This was one of the main recommendations (amongst 30 or so) from a Nuclear industry panel (Westinghouse, General Electric, Bechtel, Sargent & Lundy, Northern States Power and Commonwealth Edison) commissioned by the NRC. These should have been included in standardised Nuclear power station designs like the AP-1000, however they made the plants more expensive.

No, the AC is wrong. Yucca mountain has ground water issues that affect the storage of the material. CSIRO research showed that groundwater issues are mitigated by granite storage which can capture the isotope in its structure. DOE itself called for 'defence in depth' and it's own report judged Yucca to be unsuitable as groundwater penetrated the facility in as little as 50 years.

Except that it's not been done. When Dixie Lee Ray was the head of the Atomic Energy Commission he proclaimed that the disposal of nuclear fuel would be “the greatest non-problem in history” and would be accomplished by 1985, yet here we are almost thirty years past that date and still there is no High level waste disposal site anywhere. The closest anyone has come is the Swiss and even there project is a multi-decade test project and extremely expensive.

As for burner reactor technology, such as IFR, there are no materials technologies to support a plutonium economy.

Yeah, her record as SecState is only less marked by accomplishment than her stint in the Senate.

Actually, "client" workloads (personal computers) aren't very parallel so the requests are served sequentially. As such, this won't help too much.
      Even now, hard drive performance (if computed separately for client-workloads and server-workloads) uses queue length 1 for client and queue length 32 for server.
      By the way, if I remember correctly multiple requests on flight were implemented on SATA standard for client drives, 10 years ago or so on (SCSI had them for quite a while). I'm not sure Windows XP uses these queues.

>which is the fork and which the original ... is a matter of sometimes heated opinion.

I don't care too much about that. I am interested in which is better though.

Which is better?

10 times ago I heard about IBM and others working on new technologies to replace memory. Holographic cubes, MRAM, ... Are they still 10 years away?