|[eigen] Parallel .solve(B) for B.cols() > 1|
[ Thread Index |
| More lists.tuxfamily.org/eigen Archives
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: [eigen] Parallel .solve(B) for B.cols() > 1
- From: Julian Kent <julian.kent@xxxxxxxxx>
- Date: Wed, 4 Jan 2017 10:23:02 +0100
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=pix4d.com; s=google; h=mime-version:from:date:message-id:subject:to; bh=DI8Xq3t0il/KzE2o+F0Dm/Qvw+YZA61Hoh3EgNoda5g=; b=VQ5QksrQ0aKqdZzd+zrAkUx2J72/qZr0RlSUbEW38yp71UZO3bC2HxlD/0stv8QYR5 wSBnYOoBB3yBheTRHc1YaoAiIWNThOyxi1Zj1qC6TRZCc9Bm1TF7IIKBdLktt6/dM7Jk wT2gfTI0qvDk48bue7XOfrQ9MYY1eAY9OPWHI=
Is there a reason that Eigen doesn't do a parallel x = decomposition.solve(B) by splitting B into column-wise chunks?
Using CompleteOrthogonalDecomposition and solving for B=Identity (to solve pinv) I'm getting a 5.5x speedup on a 6-core CPU, for matrix size 4000 x 4000
Is it just considered a simple enough problem that it isn't necessary to have it done automatically, is this something that was overlooked, or am I missing something here?