Advanced Speed Optimization Techniques for QB

by Toshi (Toshihiro Horie)

• Optimize your algorithm first before you try these optimizations! That's usually where you get the biggest speed increases.

  Before:

  ' a very slowly converging formula for pi
  ' This takes a minute or so on a fast Pentium III.
  r# = 0
  s% = 1
  FOR i& = 1 TO 90000000 STEP 2
      r# = r# + s% * 1# / i&
      s% = -s%
  NEXT
  pi! = CSNG(r# * 4)
  PRINT pi!
  

  After:

  ' a fast! constant time formula for pi
  ' This is faster than you can blink.
  PI! = ATN(1) * 4
  PRINT pi!
  

  Once you've optimized the algorithm all that you can, you can start looking at algebraic and loop optimizations.

• The classic one is to use DEFINT A-Z. This forces you to use as many integer variables as possible.

• Use integer variables to index FOR loops. This may require substitution and algebraic simplification.

Before:

FOR i! = 0 to 0.3 STEP 0.01
    p! = i! * 3
NEXT


After:

FOR i% = 0 to 30
    p! = i% * 0.03
NEXT


• If your code has a lot of floating point calculations that need high accuracy, compile with QB 4.0.

  e.g. a raytracer

• If your code has a lot of floating point calculations that don't need more than 8 bits of accuracy, then definitely convert it to fixed point. Even if it needs up to 16 bits of accuracy, it might be worth converting to fixed point, if it is being used in the main loop.

  e.g. a rotozoomer or voxel terrain.

• Don't use IFs (conditional branches). Some comparison results can be directly be used in a calculation. Note that in QB, a TRUE boolean expression equals -1, and a FALSE one equals 0.

  Before:

  IF a > 4 THEN
      b = 5
  ELSE
      b = 0
  END IF
  

  After:

  b = -5 * (a > 4)
  

  Actually, the above example is too simple for the After: version to be faster. But for more complicated expressions involving multiplication and division, it can make a difference.

• Buffer your reads from a file. This is especially useful in a non-disk-cached environment like DOS 4.0.

  Before:

  DIM c AS STRING * 1
  OPEN "file.bin" FOR BINARY AS #1
  FOR i = 0 TO 10 * 256
      GET #1, , c
  NEXT
  CLOSE #1
  

  After:

  DIM buffer(0) AS STRING * 256
  OPEN "file.bin" FOR BINARY AS #1
  FOR i = 0 TO 10
      GET #1, , buffer(0)
  NEXT
  CLOSE #1
  

• Use an assembler keyboard handler or INP(&H60) plus keyboard buffer clearing routines instead of INKEY$.

  e.g. For a user controlled floormapper routine, this made a huge difference in rending fps.

• For a straight QB multikey handler, don't bother to clear the keyboard buffer every vertical retrace. Instead, slow down the keyboard repeat rate, and check every few frames.

  e.g. This made a huge difference in QBMKEY.BAS

• Use integer division for integers.

  Before:

  x% = x% / y%
  

  After:

  x% = x% \ y%
  

• Make an integer division lookup table if there is a division slowing down the inner loop.

• Store the results of complicated expressions in look-up tables.

  Before:

  pi = ATN(1) * 4
  DO
      FOR i = 0 to 360
          x! = 100 + COS(i * pi / 180!)
          y! = 100 + SIN(i * pi / 180!)
          PSET (x!, y!), c
      NEXT i
  LOOP UNTIL LEN(INKEY$)
  

  After:

  pi = ATN(1) * 4
  
  Editor's note: I think some code is missing here...

• If several complicated expressions in a loop has common subexpressions, move the common subexpressions out of the loop.

  Before:

  FOR x = 1 to 32767
      FOR y = 1 to 10
          c = sin(x) * 30 + sqr(x) + y
      NEXT y
  NEXT x
  

  After:

  FOR x = 1 to 32767
      xc = sin(x) * 30 + sqr(x)
      FOR y = 1 to 10
          c = xc + y
      NEXT
  NEXT
  

• Make constants CONST. Unfortunately, you can't use transcendental functions like ATN on the right side anymore.

  Before:

  pi = ATN(1) * 4
  piover2 = pi / 2
  

  After:

  CONST pi = 3.14159265358979#
  CONST piover2 = pi/2
  

• Unroll short loops.

  Before:

  FOR a = 1 to 8
      POKE a, a
  NEXT
  

  After:

  POKE 1,1
  POKE 2,2
  POKE 3,3
  POKE 4,4
  POKE 5,5
  POKE 6,6
  POKE 7,7
  POKE 8,8
  

• Partially unroll long loops.

