Intel Core i9-13900K "Raptor Lake" Tested Again, 30% Faster Than Predecessor in Cinebench R23

[JustBenching]

50% more cores than 12900 and it amazes with 30% more performance. Poor result IMO.

7950X3D will be a beast, although for me the 7900X3D is my sweet spot this gen.

33% more threads for 30% more performance. Great result actually

 

[ratirt]

Again, you dont really know what you are talking about. According to intel, pcores are faster at same power than ecores. Therefore they require less power for the same performance. Therefore 16 p cores would be faster than 16 ecores.

Conclusion is, power isnt the reason they dont have a 16p cpu. Its die size.

Power and die space that all I will say you explore more since you have shown you dont understand things the way they are but rather the way you want them to be. Common problem for 70% of teenagers nowadays when they literally want to create their own reality instead of accepting the reality they are in. Good for you though

50% more cores than 12900 and it amazes with 30% more performance. Poor result IMO.

7950X3D will be a beast, although for me the 7900X3D is my sweet spot this gen.

This 50% more cores are literally ecores for MT threaded workloads to boots performance. There is literally nothing to celebrate about to be fair. Half cores added to keep up with the competition and show some improvement over the previous gen so that Intel does not get scolded for the stagnation.
I'm not sure AMD will release the entire lineup with 3D vcache. For the 7950x I doesn't make a lot of sense unless the 3dvcache will not bring the clocks down like we have seen with 5800x3d

 

[Gica]

The only problem i can see with the 13900k is going to be the stupid power use. I guess if you have the cash to buy the top dog, you can cool it, or know how to tweak it to keep it in check. Same as AMD users under volting and keeping 90% of the performance, i guess the same can be done with this

The same and the same plate, with consumption and cooling. AMD will release a processor that consumes less (max. 230W versus max. 290W), but it will have much bigger problems with cooling because the dissipation surface is smaller. The cooler for AMD must be much more efficient. Ironic, right?
We desperately cling to tests where everything is squeezed out of the processor (prime95 and others) without taking into account that only how many software can come close to that consumption, but what do we do with the others? We watch a movie, surf the net, read and comment on the forums, use office or CAD, etc., etc. ... those E cores clearly outperform AMD at low consumption. As far as gaming is concerned, look on youtube for 12900K versus 5950X, because you mentioned the consumption for this flagship as well (oooooh 241W LOL).

 

[Jimmy_]

oooooo... 380Watts of Power draw @100C

 

[usiname]

The same and the same plate, with consumption and cooling. AMD will release a processor that consumes less (max. 230W versus max. 290W), but it will have much bigger problems with cooling because the dissipation surface is smaller. The cooler for AMD must be much more efficient. Ironic, right?
We desperately cling to tests where everything is squeezed out of the processor (prime95 and others) without taking into account that only how many software can come close to that consumption, but what do we do with the others? We watch a movie, surf the net, read and comment on the forums, use office or CAD, etc., etc. ... those E cores clearly outperform AMD at low consumption. As far as gaming is concerned, look on youtube for 12900K versus 5950X, because you mentioned the consumption for this flagship as well (oooooh 241W LOL).

Offtopic: I liked a comment on a car forum: You want a sports car, but you complain about consumption? Dude, you're an idiot!

You explain and show how CPUs with 16 cores are effective in multimedia and browsing and only idiot should complain for the power consumption under full load? Apparently you are not aware that this 16 cores are not for 100% of the time to be spend on gaming or browsing, this 16 cores are for doing work, and if you don't reach 100% of the potential of this cpu then you are fine with lower tier, because you get 95% of the performance for half money.

 

[JustBenching]

You explain and show how CPUs with 16 cores are effective in multimedia and browsing and only idiot should complain for the power consumption under full load? Apparently you are not aware that this 16 cores are not for 100% of the time to be spend on gaming or browsing, this 16 cores are for doing work, and if you don't reach 100% of the potential of this cpu then you are fine with lower tier, because you get 95% of the performance for half money. I guess the idiot is someone else

But in the same context, only a fool will run a cpu with a 240w power limit at 100% workloads 24/7

There is a button called power limit, use it if you intend to do heavy mt workload and stop complaining. Problem solved

 

[ratirt]

But in the same context, only a fool will run a cpu with a 240w power limit at 100% workloads 24/7

That's where the performance is and how it has been advertised so people would want to have that performance for which they have paid.

There is a button called power limit, use it if you intend to do heavy mt workload and stop complaining. Problem solved

If you need to push that button, that CPU is not for you, since you are after efficiency not performance.

 

[JustBenching]

That's where the performance is and how it has been advertised so people would want to have that performance for which they have paid.

If you need to push that button, that CPU is not for you, since you are after efficiency not performance.

