Polyphemus: "Is the gigantic one-eyed son of Poseidon and Thoosa in Greek mythology" (quote Wikipedia). Henceforth, Polyphemus is the strongest computer chess engine\account on the Internet Chess Club without factual question.
Polyphemus uses: [[asmFishW_2016-07-17_bmi2.exe Haswell BMI2 64-bit compile]]. It has been evidently proven that this compile by Mohammed Li is 22+ percent faster than regular Stockfish compiles. BMI2 compiles are more than likely even faster - with greater performance.
Hardware used: [2x] 2686 v3s Haswell Intel architecture (18 cores per CPU) = 36 cores/72 threads total. Furthermore, I have 64 gigabytes of DDR4 KINGSTON RAM (8x8 memory modules) running at a whopping 2133 megahertz.
I am using all three, four, five, and six men Syzygy endgame tablebases running on a Samsung 850 PRO solid state drive. The total gigabyte usage for these tables is 150 gigabytes. I wholeheartedly believe that endgame tablebases add about 10+ ELO to the overall strength of today's best computer chess engines.
Polyphemus uses one of the best computer chess opening books on the face of this planet. Firstly, it is continually being worked on and finely tuned 24/7.
More importantly, it is a titan god-like book that uses both Stockfish development and the latest Komodo version(s) (now 9.42) to run painfully extensive\thorough analysis of many different variations and deviations.
My PC's dual-socket CPU performance obtains a whopping 3793+ Cinebench score. Even more curiously, my two 2686 v3s achieved a 32,619 PassMark CPU score. Most regular users, those being 99.9% percent of the computer-user population will be hard-pressed to achieve a 9500+ PassMark CPU score.
Here are some good links that I would like to share for some individuals\users who are interested in computer-chess engine development and fortunes: Here is a link to my PC's PassMark score dated as of the July 14th, 2016:http://www.passmark.com/baselines/V8/display.php?id=64882122964
Here is a link also to Mohammed Li's github where one can find AsmFish: https://github.com/tthsqe12/asm. Stockfish on my hardware is using 72 threads - quite the equivalent to using 72 cores. I obtain approximately 63 million nodes per second in opening-type positions.
I do have hyperthreading enabled. AsmFish is Numa aware; I see no reason to disable one of the most crucial of Intel CPU features. The only thing that I see, unequivocally, is a performance gain.