  Before:

           
  FOR x = 0 TO 319
      POKE x,a
  NEXT
  

  After:

  ' this is a silly example, you should be using
  ' MMX filling or REP STOSB at least.
  FOR x = 0 TO 319 STEP 4
      POKE x, a
      POKE x + 1, a
      POKE x + 2, a
      POKE x + 3, a
  NEXT x
  

• Move junk outside of the inner loops (code movement).

  Before:

  FOR y = 0 TO 199
      FOR x = 0 TO 319
          a = x * 4 + COS(t)
          b = y * 3 + SIN(t)
      NEXT
  NEXT
  

  After:

  FOR y = 0 TO 199
      b = y * 3 + SIN(t)
      FOR x = 0 TO 319
          a = x * 4 + COS(t)
      NEXT
  NEXT
  

• Use cache sensitive programming. This means, try to access your arrays in a sequential manner if possible. If not, access them in small blocks that are adjacent to each other. For example, QB arrays are usually stored in a column major order, so dimension your arrays as vscreen(xmax,ymax) if you are doing scanline-based algorithms, and only change move in the x (scanline) direction in the inner loop.

  Before:

  '$DYNAMIC
  xmax = 319: ymax = 199
  DIM buf(xmax, ymax)
  FOR x = 0 TO xmax
      FOR y = 0 TO ymax
          buf(x, y) = INT(RND * 256)
      NEXT
  NEXT
  DEF SEG
  

  After:

  '$DYNAMIC
  xmax = 319: ymax = 199
  DIM buf(xmax, ymax)
  FOR y = 0 TO ymax
      FOR x = 0 TO xmax
  buf(x, y) = INT(RND * 256)
      NEXT
  NEXT
  DEF SEG
  

• Use a precalculated (canned) pseudo-random number sequence.

  Before:

  'main loop
  FOR i = 1 TO 1000
      x = INT(RND * 256)
      y = INT(RND * 256)
      c = INT(RND * 256)
      PSET (x, y), c
  NEXT i
  

  After:

  'precalculation
  DIM rand(8191)
  FOR i = 0 TO 8191
      rand(i) = INT(RND * 256)
  NEXT i
  
  'main loop     
  count = 0
  FOR i = 1 TO 1000
      x = rand(count)
      y = rand(count + 1)
      c = rand(count + 2)
      PSET (x, y), c
      count = (count + 3) ' AND 8192 (needed in general)
  NEXT i
  

• Prefer array indexing over user defined TYPEs. (1)

  Warning: This makes code unreadable unless it is well commented.

• Cache often-used array elements in scalar variables. (2)

• Cache intermediate values into temporary variables. (3)

  Example of both optimizations being used.

  Before:

   
  TYPE PtType
      x AS INTEGER
      y AS INTEGER
      z AS INTEGER
  END TYPE
  TYPE TriType
      pt1 AS INTEGER 'index of first point in points array
      pt2 AS INTEGER 'index of second point in points array
      pt3 AS INTEGER 'index of third point in points array
  END TYPE
  DIM points(numpoints, 1 TO 3) AS PtType
  DIM tri(numtriangles) AS TriType
  CONST screendist = 200
  CONST lightx = 1, lighty = 0, lightz = 0
  CALL loadobject(filename$, points())
   
  FOR i = 1 TO numtriangles
      V1x = points(tri(i).pt2).x - points(tri(i).pt1).x
      V2x = points(tri(i).pt3).x - points(tri(i).pt1).x
      V1y = points(tri(i).pt2).y - points(tri(i).pt1).y
      V2y = points(tri(i).pt3).y - points(tri(i).pt1).y
      V1z = points(tri(i).pt2).z - points(tri(i).pt1).z
      V2z = points(tri(i).pt3).z - points(tri(i).pt1).z
  
      length1 = SQR(V1x * V1x + V1y * V1y + V1z + V1z)
      length2 = SQR(V2x * V2x + V2y * V2y + V2z + V2z)
  
      vx = V1y * V2z - V2y * V1z
      vy = V2x * V1z - V1x * V2z
      vz = V1x * V2y - V2x * V1y
      CALL normalize(vx, vy, vz)
      brightness = vx * lightx + vy * lighty + vz * lightz
  
      xp1 = screendist * x1 / z1
      yp1 = screendist * y1 / z1
      xp2 = screendist * y1 / z1
      yp2 = screendist * y2 / z2
  
      '... and so on...     
  NEXT i
  

  After:

  ' index 1 = x coordinate of point
  ' index 2 = y coordinate of point
  ' index 3 = z coordinate of point
  DIM points(numpoints, 1 TO 3)
  DIM tri(numtriangles, 1 TO 3)
  CONST screendist = 200
  CONST lightx = 1, lighty = 0, lightz = 0
  CALL loadobject(filename$, points())
  
