Massively Parallel Reductions in Multivariate Polynomial Systems: Bridging the Symbolic Preprocessing Gap on GPGPU Architectures
v⊤A=0⟺∑i=1rvi(tigki)=0inݔ�p[x1,…,xn].v^{\top}A=0\quad\Longleftrightarrow\quad\sum_{i=1}^{r}v_{i}(t_{i}g_{k_{i}})=0\ \text{in}\ \mathbb{F}_{p}[x_{1},\dots,x_{n}]. Proof 11.5. The forward direction holds becausev⊤Av^{\top}Alists all coefficients of∑ivi(tigki)\sum_{i}v_{i}(t_{i}g_{k_{i}})in the basisݒ�\mathcal{T}. Support-closur...
