diff --git a/PERFORMANCE.md b/PERFORMANCE.md
index 58397f1..a88b3ef 100644
--- a/PERFORMANCE.md
+++ b/PERFORMANCE.md
@@ -50,7 +50,9 @@ Currently the FFT benchmark shows the following results (average):
 We want to positively note that the overall trend of 'performance drops'
 at the hyperthreading mark is not really bad anymore, it rather
 seems similar to EMBB now (with backoff + lockfree deque + top level
-reader-writers lock).
+reader-writers lock). This comes partly because the spike at 4 threads
+is lower (less performance at 4 threads). We also see better times
+on the multiprogramed system with the lock-free deque.
 
 This is discouraging after many tests. To see where the overhead lies
 we also implemented the unbalanced tree search benchmark,
diff --git a/media/18b2d744_fft_average.png b/media/18b2d744_fft_average.png
index f885ee8..d8a6017 100644
Binary files a/media/18b2d744_fft_average.png and b/media/18b2d744_fft_average.png differ
diff --git a/media/18b2d744_unbalanced_average.png b/media/18b2d744_unbalanced_average.png
index bd2ac17..2a2fded 100644
Binary files a/media/18b2d744_unbalanced_average.png and b/media/18b2d744_unbalanced_average.png differ
diff --git a/media/aa27064_fft_average.png b/media/aa27064_fft_average.png
index 3441678..74229a0 100644
Binary files a/media/aa27064_fft_average.png and b/media/aa27064_fft_average.png differ
diff --git a/media/d16ad3e_fft_average.png b/media/d16ad3e_fft_average.png
index 78730f9..a29018b 100644
Binary files a/media/d16ad3e_fft_average.png and b/media/d16ad3e_fft_average.png differ