Microoptimization: reduced number of allocations

README.md

@@ -43,13 +43,13 @@ Anagrams generation is not parallelized, as even single-threaded performance for
 
 Multi-threaded performance with RyuJIT (.NET 4.6, 64-bit system) on quad-core Sandy Bridge @2.8GHz is as follows (excluding initialization time of 0.2 seconds):
 
-* If only phrases of at most 4 words are allowed, then it takes **1.5 seconds** to find and check all 7433016 anagrams; **all hashes are solved in first 0.2 seconds**.
+* If only phrases of at most 4 words are allowed, then it takes **1.1 seconds** to find and check all 7433016 anagrams; **all hashes are solved in first 0.2 seconds**.
 
-* If phrases of 5 words are allowed as well, then it takes 3.5 minutes to find and check all 1348876896 anagrams; all hashes are solved in less than 5 seconds.
+* If phrases of 5 words are allowed as well, then it takes 2:45 minutes to find and check all 1348876896 anagrams; all hashes are solved in first 4 seconds.
 
-* If phrases of 6 words are allowed as well, then "more difficult" hash is solved in 4.5 seconds, "easiest" in 28 seconds, and "hard" in 70 seconds.
+* If phrases of 6 words are allowed as well, then "more difficult" hash is solved in 3.5 seconds, "easiest" in 21 seconds, and "hard" in 54 seconds.
 
-* If phrases of 7 words are allowed as well, then "more difficult" hash is solved in 27 seconds, "easiest" in less than 3.5 minutes, and "hard" in 9.5 minutes.
+* If phrases of 7 words are allowed as well, then "more difficult" hash is solved in 20 seconds, "easiest" in less than 2.5 minutes, and "hard" in 6:45 minutes.
 
 Note that all measurements were done on a Release build; Debug build is significantly slower.
 
@@ -111,6 +111,4 @@ There is no need in processing all the words that are too large to be useful at
 11. Filtering the original dictionary (e.g. throwing away all single-letter words) does not really improve the performance, thanks to the optimizations mentioned in notes 7-9.
 This solution finds all anagrams, including those with single-letter words.
 
-12. MD5 computation could be further optimized by:
+12. MD5 computation could be further optimized by leveraging CPU extensions; however, it could not be done with current .NET (see readme for https://github.com/penartur/TrustPilotChallenge/tree/simd-md5)
-    * Using CPU instructions for rotation (implemented in not yet released version of RyuJIT): https://github.com/dotnet/coreclr/pull/1830
-    * Computing several MD5 hashes in parallel on each core, using SSE (4 hashes / core) or AVX2 (8 hashes / core). However, even bit shifts on vectors are not yet supported by .NET: https://github.com/dotnet/coreclr/issues/3226

dotnet/WhiteRabbit/Phrase.cs

@@ -5,13 +5,13 @@
     {
         public fixed uint Buffer[8];
 
-        public Phrase(byte[][] words, int numberOfCharacters)
+        public Phrase(byte[][] words, PermutationsGenerator.Permutation permutation, int numberOfCharacters)
         {
             fixed (uint* bufferPointer = this.Buffer)
             {
                 var length = numberOfCharacters + words.Length - 1;
 
-                byte[] currentWord = words[0];
+                byte[] currentWord = words[permutation[0]];
                 var j = 0;
                 var wordIndex = 0;
                 var currentPointer = (byte*)bufferPointer;
@@ -22,7 +22,7 @@
                     {
                         j = 0;
                         wordIndex++;
-                        currentWord = words[wordIndex];
+                        currentWord = words[permutation[wordIndex]];
                     }
 
                     *currentPointer = currentWord[j];

dotnet/WhiteRabbit/StringsProcessor.cs

@@ -62,23 +62,7 @@
             return sums
                 .Select(this.ConvertVectorsToWords)
                 .SelectMany(Flattener.Flatten)
-                .SelectMany(GeneratePermutations)
+                .SelectMany(this.ConvertWordsToPhrases);
-                .Select(this.ConvertWordsToPhrase);
-        }
-
-        private static IEnumerable<T[]> GeneratePermutations<T>(T[] original)
-        {
-            var length = original.Length;
-            foreach (var permutation in PrecomputedPermutationsGenerator.HamiltonianPermutations(length))
-            {
-                var result = new T[length];
-                for (var i = 0; i < length; i++)
-                {
-                    result[i] = original[permutation[i]];
-                }
-
-                yield return result;
-            }
         }
 
         private byte[][][] ConvertVectorsToWords(int[] vectors)
@@ -93,9 +77,12 @@
             return words;
         }
 
-        private unsafe Phrase ConvertWordsToPhrase(byte[][] words)
+        private IEnumerable<Phrase> ConvertWordsToPhrases(byte[][] words)
         {
-            return new Phrase(words, this.NumberOfCharacters);
+            foreach (var permutation in PrecomputedPermutationsGenerator.HamiltonianPermutations(words.Length))
+            {
+                yield return new Phrase(words, permutation, this.NumberOfCharacters);
+            }
         }
     }
 }