Usenet comes to mind.
Use tables that only have a ++id, and a blob.
Windows and Flash are of straw.
Just change the name to Wobbit. Or Habbit.
Amazing that nobody mentioned Cascalog https://github.com/nathanmarz/cascalog
My favorite boxes currently are HP-G7 with 2x 640 GB Fusion-IO and some hard disk. The data is in partitioned tables (MySQL 5.5), and I run for example 20 SELECT (partition-optimized) queries in parallel on them.
This is much faster (for me) than for example a Hadoop cluster. Mainly because the data is already where I query it. For me the copying of big data sets is the main bottle neck.
My aggregations normally finish in 1-5 minutes. One box does 60 GB per minute, 4 boxes do 90 GB per minute, so no big win by adding boxes.
A one hour flight on average takes 3 hours of your life, doesn't it?
The new Gotthard was built for USD 10 bln, took 20 years:
Just beat the speed of light, by pre-calculating many scenarios, such that you don't need to calculate anymore once the fresh parameters arrive.
When people dialed in on our conference call systems with Skype, they most of the times sounded too unclear to be properly understood, which slowed down the groups in the meetings. So we had to ban it.
I use for example voipbuster, about EUR 40 per year for free calls to almost any number in the world that I might want to ever call, and it comes with a phone number too. Sound quality is just as good as any land line.
The DNA replication in a human body does about 500 TB per second.