Beema A8-6410 faster than Kaveri A10-7300

[lefty2]

Notebook check has reviewed the HP Pavilion 17z, which sports a Beema A8-6410 CPU.
http://www.notebookcheck.net/HP-Pavilion-17z-Notebook-Review.124669.0.html
Performance is better than the older A4 5000 and also power consumption is much lower for light workloads.
Unfortunately, in the review they are comparing it with 35W Intel CPUs, not 15W.
The most curious thing is that if you compare the benchmarks on the CPU page, the A8-6410 is actually faster than the Kaveri A10-7300 (i.e taking an average value)

 

[monstercameron]

It shouldn't be, I'm guessing that kaveri is throttling the cpu, especially in something like cinebench St where it should turbo to greater than 3Ghz.

 

[sm625]

Those are some really bad cpu benchmarks. (Bad as in uninformative, not in terms of product performance.) They dont really say anything about how it is going to perform in real usage scenarios. Even one single sunspider test, which takes all of 10 seconds to load and run, would tell you more than all that cinebench garbage.

 

[Erenhardt]

I use notebookcheck only for CHECKING mobile GPUs specifications. Their performance results are dodgy to say the least. I take performance numbers from product reviews done by proper hardware review sites.

 

[lefty2]

I use notebookcheck only for CHECKING mobile GPUs specifications. Their performance results are dodgy to say the least. I take performance numbers from product reviews done by proper hardware review sites.

Unfortunately, we may never see a review from a "proper" hardware site, because they only do reviews when they are given product samples from the manufacture and that doesn't seem to be happening this time around

 

[jdubs03]

I don't have much of an aversion towards notebookcheck, averaged out over enough samples should give reliable results, but I always use other review sites in conjunction with notebookcheck. The more results to average the better.

A lot of the benchmarks they use are the ones we throw around here, i.e. cinebench, so they have validity. Just not in isolation.