  FOR i = 1 TO numtriangles
      x1 = points(tri(i, 1), 1) ' example of optimization 1
      y1 = points(tri(i, 1), 2) ' and optimization 2
      z1 = points(tri(i, 1), 3)
      x2 = points(tri(i, 2), 1)
      y2 = points(tri(i, 2), 2)
      z2 = points(tri(i, 2), 3)
      x3 = points(tri(i, 2), 1)
      y3 = points(tri(i, 2), 2)
      z3 = points(tri(i, 2), 3)
      V1x = (x2 - x1): V2x = (x3 - x1)
      V1y = (y2 - y1): V2y = (y3 - y1)
      V1z = (z2 - z1): V2z = (z3 - z1)
  
      length1 = SQR(V1x * V1x + V1y * V1y + V1z + V1z)
      length2 = SQR(V2x * V2x + V2y * V2y + V2z + V2z)
  
      vx = V1y * V2z - V2y * V1z
      vy = V2x * V1z - V1x * V2z
      vz = V1x * V2y - V2x * V1y
      CALL normalize(vx, vy, vz)
      brightness = vx * lightx + vy * lighty + vz * lightz
  
      xp1 = screendist * x1 / z1
      yp1 = screendist * y1 / z1
      xp2 = screendist * y1 / z1
      yp2 = screendist * y2 / z2
  
      '... and so on...     
  NEXT i
  

• Use REDIM to clear a large array instead of using a FOR loop to set each element to zero.

  Before:

  DIM x(32000)
  DO
      FOR i = 0 TO 32000
          x(i) = 0 'clear array slowly
      NEXT i
      x(RND * 32000) = 50
      x(RND * 32000) = 93
  LOOP UNTIL LEN(INKEY$)
  

  After:

  DIM x(32000)
  DO
      REDIM x(32000) 'clear array faster
      x(RND * 32000) = 50
      x(RND * 32000) = 93
  LOOP UNTIL LEN(INKEY$)
  

• Avoid multidimensional arrays. Use array head lookup tables like in the POKE vs. PSET example for faster access of single dimension arrays as multidimensional ones.

  Before:

  DIM x(63, 63)
  

  After:

  DIM x(4095)
  

• Don't waste an extra element. Unlike C arrays, the declaration of QB arrays specify the first and last element indicies rather than the size of the array. This matters when you want to make a 64KB array without using '$DYNAMIC.

  Before:

  DIM x(256, 256) 'allocate 66049 elements
  

  After:

  DIM x(0 TO 255, 0 TO 255) 'allocate 66536 elements
  

  or

  'OPTION BASE 0
  DIM x(255, 255)
  

• Use incremental calculation instead of evaluating the entire equation every loop. This usually means multiplies will be replaced by addition. It's very important that you do this in any linear interpolation function you use for Gouraud Shading, Texture Mapping, etc. Most line DDAs (digital difference analyzers) use this method.

  Before:

  slope! = 0.1
  FOR x = 0 TO max
      y! = slope! * x
  NEXT x
  

  After:

  slope! = 0.1
  y! = 0
  FOR x = 0 TO max
      y! = y! + slope!         
  NEXT x
  

• Use POKE instead of PSET. This is a simple way to get 2x performance in graphics intensive apps.

  Before:

  CONST xmax = 319, ymax = 199, scansize& = 320
  FOR i = 0 TO 255
      PSET (i, 0), i
  NEXT i
  FOR i = 0 TO 255
      PSET (i, 10), i
  NEXT i
  

  After:

  CONST xmax = 319, ymax = 199, scansize& = 320
  DIM ytab&(ymax)
  FOR y = 0 to ymax
      ytab&(y) = y * scansize&
  NEXT y
       
  DEF SEG = &HA000
  FOR i = 0 TO 255
      POKE i, i
  NEXT i
  FOR i = 0 TO 255
      POKE ytab&(10) + i, i
  NEXT i
  DEF SEG       
  

• Use INTEGER variables instead of LONGs for unsigned integers in the range 0 to 65535. This will only work when the program is compiled.

• PEEKing from video memory is slower than PEEKing from system memory. Therefore, use double buffering when you need to do feedback effects.

• Use DEF SEG sparingly. You don't need to DEF SEG back to the default segment when you are accessing arrays in the default segment. DEF SEG only applies to PEEK and POKE and SETMEM.

• Don't use '$DYNAMIC. QB arrays in the default segment are accessed at blazing speed, because there is no segment switching. However, '$DYNAMIC puts them in different segments, which need extra instructions to accessed, slowing them down. This makes a big difference in programs that use large lookup tables in their inner loop. It seems that huge arrays (allowed using the QB/AH command) are the slowest to access.

