Originally Posted by Insan1tyOne
Haha "only" 14c / 28t with 35MB of L3 Cache you say? According to Intel's ARK this is quite a nice CPU
, however, the speeds of it are just far too low for it to even be considered for gaming or anything like that. Although for pretty much any other application $235 for a CPU of that caliber
is one of the best deals I have ever seen. Seriously, someone please go buy a bunch of these right now!
Now this is an interesting model indeed... This particular SKU is not even listed on Intel's ARK. That leads me to believe that it was a custom request by a company and was only produced for OEM purposes. What I find the most interesting is that its boost range is so large
at ~1.5 Ghz. For 18c / 36t with 45mb of L3 Cache $399 is not a bad price at all
. Although it would be terrible for gaming. Unless it could always run all cores at 3.5 Ghz, and even then, 3.5 Ghz is sadly still quite low for gaming purposes.
I have been throwing around an idea in my head when it comes to these V1 / V2 / V3 / V4 Xeons and "overclocking". I have no idea how / if it could even be done, but what if instead of finding a way to unlock the "hardware" locks on the multiplier, etc. Someone just found a way to unlock the "software" limitations on these Xeons? And by software limitations, I mean the Intel TurboBoost technology.
Each Xeon processor has a "Max Turbo Frequency" which is defined as "the maximum single core frequency at which the processor is capable of operating using Intel® Turbo Boost Technology." I also believe that there is a deeper feature of Intel TurboBoost which allows for all
cores to simultaneously boost up to a certain (slightly lower than max) speed under the correct (very high system load) circumstances.
So now, what if someone were to create a program that just told Intel TurboBoost (because it is a piece of software that communicates / functions within the OS if I am not mistaken) to always
processor cores at their maximum TurboBoost frequency regardless
of system load?
In my opinion, the above solution seems like it would be several measures of magnitude easier to accomplish than trying to modify the pins / contacts of the CPU itself to fake less QPI connections or "disable" the CPUs overclocking security features. In my opinion, disabling pins will not result in any processor features being restored, as the only way to bring back these lost features would be to re-add the pins that Intel removed to block off these features in the first place. That is just the musings of someone who is neither an Hardware or Software Engineer though.