4900MQ turbo boost broken?

Based on everything we've seen so far, I'm starting to wonder if 4900MQ and 4930MX are actually in the same bin. If so, then there really is little point in getting getting the 4930MX, unless you're a hardcore enthusiast., and don't mind paying $500 for unlocked multipliers and a 200MHz bump in base clock.

 

Probably not the exact same bin but the two have always been close (Extreme and one down).

 

In the same bin, no. 57W TDP vs 47W TDP should mean better thermal characteristics with the MX.

 

Intel will use their own binning procedures looking at frequency vs power.

 

Not much difference eh? Even your 4700MQ has 160+ GFlops in Linpack. The highest I got was 177 GFlops in the 20000 sample size test. That's about the same as what a desktop 4770K would get.

Seems like the higher end CPU is binned to reach higher clock speed, and I am not entirely sure that it's more efficient than the lower end CPUs at all clock speed.

It's just sad that the 3920XM/3940XM can reach 4.3+ GHz stable, and my Haswell 4930MX can't. Because Ivybridge & Sandybridge don't support AVX2, so nothing can really push that much heat. It's like having a latent but unusable and unpractical performance with the Haswell.

With Broadwell, if Intel scale down the IVR to 14nm, I'd expect it to have a lower efficiency, which means more heat. I am not sure if they can pull any magic trick there with the analog parts.

 

The analogue parts are usually larger structures anyway.

per unit of performance haswell is more efficient, it's just hotter as a denser core.

 

Yes, but I can only run for a few seconds like that before I have thermal problems.

You also have all those extra multipliers for other tasks other than Linpack. Would have liked to have tried the 1.25 strap for BCLK but this GE BIOS has all those nice things removed. :/

 

The strap has not been working in any machines as I have seen.