  Before:

  '$DYNAMIC
  DIM hugetable(319, 199)
  FOR y = 0 TO 199
      FOR x = 0 TO 319
          xo = (x - 160) \ 2
          yo = (y - 100) \ 2
          hugetable(x, y) = xo * xo + yo * yo
      NEXT
  NEXT
  

  After:

  '$STATIC
  DIM hugetable1(319, 99)
  DIM hugetable2(319, 99)
  
  FOR y = 0 TO 99
      FOR x = 0 TO 319
          xo = (x - 160) \ 2
          yo = (y - 100) \ 2
          hugetable(x, y) = xo * xo + yo * yo
      NEXT
  NEXT
  FOR y = 100 TO 199
      FOR x = 0 TO 319
          xo = (x - 160) \ 2
          yo = (y - 100) \ 2
          hugetable(x, y - 100) = xo * xo + yo * yo
      NEXT
  NEXT
  

• Use SELECT CASE instead of a bunch of ELSEIFs. The only exception is when one case executes much more often than the others.

  Before:

  IF i = 1 THEN
      CALL DrawSprite
  ELSEIF i = 6 THEN
      CALL PlaySound
  ELSEIF i > 9 AND i < 16 THEN
      CALL Calculate(i)
  ELSE
      PRINT "."
  END IF
  

  After:

  SELECT CASE i
      CASE 1
          CALL DrawSprite
      CASE 6
          CALL PlaySound
      CASE 10 TO 15
          CALL Calculate(i)
      CASE ELSE
          PRINT "."
  END SELECT
  

• Use AND instead of MOD for MODing by a power of 2.

  Before:

  a = b MOD 64
  

  After:

  a = b AND 63
  

• Simplify compares against zero.

  Before:

  IF a <> 0 THEN
      b% = b% - 1
  END IF
  

  After:

  IF a% THEN  'note <>0 is gone
      b% = b% - 1
  END IF
  

• Use -x to find the negative of a number instead of -1*x. This is an obvious optimization if you know that the CPU has a NEG instruction, which is faster than IMUL.

• Don't put the main loop in the main code-- put it in a SUB. This makes a difference in the IDE, probably because the p-code interpreter has less variables to wade through when you are in a SUB.

• Use static storage for non-recursive SUB parameters. This makes very little improvement in speed, unless there are tons of variables passed to a SUB.

  Before:

  SUB drawcircle(x%,y%,r%)
      'routine to draw a circle
  END SUB
  

  After:

  SUB drawcircle(x%,y%,r%) STATIC
      'routine to draw a circle
  END SUB
  

• Pass dummy parameters to functions that take an odd number of arguments in order to improve data alignment. The dummy parameter is not used by the function, but is there to encourage burst memory writes. This only makes a minimal difference in speed.

  Before:

  CALL drawcircle(x%, y%, r%)
  

  After:

  CALL drawcircle(x%, y%, r%, dummy%)
  

• Don't initialize QB array elements to zero. Warning: this is a dangerous habit to get into, if you plan to use C or C++ later on. This is because C does not initialize variables by default.

  Before:

  DIM div320(32767)
  FOR i = 0 TO 32767
      div320(i) = i \ 320
  NEXT
  

  After:

  DIM div320(32767)
  FOR i = 320 TO 32767
      div320(i) = i \ 320
  NEXT
  

• For floating point, multiply by the reciprocal of a number instead of dividing by a number.

  Before:

  SUB normalize (x!, y!, z!)
      norm! = SQR(x! * x! + y! * y! + z! * z!)
      x! = x! / norm!
      y! = y! / norm!
      z! = z! / norm!
  END SUB
  

  After:

  SUB normalize (x!, y!, z!)
      recipnorm! = 1 / SQR(x! * x! + y! * y! + z! * z!)
      x! = x! * recipnorm!
      y! = y! * recipnorm!
      z! = z! * recipnorm!
  END SUB
  

• Simplify comparisons using simpler monotonic functions. Monotonic functions are functions that always grow upwards or always grow downwards. For example, x^2 is a monotonic function of x, so is 2*x. In the example, an expensive square root was removed by squaring both sides, since squaring is a monotonic function.

  Before:

  dist = SQR(x * x + y * y)
  IF dist < radius THEN
      'inside circle
  END IF
  

  After:

  r2 = radius * radius
  distsquared = x * x + y * y
  IF distsquared < r2 THEN
      'inside circle
  END IF
  

Advanced optimizations for C++

Thanks for Qasir, entropy, Pasco, and Eclipzer for their critiques and suggestions.