It's 6 AM and I've been playing Halo Reach all night with the mates.
Originally Posted by kartcrg84
Just because hand crafting/automated design didn't cause the delays doesn't mean that hand crafting as apposed to automated design would have caused things to take even longer.
Except they didn't, issues with production and process caused the delays. As well as some (semi)poor design decisions on AMD's part. To a large extent, AMD was too forward thinking in design.
Originally Posted by CULLEN
The questions is, do you see your self with BD any time soon? Please explain.
I've got a pre-production sample. It's a good chip, can do 5.5Ghz on cheap DIY water solutions (like a RASA RX240.)
My current rig is mATX and I'm not even going to try and take it apart to put BD inside it. If I was running a tower, I most likely would, and I would disable core 1,3,5, & 7 so that I could obtain higher IPC.
Or set Core Affinity for most tasks to Core 0, 2, 4, & 6.
If I was looking to upgrade from what I had and performance was my only concern (usually is.) Then I would most likely just put more money into a better/faster GPU instead of BD.
For gaming purposes, Thuban does just fine. This particular chip can do 4.4Ghz easy on a Silver Arrow so I've not really got a workload that demands more of my system than what the CPU can produce.
Originally Posted by t00sl0w
so these global foundries mishaps could actually be causing the internals of the chip to not "communicate" properly and be extremely inefficient at the same time?
No, high leakage (which increases power draw & heat output) is an issue of the wonky process. Though, if you had a communication breakdown (like the first generation of Phenom chips) then it's most likely an on chip issue. Not a process/manufacturing issue.