Description of fast matrix multiplication algorithm: ⟨4×7×10:203⟩

Algorithm type

8X4Y7Z4+8X2Y7Z2+24X2Y4Z2+29X2Y2Z2+24XY4Z+2X2YZ2+6XY3Z+18XY2Z+84XYZ8X4Y7Z48X2Y7Z224X2Y4Z229X2Y2Z224XY4Z2X2YZ26XY3Z18XY2Z84XYZ8*X^4*Y^7*Z^4+8*X^2*Y^7*Z^2+24*X^2*Y^4*Z^2+29*X^2*Y^2*Z^2+24*X*Y^4*Z+2*X^2*Y*Z^2+6*X*Y^3*Z+18*X*Y^2*Z+84*X*Y*Z

Algorithm definition

The algorithm ⟨4×7×10:203⟩ could be constructed using the following decomposition:

⟨4×7×10:203⟩ = ⟨2×4×5:32⟩ + ⟨2×4×5:32⟩ + ⟨2×3×5:25⟩ + ⟨2×4×5:32⟩ + ⟨2×3×5:25⟩ + ⟨2×4×5:32⟩ + ⟨2×3×5:25⟩.

This decomposition is defined by the following equality:

TraceMulA_1_1A_1_2A_1_3A_1_4A_1_5A_1_6A_1_7A_2_1A_2_2A_2_3A_2_4A_2_5A_2_6A_2_7A_3_1A_3_2A_3_3A_3_4A_3_5A_3_6A_3_7A_4_1A_4_2A_4_3A_4_4A_4_5A_4_6A_4_7B_1_1B_1_2B_1_3B_1_4B_1_5B_1_6B_1_7B_1_8B_1_9B_1_10B_2_1B_2_2B_2_3B_2_4B_2_5B_2_6B_2_7B_2_8B_2_9B_2_10B_3_1B_3_2B_3_3B_3_4B_3_5B_3_6B_3_7B_3_8B_3_9B_3_10B_4_1B_4_2B_4_3B_4_4B_4_5B_4_6B_4_7B_4_8B_4_9B_4_10B_5_1B_5_2B_5_3B_5_4B_5_5B_5_6B_5_7B_5_8B_5_9B_5_10B_6_1B_6_2B_6_3B_6_4B_6_5B_6_6B_6_7B_6_8B_6_9B_6_10B_7_1B_7_2B_7_3B_7_4B_7_5B_7_6B_7_7B_7_8B_7_9B_7_10C_1_1C_1_2C_1_3C_1_4C_2_1C_2_2C_2_3C_2_4C_3_1C_3_2C_3_3C_3_4C_4_1C_4_2C_4_3C_4_4C_5_1C_5_2C_5_3C_5_4C_6_1C_6_2C_6_3C_6_4C_7_1C_7_2C_7_3C_7_4C_8_1C_8_2C_8_3C_8_4C_9_1C_9_2C_9_3C_9_4C_10_1C_10_2C_10_3C_10_4=TraceMulA_2_3A_1_4+A_2_5A_1_1+A_2_2A_1_6+A_2_7A_4_3A_4_5+A_3_4A_3_1+A_4_2A_3_6+A_4_7B_3_6B_3_3B_3_8B_3_5B_3_10B_4_1+B_5_6B_4_2+B_5_3B_4_7+B_5_8B_4_4+B_5_5B_4_9+B_5_10B_1_1+B_2_6B_2_3+B_1_2B_1_7+B_2_8B_2_5+B_1_4B_1_9+B_2_10B_6_1+B_7_6B_6_2+B_7_3B_6_7+B_7_8B_6_4+B_7_5B_6_9+B_7_10C_1_1+C_6_2C_1_3+C_6_4C_2_1+C_3_2C_2_3+C_3_4C_7_1+C_8_2C_7_3+C_8_4C_4_1+C_5_2C_5_4+C_4_3C_9_1+C_10_2C_9_3+C_10_4+TraceMulA_1_3-A_2_3A_1_5-A_2_5A_1_2-A_2_2A_1_7-A_2_7A_3_3-A_4_3-A_4_5+A_3_5-A_4_2+A_3_2A_3_7-A_4_7B_3_1+B_3_6B_3_2+B_3_3B_3_7+B_3_8B_3_4+B_3_5B_3_9+B_3_10B_5_1+B_5_6B_5_2+B_5_3B_5_7+B_5_8B_5_4+B_5_5B_5_9+B_5_10B_2_1+B_2_6B_2_2+B_2_3B_2_7+B_2_8B_2_4+B_2_5B_2_9+B_2_10B_7_1+B_7_6B_7_2+B_7_3B_7_7+B_7_8B_7_4+B_7_5B_7_9+B_7_10C_1_1C_1_3C_2_1C_2_3C_7_1C_7_3C_4_