Speedy Matrix Multiplication in Lisp, Again… June 30th, 2009
Patrick Stein

In response to my earlier post about hidden memory allocations, Jason Cornez pointed out how to rework my loop so that Allegro wouldn’t need to allocate memory. The new structure avoids using the summing keyword in (loop …) and doesn’t use the return-value from (loop …).

(declaim (ftype (function ((simple-array single-float (12))
                           (simple-array single-float (3))
                           (simple-array single-float (3)))
                          (simple-array single-float (3))) mvl*-na))
(defun mvl*-na (matrix vec ret)
  (declare (type (simple-array single-float (12)) matrix)
           (type (simple-array single-float (3)) vec)
           (type (simple-array single-float (3)) ret)
           (optimize (speed 3) (safety 0)))
  (loop for jj fixnum from 0 below 3
     do (let ((offset (* jj 4)))
          (declare (type fixnum offset))
          (let ((current (aref matrix (+ offset 3))))
            (declare (type single-float current))
            (loop for ii fixnum from 0 below 3
               for kk fixnum from offset below (+ offset 3)
               do (incf current
                        (* (aref vec ii)
                           (aref matrix kk))))
            (setf (aref ret jj) current))))
  ret)

Now, the Allegro moved up in the rankings. All of the implementations (except 32-bit Clozure) ran slightly faster with this version. Allegro improved the most here though as its allocation/garbage dropped to nothing.

  wall non-GC alloced
SBCL 1.0.29 0.406s 0.406s 0
CMUCL 19f 0.539s 0.539s 32
Allegro 8.1 (Personal) 0.720s 0.720s 0
Clozure-64bit 1.3-r11936 1.100s 1.100s ?? 0
Clozure-32bit 1.3-r11936 6.884s 5.023s 1,680,000,000
LispWorks 5.1 (Personal) 12.587s ?? 3,120,000,516
ECL 9.6.2 30.875s ?? 17,280,000,256
GNU CLISP 2.47 88.008s 74.455s 2,160,000,000

CMUCL compiled with 30 Notes. I should look into those some time. It still ran quite fast though.

Allegro definitely liked this version better. It kept from allocating anything.

For comparison, I am including the previous runtimes here as well:

  wall non-GC alloced
SBCL 1.0.29 0.444s 0.444s 0
CMUCL 19f 0.567s 0.567s 0
Clozure-64bit 1.3-r11936 1.272s 1.272s ?? 0
Clozure-32bit 1.3-r11936 5.009s 4.418s 1,200,000,000
Allegro 8.1 (Personal) 6.131s 2.120s 1,440,000,000
LispWorks 5.1 (Personal) 14.054s ?? 3,360,000,480
ECL 9.6.2 33.009s ?? 18,240,000,256
GNU CLISP 2.47 93.190s 77.356s 2,520,000,000

Related posts:

  1. Lisp Troubles: Fabricating a Closure… Edit: lispnik on Livejournal pointed out this section of the Hyperspec which gaurantees that my pointer math is valid. I’d still like to know how to make a closure like this though, for future reference. Edit(2): See tfb’s comment below for a solution that does make a closure using a...
  2. To CLOS or not to CLOS I am working on some Lisp code. I am trying to mimic the basic structure of a large C++ project. I think the way the C++ project is structured is a good fit for the tasks involved. Most of the C++ stuff is done with classes. Most of the methods...
  3. Trying to Unconfound Lisp Speeds In the original version of my Optimizing Lisp Some More article, I did a bad comparison between SBCL and Clozure. SBCL supports two different ways to declare the arguments to a function. Clozure only supports one of those ways. As such, my declarations didn’t matter at all to Clozure. I...
  4. Even More Optimization In the previous article, I showed a chunk of code for multiplying a vector by a matrix. Nikodemus Silvola, on the sbcl-help mailing list, recommended that I change the matrix to a one-dimensional array instead of a two dimensional array. Doing that takes me from ten million iterations in 1.038...
  5. Optimizing Lisp Some More Today, I spent a large chunk of the day trying optimize a very small chunk of code. The code takes a matrix with three rows and four columns. It multiplies it by a vector that has three rows (and a pretend fourth row that is set to one). Here is...

RSS Feeds

l