as far as i know, there are many differences from P6 architecture to the Core architecture which is 2 generations ahead,
first thing you might want to go through, is the netbusrt microarchitecture, which is the successor to the P6 and you can find info of the changes here:
http://en.wikipedia.org/wiki/Intel_NetBurst#Technology
seems that some over-heating issues raised with netburst, limiting it's clock speed, and excessive pipelining which has actually caused latencies with certain applications,
has mainly led intel to discard it and base the Core architecture foundations on the P6.
you can see the main differences here:
http://hothardware.com/Articles/Int...ors-Chipsets-And-Performance-Analysis/?page=3
and through the earlier page,
basically,
more cache, linked L1, and what is called, "smart cache" where the cores share the same L2 cache pool, and which is also dynamically allocated.
smart memory access with data fetching and prediction algorithms to withdraw data from main memory and pipelining it, prior to it's addressing by the CPU,
shorter instruction pipelining and increased, up to 4 instructions per clock cycle,
they have some fusion process which is combining two operations into a single one,
multi-platform architecture, which is a move toward creating all of intel's cpu's on the same architecture, rather then using different ones for the desktop and server,
yet not absolutely sure about that,
this is more of a strategic move, which makes the Xeons more available to the mainstream or hi-end markets.
you can also go over the Core architecture, wikipedia article, which might give you a closer look at the technology,
this is just a quick answer, which is made quite briefly...
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
Discussion Intel current and future Lakes & Rapids thread
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products
Facebook
Twitter