> An independent panel was set up to scrutinise its plans and released a report suggesting some of its ideas were "tech for tech's sake", and potentially unnecessary.
I wonder what could have given them that idea.
> In his blog, Dan Doctoroff said the firm continued to invest in start-ups "working on everything from robotic furniture to digital electricity".
I'm always a bit suspicious about x86 vs. non-x86 micro-benchmarks. I remember all the fun people had with ByteMark back in the day, and while I assume that Geekbench doesn't play those sorts of games with compiler optimizations, I would really like to see data from something a bit more representative of a real-world CPU-bound application (short-lived is fine, just not synthetic).