1C_4_3C_9_1C_9_3+TraceMul-A_1_4+A_2_4A_2_1-A_1_1-A_1_6+A_2_6-A_3_4+A_4_4-A_3_1+A_4_1-A_3_6+A_4_6B_4_1+B_4_6B_4_3+B_4_2B_4_7+B_4_8B_4_4+B_4_5B_4_9+B_4_10B_1_1+B_1_6B_1_2+B_1_3B_1_7+B_1_8B_1_4+B_1_5B_1_9+B_1_10B_6_1+B_6_6B_6_2+B_6_3B_6_7+B_6_8B_6_5+B_6_4B_6_9+B_6_10C_6_2C_6_4C_3_2C_3_4C_8_2C_8_4C_5_2C_5_4C_10_2C_10_4+TraceMulA_1_3A_1_4+A_1_5A_1_2+A_1_1A_1_6+A_1_7A_3_3A_3_4+A_3_5A_3_1+A_3_2A_3_6+A_3_7B_3_6B_3_3B_3_8B_3_5B_3_10B_5_6B_5_3B_5_8B_5_5B_5_10B_2_6B_2_3B_2_8B_2_5B_2_10B_7_6B_7_3B_7_8B_7_5B_7_10-C_1_1+C_6_1-C_1_3+C_6_3-C_2_1+C_3_1C_3_3-C_2_3-C_7_1+C_8_1C_8_3-C_7_3-C_4_1+C_5_1-C_4_3+C_5_3-C_9_1+C_10_1-C_9_3+C_10_3+TraceMulA_1_4A_1_1A_1_6A_3_4A_3_1A_3_6B_4_6-B_5_6B_4_3-B_5_3B_4_8-B_5_8B_4_5-B_5_5B_4_10-B_5_10-B_2_6+B_1_6-B_2_3+B_1_3B_1_8-B_2_8-B_2_5+B_1_5B_1_10-B_2_10B_6_6-B_7_6B_6_3-B_7_3B_6_8-B_7_8B_6_5-B_7_5B_6_10-B_7_10C_6_1+C_6_2C_6_4+C_6_3C_3_1+C_3_2C_3_3+C_3_4C_8_1+C_8_2C_8_3+C_8_4C_5_1+C_5_2C_5_4+C_5_3C_10_1+C_10_2C_10_3+C_10_4+TraceMulA_2_3A_2_5A_2_2A_2_7A_4_3A_4_5A_4_2A_4_7B_3_1B_3_2B_3_7B_3_4B_3_9-B_4_1+B_5_1-B_4_2+B_5_2-B_4_7+B_5_7-B_4_4+B_5_4B_5_9-B_4_9B_2_1-B_1_1B_2_2-B_1_2-B_1_7+B_2_7B_2_4-B_1_4-B_1_9+B_2_9-B_6_1+B_7_1-B_6_2+B_7_2-B_6_7+B_7_7-B_6_4+B_7_4-B_6_9+B_7_9C_1_1+C_1_2C_1_4+C_1_3C_2_1+C_2_2C_2_4+C_2_3C_7_1+C_7_2C_7_3+C_7_4C_4_1+C_4_2C_4_3+C_4_4C_9_1+C_9_2C_9_3+C_9_4+TraceMulA_2_4+A_2_5A_2_2+A_2_1A_2_6+A_2_7A_4_5+A_4_4A_4_1+A_4_2A_4_6+A_4_7B_4_1B_4_2B_4_7B_4_4B_4_9B_1_1B_1_2B_1_7B_1_4B_1_9B_6_1B_6_2B_6_7B_6_4B_6_9C_1_2-C_6_2C_1_4-C_6_4C_2_2-C_3_2C_2_4-C_3_4C_7_2-C_8_2C_7_4-C_8_4C_4_2-C_5_2-C_5_4+C_4_4C_9_2-C_10_2C_9_4-C_10_4

N.B.: for any matrices A, B and C such that the expression Tr(Mul(A,B,C)) is defined, one can construct several trilinear homogeneous polynomials P(A,B,C) such that P(A,B,C)=Tr(Mul(A,B,C)) (P(A,B,C) variables are A,B and C's coefficients). Each trilinear P expression encodes a matrix multiplication algorithm: the coefficient in C_i_j of P(A,B,C) is the (i,j)-th entry of the matrix product Mul(A,B)=Transpose(C).

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