Why so stingy with the kit?
I wonder why the author of this report chose to build TSM (for it is indeed a TSM 6.3 server rather than Spectrum Protect 7.1 server) on a Windows platform with less than the supported amount of memory. 8GB is scant for a laptop these days, and below the supported minimum for a TSM 6.3 server, but this is what has been demonstrated in the paper.
A typical Spectrum Protect server these days (for example, the "large" blueprint) would have about 20x more than that - memory isn't all that expensive.
Indeed, why bother demonstrating a TSM 6.3 server at all? TSM 6.3 came out in 2011 and will go end of service in April next year, it was withdrawn from marketing last month.
TSM 7 has Aspera FASP (guess what, a software defined WAN accelerator) as an optional feature (it's shipped within Protect from v7.1.6+).
Perhaps that's part of the reason for this report being based on v6?
Anon for obviously-too-well-informed reasons.