• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Intel hex-core processor

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Originally posted by: Idontcare
It just strikes me as REALLY odd that Intel would go from dual-core straight to hexa-core monolithic Penryn die and completely skipping the quad-core monolithic rev.
That's a very good point. Perhaps these really are "glued" dice, somehow, with a seperate L3 controller chip? That would still require some pretty impressive engineering to pull off though, and the L3 would be sub-optimal in performance if all of the three dice had to access it via FSB.
Originally posted by: Idontcare
And all this effort for an EOL architecture? How long will they make Dunnington? 9 months? It just makes no sense NOW...had they released it 6 months ago, or had a 45nm 6-core Dunnington been the follow-up to a prior released 65nm monolithic quad-core (or tri-core with L3$) then I could see the logic in the SKU progression.
Wouldn't it make much more sense to make a monolithic quad-core, with shared L3, and then stick two of those underneath the heatspreader? It would make a sweet octo-core drop-in chip.

 
Originally posted by: VirtualLarry
That's a very good point. Perhaps these really are "glued" dice, somehow, with a seperate L3 controller chip? That would still require some pretty impressive engineering to pull off though, and the L3 would be sub-optimal in performance if all of the three dice had to access it via FSB.

If you think about it though Intel pioneered (to my knowledge) the MCM technique with the Pentium Pro and the seperate L2$ chip wired to the CPU on the same ceramic package.

Maybe not out of their capability. And in theory L3$ should be faster than FBDIMM DDR2 accesses. Don't ask me how Phenom managed to screw the pooch on this one, I freely admit to not being a cache hierarchy expert.

Originally posted by: VirtualLarry
Wouldn't it make much more sense to make a monolithic quad-core, with shared L3, and then stick two of those underneath the heatspreader? It would make a sweet octo-core drop-in chip.

Absolutely. Yet another reason the 6-core monolithic SKU continues to make no sense in the absence of all these other options.
 
That's a good point. Probably to avoid market confusion? (i.e. self-competition eroding margins by having similarly spec'ed two quad-cores)
 
Back
Top