diff --git a/Eigen/src/Core/products/GeneralBlockPanelKernel.h b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
index 229e96ceb..37e2e39f1 100644
--- a/Eigen/src/Core/products/GeneralBlockPanelKernel.h
+++ b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
@@ -1628,9 +1628,10 @@ void gebp_kernel<LhsScalar,RhsScalar,Index,DataMapper,mr,nr,ConjugateLhs,Conjuga
           prefetch(&blA[0]);
           const RhsScalar* blB = &blockB[j2*strideB+offsetB*nr];
 
-          if( (SwappedTraits::LhsProgress % 4)==0 )
+          // NOTE The following piece of code doesn't work for 512 bit registers,
+          // so we don't call it for registers that contain more than 8 values.
+          if( ((SwappedTraits::LhsProgress % 4)==0) && (SwappedTraits::LhsProgress <= 8))
           {
-            // NOTE The following piece of code wont work for 512 bit registers
             SAccPacket C0, C1, C2, C3;
             straits.initAcc(C0);
             straits.initAcc(C1);