| [../../index.menu.top.htm] | |||
Problems & Puzzles: Puzzles Puzzle 92. A pile of prime-spheres Days ago my friend Enoch Haga and me starting puzzling each other to construct a pile of balls (a tetrahedron) with the following properties: a) every ball contains a distinct prime, b) each prime-ball must be the sum of the prime numbers contained in the three balls from the immediate inferior level and in contact with the mentioned prime-ball. Can you imagine a pile of balls? A new friend of these pages, Chuck Henry, kindly and quickly provided several beautiful photos generated by him that should help to visualize a pile of balls like the one we are talking about. Please click here 1 and here 2 (*). Question: Get the least solution for a pile of n levels, for n=5, 6 & 7. _______ Hints: Here are (what we think) the minimal solutions for n (levels)= 2, 3 & 4: *** 19 19=3+5+11 3, 5 71
71=23+19+19 431 For n= 5 we have not found a "complete solution". This is the "best" I have found, 33/35= 94.28%, composites in bold & black. 22879 *** For n=5 I have gotten a 35/35 solution but only after accepting (9)
repeated (bold & black) primes: *** Solution At last, one complete solution for n=5, 35/35 primes without any repetition! Enoch Haga got it the Sunday 08/05/2000: 20431 72493 165799 357787 880211 *** Enoch Haga found a smaller solution (13/05/2000) for n=5 35/35: 15991 53503 107449 245587 837451 *** One more & smaller solution by Enoch Haga (11/3/2001) for n=5 is this: 15373 23173 8443 30253 8053 40609 35899 45979 44959 84067 104347 119797 237361 317971 804313 *** On October 29, 2026, Bas van Dorst launched a massive attack on 61 puzzles (64 results), including this one. Please see his contribution to this puzzle on the page https://www.primepuzzles.net/puzzles/puzz_1288.htm*** On Nov 20, 2026, Taichi Tsuneda wrote:I would like to submit a minimal solution for Puzzle 92 (A pile of prime-spheres) for n = 5, found in collaboration with an AI model (DeepSeek V4.1 Flash). Result (n = 5): - Total sum S = 5,811 (previous record: 16,561) - Top sphere T = 1,621 (previous record: 4,999) Bottom level (row-major, 15 primes): 67 | 29 31 | 17 7 5 | 41 3 13 19 | 79 59 11 47 97 All 5 levels (from top to bottom): 1621 463 | 617 541 223 | 137 103 | 313 167 271 127 | 53 43 | 61 23 37 | 179 73 71 163 67 | 29 31 | 17 7 5 | 41 3 13 19 | 79 59 11 47 97 Optimality check (computer-assisted exhaustive search): This is a computational proof by exhaustive search (not a manual analytical proof). Using two independent deterministic implementations (C and Python, without heuristic cuts), we verified that: - 0 valid piles exist with S <= 5,810 - 0 valid piles exist with T <= 1,620 - Exactly 6 valid piles exist with S <= 5,811 (and T <= 1,621), which are the 6 symmetries (rotations and reflections) of the single pile above. I have attached the archive `prime_tetra_n5_5811.zip` containing the independent verification script (`verify.py`), both exhaustive search codes (`pt.c` and `enum_indep.py`), and reproduction instructions (`README.md`). ***
|
|||
|
|||
|
|
|||
| [../../index.menu.bottom.htm] |