How big is TOO big?

I'm doing some simulation work with polymers and I've come across what, to me, seems like an arbitrary limitation on the chain lengths I can perform MD simulations on.  Specifically, I seem to be having a problem with polymers containing heteroatoms.

Here's what I'm doing, feel free to try it yourself and see what happens for you:

1.  Build a Nylon 6 polymer with the Polymer Builder (Build-->Build Polymers-->Homopolymer, amides library, nylon6 repeat unit).

2.  Build two polymers, one that is 334 repeat units long and one that is 335 units long.

3.  Run a simple Disco Dynamics run (see my error report below).

The 334 unit polymer runs fine, the 335 unit polymer returns the following error:

-----------------------------------------------------------------------------------------------------

Error: Miscellaneous:

* In dyn_remove_momentum(), after removing angular momentum:

* C.O.M. of coords used     = 0.000000 0.000000 -0.000000

* Adjusted linear  momentum = -0.000000 0.000000 -0.000000

* Adjusted angular momentum = -1.255456 -0.100612 1.192369

    invoked from within

"dynamics \

    time = 1000.00 \

    timestep = 1.00 \

    initial_temperature = 273.00 \

    ensemble = NVT \

    temperature_control_method = anderse ..."

Total time used by DISCOVER: 0.98 secs

---------------------------------------------------------------------------------------------------------

The 334 unit molecule has 6,348 atoms while the 335 unit molecule has 6,367.  The mass of each is around 37,900 g/mol which is a bit lower than I'd like.

What's the difference between 334 repeat units and 335 repeat units?  I've been able to get MD simulations to run on 150,000 g/mol polystyrene molecules with no* problem. 

Any input on the nature of the problem will be greatly appreciated -- I'm slowly hacking my way through Materials Studio and I love learning new things about it.

CS