 8K screens and presentations with high-resolution beamers. Click the image for the full size screenshot.

MacOS Screen shot