For those who have been following the front page and the submission discussion sub-forum this will come as no surprise, but 0.0 from Thailand has pushed the dual core Global First Place to over 600 points using his Core i7 4600M mobile Haswell processor. Even though this processor is not unlocked by default, a bug in an early microcode allows for ratio overclocking just like on the K-sku processor series.
Spread over a couple of forum topics, 0.0 explains how to “unlock” the factory-locked mobile processors exploiting a bug in an old microcode for the Haswell processors. To make a long story short, the old microcode allows for higher turbo boost bins to be used on any (mobile) processor. On power user forums (like win-raid.com) this method is explained a bit more in detail, although due to the technical nature of the matter there’s no simple guide on how to unlock your processor available. The process of “unlocking” or exploiting involves modifying your bios by inserting the old microcode and then flashing your notebook with this modified BIOS.
In the threads below you can find more information posted by 0.0 and fellow tested ice.cold
The results are pretty amazing. 0.0 pushed the frequency of the Core i7 4600M to 4.8GHz in XTU and 4.9GHz CPU-Z validation on his MSI MS-16GD notebook. Needless to say the CPU does throttle rather quickly inside the notebook casing and this is something you might not want to do to your own notebook at home.
The sweet/awesome part of the story is that this is in fact now the fastest Intel dual core configuration! Since Intel does not have an unlocked dual core with hyper-threading product on the market, the closest you can get with a desktop part is using liquid nitrogen and pushing the base clock frequency to the max. Dancop currently has the fastest desktop dual core with the Core i3 4370 clocked to a mere 4225 MHz with both processor and memory cooled by liquid nitrogen.
On the one hand, it’s hilarious to see a notebook on air beating a desktop on liquid nitrogen. On the other hand, it’s also remarkable how close a 4.2GHz desktop CPU with proper memory configuration can get to a 4.8GHz mobile CPU. It’s hard to tell which is most impressive too as both 0.0 and Dancop have put hours in getting the scores they have.
All in all I am impressed with the research and efforts of 0.0 of exploiting the microcode to push the mobile Haswell to 4.8GHz. It’s refreshing (no pun intended) to see something else than liquid nitrogen at the top. But quite frankly, I am even more curious to find out what extreme power users and overclockers are going to with this information. Can we expect people to run liquid nitrogen on notebook PCBs soon? I, for one, would be really interested to see a 5.5GHz mobile CPU take the 2xCPU XTU Global First Place to new heights!