Can Bulldozer save AMD?
Nothing about Bulldozer looks like a huge gamble. When considered in the context of other very wide SMT designs from IBM and Intel, Bulldozer is actually a conservative, evolutionary step forward from what has gone before.
In the world of processor design, evolution is always a lot better than revolution. It's the radical designs that fail to live up to expectations (e.g., Itanium, Pentium 4, IBM Cell), while more conservative, incremental approaches tend to win out in the end.
That said, as incremental improvements go, adding a whole separate set of four integer units is a pretty large increment. This introduces many changes, and there are a ton of knobs that will need to be dialed in to exactly the right value (cache size, cache associativity, cache latency, instruction buffer size, partition policy, decode bandwidth, etc.). Of course, this is always the case with a brand new design, but that's what's perilous for AMD—these values often get tweaked as the design matures, but Bulldozer won't be mature for some time. AMD needs Bulldozer to deliver immediately, though, so the margin for error is zero.
But even if the first Bulldozer products ship on-time and are fully price/performance and performance/watt competitive with Intel, Bulldozer (and Bobcat, which I'll talk more about in a separate piece) may still not be the home run that AMD needs.
Fighting the last war
AMD always succeeds when it attacks Intel not where the latter is strong, but where it is weak. Historically, AMD's biggest wins have come when the company moved into an obvious hole in Intel's product line. For example, when Intel announced that EPIC and Itanium would be its 64-bit upgrade path, AMD countered with x86-64 and scored a huge victory in the server market. Or, when delays with the QuickPath Interconnect forced Intel to stick with its aging frontside bus architecture for way too long, AMD exploited its superior HyperTransport interconnect to pursue the multisocket server market. When Intel was pushing RAMBUS and, later, the power-hungry FB-DIMM, AMD stuck with cheaper DDR and gained a platform-level performance/watt advantage.
Right now, there are no obvious weak spots in Intel's conventional server platform; indeed, Intel's Xeon line is as strong as it has ever been. (Mobile is a different story, but that's a topic for later.) Insofar as Bulldozer is aimed at the server market, AMD is attacking Intel when and where the larger chipmaker is at its absolute strongest.
