SSDF

Comments

to the Swedish Rating List 2/2018, May 14

On our latest list we can present two new manually tested chess computers. The first one are Richard Lang's ChessGenius 3, which we have tested on an Android mobile phone, named ZTE Blade Apex 3. This mobile has an ARM Cortex-A53 Quad-core processor running at 1.3 GHz/core. As ChessGenius is a single-core program, it has used one of these cores. Despite using only one of four cores, it has achieved a strong rating of 2454 after the first 100 games. This is just 26 points shy of being the strongest ChessGenius version we have ever tested for the rating list, trailing the Genius 6.5 version which we tested on PC.

The second new entrant on the rating list is the Revelation Ruffian which is Per Ola Valfridsson's Ruffian 2.1 for Ruud Martin's Revelation 1 Chess computer. The Revelation 1, which in it's modular form for the old Mephisto chess computers also is named Resurrection 2, contain an Intel XScale PXA255 processor running at 500 MHz. On this hardware, and after its first 100 games, the Revelation Ruffian 2.1 has achieved a rating of 2357. With this addition, we have now tested most of the native engines contained in Ruud Martin's versatile and strong Resurrection 1-2/Revelation 1 series of chess computers. They span a range of approximately 400 rating points on the stronger Revelation 1 hardware, and perhaps a range of 500-600 points if one would include the untested engines on the Resurrection 1 chess computer.

The Rating list is also somewhat overhauled since last time. We have been going through the computers and simplified, standardised and changed some of the names so they are more uniform, less redundant and internationally correct. We have opted to take away the memory-size from the names. It was necessary for the Pentium MMX 200 MHz hardware level, and there it still indicates the average size of the RAM on the computers used for the testing of that specific program. From then on we have always used the same amount of memory on all test machines.

We have also included hardware-information in the cases where it was missing and decided to skip the MP designation, which we used to show that we used all cores on our Q6600 and 1800X hardware levels. This is because all programs we have tested on these two levels has been Multi processor compatible. The x64 designation is still there, or missing, to show if the program tested was 32 bit or 64 bit. A more complete review of all changes can be found on our homepage together with the information about our different hardware levels.

At the moment we are testing Stockfish 9 and will include it in our next ratinglist which we hope to have ready by late summer/early autumn.

Lars Sandin