Sorry if this has been discussed elsewhere.
I suspect I'm not alone in taking a lot of interest in the new quad-core Sandy Bridge mobile processors, which promise desktop performance in a laptop. I notice that most of the CPU tests shown so far for the i7-2630QM and i7-2720QM have been classified as overclocked. These processors are not, in fact, overclockable in the sense that the turbo boost margins are factory-set. The processor is therefore working as its manufacturer intended. This makes me wonder whether turbo boost within factory limits should count as overclocking.
I notice that some of the baselines show incorrect clock speeds (either none, or, in the case of one of the 2720QMs, 24GHz and the wrong number of cores). Referring to the 2720QM, I notice also that there are some examples recorded as running at base clock rate but still yielding similar performance to those recorded as overclocked. Could it be that the clock rate is measured only once, and it is a matter of luck whether the processor is running in turbo boost mode or not? How do you measure the clock speed during the performance test? How do anomalous clock speeds and core counts occur?
I suspect I'm not alone in taking a lot of interest in the new quad-core Sandy Bridge mobile processors, which promise desktop performance in a laptop. I notice that most of the CPU tests shown so far for the i7-2630QM and i7-2720QM have been classified as overclocked. These processors are not, in fact, overclockable in the sense that the turbo boost margins are factory-set. The processor is therefore working as its manufacturer intended. This makes me wonder whether turbo boost within factory limits should count as overclocking.
I notice that some of the baselines show incorrect clock speeds (either none, or, in the case of one of the 2720QMs, 24GHz and the wrong number of cores). Referring to the 2720QM, I notice also that there are some examples recorded as running at base clock rate but still yielding similar performance to those recorded as overclocked. Could it be that the clock rate is measured only once, and it is a matter of luck whether the processor is running in turbo boost mode or not? How do you measure the clock speed during the performance test? How do anomalous clock speeds and core counts occur?
Comment