Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

There has to be a memory bandwidth limit where the number of cores is rendered moot for quad-core arm7 to be slower than dual-core arm7.

Either that or the test is somehow being fooled, I noticed that benchmark code is updated frequently.



Tegra is known for low memory bandwidth.


Googling yields a 6.4 GB/s figure for the single channel LPDDR3 supported by Tegra 3. It doesn't sound like the most likely bottleneck for a CPU benchmark given that those A9 cores plod along at about 1/10 the perf of a desktop chip.

What sort of bandwidth do other mobile chips provide?

edit: Looks more like 1/3 than 1/10, looking at the tests in geekbenchscores. I had the 1/10 figure from looking at Sunspider results (http://forum.beyond3d.com/showpost.php?s=502c8933367f315a17f...).


That's still less bandwidth than any other recent SoC and keep in mind that the GPU is also sharing that bandwidth.


While Geekbench is updated frequently, the actual benchmark code is virtually unchanged since 2.0.0 was released.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: