Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
The 3825 is rated up to 350 Kpps (for fast switching - only 25 Kpps for process switching). For minimum size Ethernet, that would support about 235 Mbps (i.e. 14.88 Kpps per 10 Mbps). That's unidirectional, for bidirectional you would need to half that to about 118 Mbps. That of course is less than your 128 Mbps. Good news, bandwidth throughput usually increases with packet size and as most traffic packets sizes average larger than minimum size, i.e. effective throughput should be even better. Bad news, though, most routers' PPS rate drops with larger packets and, of course, you need CPU for other processes and any software switched packets takes more than 10x the CPU. I.e. real traffic, on "typical" configurations, often fail to come anywhere close to theoretical maximum.
In a later ISR performance document, Cisco recommends the 2911 (rated at 353 Kpps) for only 35 Mbps of "WAN" bandwidth.
Given the forgoing, I would have reservations about using a 3825 with 128 Mbps.
However, assuming everything else will remain the same except for doubling your bandwidth, how's your CPU stats look now? You might compare your CPU stats with concurrently bandwidth throughput, and roughly scale your CPU increase. If your scaled CPU looks as it will remain under 75%, you may be okay.
