Using NetBSD we match the performance of one of Cisco's new routers utilizing their new Cisco QuantumFlow Processor - well except for the power consumption. Though our solution is non-green, it will take more than a couple of years before the additional x86 power-cost catches the Cisco equipment+IOS cost.
At a higher layer, you finally have the message-hub HW platform you never knew you needed when you first introduced your CIO to the ESB :-)
You realize that, given enough spindles, this 48 core machine actually makes the NCR/TeraData warehouse scale linearly when you query multi TB data.
Someone copyrights the CloudInABox term, and Intels sues them and wins, because it is too close the their future concept of CloudOnAChip.
You don't have to buy additional hw to handle 20.000+ concurrent Exchange users - but you will get ripped of by VMware and Microsoft.
Someone tries to upgrade their maxed out 4xQuad 3.33 GHz OLTP database server to this 48 core 2.2 GHz cloud-box - and they fail - because the query-queues fill up.
