How can I use fixed point types with more than 2080 bits?

In our Simulink / Altera DSP Builder design we use a wide array of type ufix() to distribute information from configuration registers around the system. Now this array has grown to more than 2080 bits, which seems to be the limit for fixed point types. Is there a way to get wider data types than 2080 bits?

Of course one option would be to redesign our system to use e.g. a vector of fixed point data, which would fix this issue, but we would like to avoid this redesign at this time since it would affect many parts of the design.

[Nick Choi]

It would be helpful to see what is going on within the model. Could you provide a reproduction model?

Also, how are you determining that the array is approaching 2080 bits? For simulation and code generation, there is a limit of 128 bits for fixed-point data types as described in the following documentation link:

https://www.mathworks.com/help/fixedpoint/ug/data-type-support_bs2nc0s.html

In MATLAB, the 'fi' object has a 'WordLength' limit of 65535.