Matrix4f Array optimized

public void mul4f(Matrix4fArray leftM, Matrix4fArray rightM) {
    final float[] left = leftM.buffer;
    final float[] right = rightM.buffer;
    final float[] buffer = this.buffer;

    float lm00 = left[0];
    float rm00 = right[0];

    float lm10 = left[1];
    float rm01 = right[4];

    float lm20 = left[2];
    float rm02 = right[8];

    float lm30 = left[3];

    buffer[0] = lm00 * rm00 + lm10 * rm01 + lm20 * rm02 + lm30 * right[12];

    float rm10 = right[1];
    float rm11 = right[5];
    float rm12 = right[9];

    buffer[1] = lm00 * rm10 + lm10 * rm11 + lm20 * rm12 + lm30 * right[13];

    float rm20 = right[2];
    float rm21 = right[6];
    float rm22 = right[10];

    buffer[2] = lm00 * rm20 + lm10 * rm21 + lm20 * rm22 + lm30 * right[14];

    float rm30 = right[3];
    float rm31 = right[7];
    float rm32 = right[11];

    buffer[3] = lm00 * rm30 + lm10 * rm31 + lm20 * rm32 + lm30 * right[15];

    float lm01 = left[4];
    float lm11 = left[5];
    float lm21 = left[6];
    float lm31 = left[7];

    buffer[4] = lm01 * rm00 + lm11 * rm01 + lm21 * rm02 + lm31 * right[12];
    buffer[5] = lm01 * rm10 + lm11 * rm11 + lm21 * rm12 + lm31 * right[13];
    buffer[6] = lm01 * rm20 + lm11 * rm21 + lm21 * rm22 + lm31 * right[14];
    buffer[7] = lm01 * rm30 + lm11 * rm31 + lm21 * rm32 + lm31 * right[15];

    float lm02 = left[8];
    float lm12 = left[9];
    float lm22 = left[10];
    float lm32 = left[11];

    buffer[8] = lm02 * rm00 + lm12 * rm01 + lm22 * rm02 + lm32 * right[12];
    buffer[9] = lm02 * rm10 + lm12 * rm11 + lm22 * rm12 + lm32 * right[13];
    buffer[10] = lm02 * rm20 + lm12 * rm21 + lm22 * rm22 + lm32 * right[14];
    buffer[11] = lm02 * rm30 + lm12 * rm31 + lm22 * rm32 + lm32 * right[15];

    float lm03 = left[12];
    float lm13 = left[13];
    float lm23 = left[14];
    float lm33 = left[15];

    buffer[12] = lm03 * rm00 + lm13 * rm01 + lm23 * rm02 + lm33 * right[12];
    buffer[13] = lm03 * rm10 + lm13 * rm11 + lm23 * rm12 + lm33 * right[13];
    buffer[14] = lm03 * rm20 + lm13 * rm21 + lm23 * rm22 + lm33 * right[14];
    buffer[15] = lm03 * rm30 + lm13 * rm31 + lm23 * rm32 + lm33 * right[15];
}