Core i7 Nehalem Benchmarks Preview

Sisoft Sandra
SiSoftware Sandra (the System ANalyser, Diagnostic and Reporting Assistant) is an information & diagnostic utility capable of benchmarking the performance of individual components inside a PC. Each of the benchmarks below were run a total of five times with the highest and lowest scores being discarded and an average being calculated from the remaining three.
Despite initial problems getting Sandra to run on the Nehalem (2008 > 2009 update required) the results certainly filled us with excitement as to what we may see in the real-world benchmarks over the next few pages. Starting with the CPU Arithmetic results, the i7 kicks ass and kills the hostages by more than doubling the Whetstone results obtained by the Q6600.
However, that is absolutely nothing compared to the totally insane 22GB/s memory bandwidth pumped out by the X58 / i7 combo, thanks to the tri-channel DDR3 and on-die memory controller. With the QX9650 and Q6600 pushing out 6.86GB/s and 4.89GB/s respectively, you’d be forgiven for thinking that we were comparing the i7 to some ancient SDRAM-based system from the early ’90’s
As most of us will remember, SuperPI is where the original Core2 chips really captured the hearts (and wallets) of the overclocking community. Sporting sub-20 second results from stock clocked chips, the original Conroe made the ~30second results of an overclocked AMD look slightly pathetic.
Having not seen any SuperPI results from the i7 prior to testing the chip, we had high hopes that it would wade straight in with a ~10 second 1m score. However, coming in at 13.915 seconds, the i7 only manages to bag itself a 2 second advantage over the QX9650. This small advantage does turn into more of a leap when a full 32-million iteration PI is performed, with the i7 finishing around 3 minutes sooner than the QX9650 and a full 8 minutes faster than the Q6600.