Description of fast matrix multiplication algorithm: ⟨18×18×20:3648⟩

Algorithm type

3X8Y8Z8+3X8Y4Z6+3X4Y8Z6+3X4Y4Z10+6X8Y4Z4+3X4Y8Z4+3X4Y4Z8+3X6Y2Z6+48X4Y4Z6+6X8Y2Z3+36X6Y4Z2+42X6Y2Z4+36X4Y6Z2+306X4Y4Z4+36X2Y2Z8+6X4Y4Z3+6X4Y2Z5+75X6Y2Z2+6X6YZ3+6X4Y4Z2+6X4Y2Z4+108X2Y6Z2+72X2Y4Z4+93X2Y2Z6+36X6Y2Z+48X6YZ2+66X4Y2Z3+36X2Y6Z+42X2Y4Z3+6X2Y2Z5+78X6YZ+36X4Y3Z+312X4Y2Z2+36X3Y4Z+402X2Y4Z2+114X2Y2Z4+108XY6Z+36X3Y2Z2+6X3YZ3+24X2Y2Z3+36X2YZ4+54XY4Z2+36XY2Z4+72X3Y2Z+12X3YZ2+108X2Y3Z+135X2Y2Z2+132X2YZ3+90XY4Z+54XY2Z3+6X3YZ+126X2Y2Z+126X2YZ2+126XY2Z2+78XYZ3+42X2YZ+36XY2Z+36XYZ2+42XYZ3X8Y8Z83X8Y4Z63X4Y8Z63X4Y4Z106X8Y4Z43X4Y8Z43X4Y4Z83X6Y2Z648X4Y4Z66X8Y2Z336X6Y4Z242X6Y2Z436X4Y6Z2306X4Y4Z436X2Y2Z86X4Y4Z36X4Y2Z575X6Y2Z26X6YZ36X4Y4Z26X4Y2Z4108X2Y6Z272X2Y4Z493X2Y2Z636X6Y2Z48X6YZ266X4Y2Z336X2Y6Z42X2Y4Z36X2Y2Z578X6YZ36X4Y3Z312X4Y2Z236X3Y4Z402X2Y4Z2114X2Y2Z4108XY6Z36X3Y2Z26X3YZ324X2Y2Z336X2YZ454XY4Z236XY2Z472X3Y2Z12X3YZ2108X2Y3Z135X2Y2Z2132X2YZ390XY4Z54XY2Z36X3YZ126X2Y2Z126X2YZ2126XY2Z278XYZ342X2YZ36XY2Z36XYZ242XYZ3*X^8*Y^8*Z^8+3*X^8*Y^4*Z^6+3*X^4*Y^8*Z^6+3*X^4*Y^4*Z^10+6*X^8*Y^4*Z^4+3*X^4*Y^8*Z^4+3*X^4*Y^4*Z^8+3*X^6*Y^2*Z^6+48*X^4*Y^4*Z^6+6*X^8*Y^2*Z^3+36*X^6*Y^4*Z^2+42*X^6*Y^2*Z^4+36*X^4*Y^6*Z^2+306*X^4*Y^4*Z^4+36*X^2*Y^2*Z^8+6*X^4*Y^4*Z^3+6*X^4*Y^2*Z^5+75*X^6*Y^2*Z^2+6*X^6*Y*Z^3+6*X^4*Y^4*Z^2+6*X^4*Y^2*Z^4+108*X^2*Y^6*Z^2+72*X^2*Y^4*Z^4+93*X^2*Y^2*Z^6+36*X^6*Y^2*Z+48*X^6*Y*Z^2+66*X^4*Y^2*Z^3+36*X^2*Y^6*Z+42*X^2*Y^4*Z^3+6*X^2*Y^2*Z^5+78*X^6*Y*Z+36*X^4*Y^3*Z+312*X^4*Y^2*Z^2+36*X^3*Y^4*Z+402*X^2*Y^4*Z^2+114*X^2*Y^2*Z^4+108*X*Y^6*Z+36*X^3*Y^2*Z^2+6*X^3*Y*Z^3+24*X^2*Y^2*Z^3+36*X^2*Y*Z^4+54*X*Y^4*Z^2+36*X*Y^2*Z^4+72*X^3*Y^2*Z+12*X^3*Y*Z^2+108*X^2*Y^3*Z+135*X^2*Y^2*Z^2+132*X^2*Y*Z^3+90*X*Y^4*Z+54*X*Y^2*Z^3+6*X^3*Y*Z+126*X^2*Y^2*Z+126*X^2*Y*Z^2+126*X*Y^2*Z^2+78*X*Y*Z^3+42*X^2*Y*Z+36*X*Y^2*Z+36*X*Y*Z^2+42*X*Y*Z

Algorithm definition

The algorithm ⟨18×18×20:3648⟩ is serendipitous tensor product (⟨6×3×5:68⟩ - 16) ⊗ ⟨3×6×4:54⟩ +8⟨6×6×4:105⟩.

Algorithm description

These encodings are given in compressed text format using the maple computer algebra system. In each cases, the last line could be understood as a description of the encoding with respect to classical matrix multiplication algorithm. As these outputs are structured, one can construct easily a parser to its favorite format using the maple documentation without this software.


Back to main table