Intel Core i5-12600K

[md2003]

Was watching some videos on youtube. It seems, that many alder lake cpus, including the smaller of the K series, can reach up to 5.5-5.7GHz single core overclocks, with a daily aio setup. What do you think of that?

 

[Hyderz]

Was watching some videos on youtube. It seems, that many alder lake cpus, including the smaller of the K series, can reach up to 5.5-5.7GHz single core overclocks, with a daily aio setup. What do you think of that?

View attachment 232480

seems impressive, depends on how much power it chews up on a daily use...

 

[JustBenching]

Was watching some videos on youtube. It seems, that many alder lake cpus, including the smaller of the K series, can reach up to 5.5-5.7GHz single core overclocks, with a daily aio setup. What do you think of that?

View attachment 232480

Doesn't even need an AIO at all, temps from reviews are overly exaggerated. I'm running a 12900k on a u12a, I get 70 to 72C on cinebench R20 and around 75-78 on cinebench r23. That's at 204 to 230watts. It can easily hit 5.5ghz+ on single or lightly threaded workloads, at least when it comes to temperatures.

 

[md2003]

Doesn't even need an AIO at all, temps from reviews are overly exaggerated. I'm running a 12900k on a u12a, I get 70 to 72C on cinebench R20 and around 75-78 on cinebench r23. That's at 204 to 230watts. It can easily hit 5.5ghz+ on single or lightly threaded workloads, at least when it comes to temperatures.

It is not so easy to deconstruct many reviews. My personal experience, since you are mentioning yours: the 12900k i ve played with, cooled by an ac freezer 360 get easily up to 95°C in c23 with bios limited to 250 watts (in a 10min stress test). Thus, the scenario examined in each test case might be different by many attributes hence the variance of the results. It's true though, that by unleashing power constrains of this cpu, is not that easy to keep it under normal temperatures.

 

[lexluthermiester]

@15:05
There is no way a Noctua NH-U12A is going to cool a 12900K as well as a 360 AIO.

 

[GURU7OF9]

I would have preferred K versions with no e cores of the 6 and 8 core cpus.
At least have the option to buy k without e cores?
Alas it appears those days are over with the raptor lake supposedly supporting upto 16 e cores!

 

[JustBenching]

@15:05
There is no way a Noctua NH-U12A is going to cool a 12900K as well as a 360 AIO.

I'm exaggerating again, aren't I?

It is not so easy to deconstruct many reviews. My personal experience, since you are mentioning yours: the 12900k i ve played with, cooled by an ac freezer 360 get easily up to 95°C in c23 with bios limited to 250 watts (in a 10min stress test). Thus, the scenario examined in each test case might be different by many attributes hence the variance of the results. It's true though, that by unleashing power constrains of this cpu, is not that easy to keep it under normal temperatures.

Don't know what to tell you man, for me in order to get thermal throttle I need to be clocked at 5.3ghz all core @ 1.385 volts. Then it draws around 280-300watts in cbr23 and I hit ~100C

 

[Vayra86]

Yeah pretty old and outdated program tbh. Wouldn't even run unless run as Admin either and will not populate the hardware info from cpu-z correctly either.

The value of such instruments is pretty much inherent to the fact that they're old and outdated.

It offers a means of comparing CPUs in like-for-like comparisons over long periods of time. And for CPU that matters a lot. You want to know how your current model compares against what you're looking to upgrade to. So you want historical data. 5-7 years is not a strange lifetime for a consumer CPU in a regular system. Some are even older.

For the same and similar reasons you really do want to keep your game tests in a lower res even if the majority stopped playing at it. Its a scientific method of determining relative performance. You benefit from long-term data because it provides a very no-nonsense view on progress.

Applications evolve alongside CPUs for most use cases, so keeping one or two super static things is of great value. The biggest selling point of benchmark comparisons is their trustworthiness, the fact that it is 'what you see is what you get'. Any update to these benchmarks will kill that principle in one way or another, invalidating everything it produced before the update.