techreport.com — The Tech Report has quad-core Kentsfield benchmarks straight from the Intel Developer Forum, and the chip looks like a beast in multithreaded apps like 3D rendering and video encoding. What's more, the chip will plug into standard LGA775 motherboards, and could become available before AMD's '4x4' platform, which use dual dual-core chips.
Sep 28, 2006 View in Crawl 4
mike503Sep 29, 2006
since digg won't show me the "submit" button for an edit, i wanted to add the "disclaimer" of "in all the marketing and press materials '-core' means how many cores on a specific chip, when talking about the specific chip'"a dual-core processor does not mean two single core processors, it means ONE processor with dual-cores. just like this article is a single processor with four cores.four cores is available if you want to think of it that way, with dual dual-core, or dual-hyperthreading processors (sorta)
hhcvSep 29, 2006
*tumbleweed*
weaklingSep 29, 2006
Not so impressive these benchmark results. Pure CPU tests are not very relevant imho and when looking to the other results only the 3D rendering result shines. The other tests only show about 10% increase in speed compared to a Core 2 Duo, which I find way too low for a doubling of cores. I do not expect a 100% increase, not at all, that is unreasonable, but a 30% increase at least was expected by me.
theragu40Sep 29, 2006
I was just thinking that. It's like AMD has been gloating all this time about how great it is, and Intel's just been quietly researching its revenge. Suddenly Intel turns around and says "bye the way...we have THIS", and AMD just goes "oh s**t!" and tried to throw something together (dual dual core?).
cavengerSep 29, 2006
"Not so impressive these benchmark results."I don't think the CG renderers and image editors are saying that.
robchartleySep 29, 2006
I think the difference is in core to core communication. Cores on the same die can communicate with shared cache, which is wicked fast. Cores on separate dies even if they are on the same processor package are communicating through a bus of some variety, still pretty but not as.
geminitojanusSep 29, 2006
Uh, you do know the Pentium M was intentionally kept down by low clock speeds to prevent it from competing with Pentium 4/D/EE sales, right? You know that original overclocks of the Pentium M were defeating Athlon 64s, since before the Athlon 64 was even released? The Core chip was just a bit of a tweak on the Pentium M's core, that was actually allowed to have an astronomically high clock, and allowed to compete with the Pentium D. More advanced cache, smaller process, higher clock, improved Micro-op fusion.In summary: Intel's been out to compete with AMD, not destroy it. Think about things this way: what would have happened if Intel used the Pentium M to go up against the Athlon 64, and clocked it past 2.0GHz from the start?
geminitojanusSep 29, 2006
"I think the difference is in core to core communication. Cores on the same die can communicate with shared cache, which is wicked fast. Cores on separate dies even if they are on the same processor package are communicating through a bus of some variety, still pretty but not as."Bzzz. No cigar. Cores on the same piece of silicon still have to talk using a bus, it's just that the bus can be a LOT faster when the wires are a few microns long verses 2-5cm. Two Opteron cores talk to each other via HyperTransport, for example. The Core 2 chips were actually quite advanced for their ability to dynamically remap the L2 cache so that it simulated "communication" between the cores, even if all it's really doing is remapping the L2 to each processor as its needed, but in order for two Core 2 chips to talk to each other, they still require the Front Side Bus.
pantukyOct 30, 2006
The correct version is quadcore single socket.