There are numerous differences between the two networks, both on the conceptual and technical side. It would be shorter to show the similarities 
What’s most appealing to me (especially that I’m not the right person for technical discussions…) is that the Nu network relies on separation of the speculative element (NuShares) and the store/ transmitter -of-value element (NuBits), whereas Ripple is in itself a speculative element and a store/transmitter-of-value. It’s an advantage to have the two elements separated.
To quote one of the RippleLabs employee:
“BitCoiners believe that as long as the demand to use BitCoin (<-- the system) keeps increasing, the price of BTC (<-- the currency) will keep increasing. That is a logical statement when it comes to the BitCoin world. However, that logic doesn’t apply at all to Ripple (<-- the system). There is an unlimited supple of an unbounded number of different currencies. Since you can hold any currency and make payments in any other currency, no currency (including XRP) has transactional advantages over any other. So as the Ripple System economy increases that puts no upward pressure on the price of XRP.”
https://forum.ripple.com/viewtopic.php?f=1&t=2814&hilit=peg&start=10#p10010
This was taken from a discussion about the “threat” of XRP increasing in value, it’s interesting in itself, but this is the most interesting point: There is an unlimited supple of an unbounded number of different currencies. As with Nu network, an additional number of NuBits may be created as a result of consensus among shareholders, in the Ripple network, new currencies (and new quantities of a given currency) can be created by gateways. Will this be a healthy situation in the future, as the XRP network grows? It’s still an experiment, just like Nu is an experiment, but I perceive Ripple to be more of a commercial experiment, with a company standing behind it, with centralization and a massive amount of trust needed, especially in terms of gateways. The fact of a company maintaining the network and having control over it might be a strong element for banks to consider working with such a network, of implementing the protocol for their accounting systems. The German Fidor Bank has done it. For me, a stronger argument in favour is the way Nu was designed to work, it’s obviously decentralized, and distributed, autonomous, a DAB, with B standing for bank 
