Bit-tech.net has benchmarked the new Nehalem (Core i7) processors from Intel ahead of their official launch. The hardware site compares the new processors to Intel Core 2 Quad, Core 2 Duo and AMD Phenom X4. From a Mac perspective, the Intel Core 2 Quad (3.2GHz QX9770) most closely resembles the high end Xeon processor currently available in the Mac Pro.
Bit-tech found that the memory benchmarks were dramatically better on the new processors even at the same clock speeds. This improvement reflects the new underlying architecture technologies found in the Core i7. Overall, these improvements did result in substantial real world improvements for many tasks. In encoding video, they found the new processors showed improvements ranging from 18%-22% at the same clock speeds as existing chips. Even more impressive improvements were seen in Cinebench 10 testing with clock-for-clock improvements of 29% at the high end. Gaming results, however, were a bit more mixed with only modest improvements in Crysis and Half Life 2, and performance decreases in Far Cry 2.
Based on these findings, they concluded:
If you're very interested in video encoding, transcoding, image editing or you need to compress files that much faster Core i7 is the de facto choice - even the least expensive Core i7 920 is worth the upgrade.
While this should be a good marker for performance improvements coming to the Mac Pro, it's still not clear when Intel will be introducing the appropriate server-class (Xeon) versions of Nehalem. Apple has traditionally used Intel's Xeon chips for their Mac Pros which are due to be updated.
Update: TheInquirer has organized a roundup of Core i7 reviews and benchmarks. Anandtech's conclusion was similar to Bit-Tech's:
The Core i7's general purpose performance is solid, you're looking at a 5 - 10% increase in general application performance at the same clock speeds as Penryn. Where Nehalem really succeeds however is in anything involving video encoding or 3D rendering, the performance gains there are easily in the 20 - 40% range. Part of the performance boost here is due to Hyper Threading, but the on-die memory controller and architectural tweaks are just as responsible for driving Intel's performance through the roof.
