diff --git a/README.md b/README.md
index 1faa6f9..28befd3 100644
--- a/README.md
+++ b/README.md
@@ -7,39 +7,3 @@ This is an attempt at implementing a fast version of the algorithm described her
     IEEE Trans. on Information Theory, pp. 6284-6299, November, 2016.
 ~~~
 Available here: [http://ct.ee.ntust.edu.tw/it2016-2.pdf](http://ct.ee.ntust.edu.tw/it2016-2.pdf)
-
-I wasn't able to figure out how to speed up the critical piece of code:
-
-~~~
-  // z = x + y (mod Q)
-  static inline GFSymbol AddModQ(GFSymbol a, GFSymbol b)
-  {
-      const unsigned sum = (unsigned)a + b;
-
-      // Partial reduction step, allowing for Q to be returned
-      return static_cast<GFSymbol>(sum + (sum >> kGFBits));
-  }
-
-  // vx[] += vy[] * z
-  static void muladd_mem(GFSymbol * vx, const GFSymbol * vy, GFSymbol z, unsigned symbolCount)
-  {
-      for (unsigned i = 0; i < symbolCount; ++i)
-      {
-          const GFSymbol a = vy[i];
-          if (a == 0)
-              continue;
-
-          const GFSymbol sum = static_cast<GFSymbol>(AddModQ(GFLog[a], z));
-
-          vx[i] ^= GFExp[sum];
-      }
-  }
-~~~
-
-This routine basically takes all the runtime.  Since it involves normal addition via `AddModQ()`, it is pretty slow and is hard to vectorize.
-
-Math over GF(2^^16) is a lot faster because multiplies can be decomposed linearly into separate operations that can be done via vector instructions.
-
-So unfortunately it looks like while this field does have an FFT, the field does not permit fast math.
-
-Maybe someone else can figure out how to speed this up?