So you HAVE to activate pbo on the 5950x then right? Cause thats where the performance that you paid for is. In which case the cpu draws up to 200+ watts as well.

 

[ratirt]

So you HAVE to activate pbo on the 5950x then right? Cause thats where the performance that you paid for is. In which case the cpu draws up to 200+ watts as well.

What has a PBO to do with it? It is an overclocking option for Ryzen not a default setting or stock setting. It is a controlled OC given by AMD as an option and it is out of spec. You run it according to specs or guidance given by the manufacturer and that consists of the PL1 and PL2 for alder lake. Anything aside it is running out of spec. If you change those manually, you run it out off spec and you can't expect same performance or performance Intel advertises the product but sometimes it may have a positive effect. Lowering voltage is not a given and it will vary from CPU to CPU. It is not something standard for each one so you can't rely on that part to be same for every single CPU. Which means, again, you are running the CPU out of spec which means it may not run as advertised.

 

[JustBenching]

What has a PBO to do with it? It is an overclocking option for Ryzen not a default setting or stock setting. It is a controlled OC given by AMD as an option and it is out of spec. You run it according to specs or guidance given by the manufacturer and that consists of the PL1 and PL2 for alder lake. Anything aside it is running out of spec. If you change those manually, you run it out off spec and you can't expect same performance or performance Intel advertises the product but sometimes it may have a positive effect. Lowering voltage is not a given and it will vary from CPU to CPU. It is not something standard for each one so you can't rely on that part to be same for every single CPU. Which means, again, you are running the CPU out of spec which means it may not run as advertised.

Who cares if it runs as advertised? If you care about efficiency you dont want it to run as advertised, you want it to run more efficiently than advertised!

Im not talking about voltages here, just power limit. You lose 15% performance but drop the consumption to half. So if you care about efficiency, why wouldn't you? You are not making sense my man.

The same applies for zen 3 and zen 4. Say i buy a 7950x and run it at 150 ppt instead of the stock 230. Whats the problem with that?

 

[ratirt]

Who cares if it runs as advertised? If you care about efficiency you dont want it to run as advertised, you want it to run more efficiently than advertised!

Im not talking about voltages here, just power limit. You lose 15% performance but drop the consumption to half. So if you care about efficiency, why wouldn't you? You are not making sense my man.

The same applies for zen 3 and zen 4. Say i buy a 7950x and run it at 150 ppt instead of the stock 230. Whats the problem with that?

I don't make sense? OK bro
15% off the 12900K's general performance due to power limit is less performance than what 5950x can do. If you drop the power to half that is around what 5950x is using.
Great choice with the CPU and your limitation.

 

[JustBenching]

I don't make sense? OK bro
15% off the 12900K's general performance due to power limit is less performance than what 5950x can do. If you drop the power to half that is around what 5950x is using.
Great choice with the CPU and your limitation.

Yes, it is less performance in heavy mt scenarios like rendering. But there are a bunch of other scenarios where the 12900k is way faster even at 125w. Autocad premiere photoshop matlab etc. If all you do is 24/7 rendering of course the 5950x is a better option, although not by much. In more mixed workloads the 12900k is way faster. Especially for live work there is no contest, everything is way more smooth in live preview for example on the 12900k. Basically heavy workloads that require user input the 12900k is by far the better option.

 

[Valantar]

Doesnt make a difference really, this is a desktop cpu. The low threshold is what, just 2-3 watts per core. You ain't going to run a cpu at 35w

No, but when you're talking per-core efficiency, that threshold is important when comparing architectures. It's one of the major advantages of E cores - they scale much, much lower than P cores while maintaining performance.

Depending on if the task is int or fp, but generally speaking anything above 2-3w per core pcores become more efficient than ecores. So yeah, we kinda know that

I'd love to see some data on that. Has anyone tested this in depth? I've seen Anandtech's E core testing (which only tested against stock P cores) and TPU's E core testing, which show a pretty complex image overall. AT saw their E cores peak at just 48W combined, while TPU sadly still does useless full system power testing, which is especially sad as they did iso-frequency testing. (Though that does tell us some interesting things even if the methodology is fundamentally flawed - like the P cores only configuration idling at measurably higher power levels. And if we go off of TPU's numbers - flawed as they are - it seems the P cores at 3.9GHz are just marginally more efficient than the E cores at the same frequency, with 43% more power for 52% more performance (using load power minus idle power as the power gauge). Of course that power number is just from a single nT test, so mapping it onto the whole of the test suite is rather problematic in and of itself, on top of the massive variability introduced by measuring full system power rather than CPU power only.

But that is not the point. The point is the main or even only reason we dont see 16p cores is the die size. Wattage is really not a problem. I dont see how it could be, since we already know that pcores are more efficient than e cores in everything unless you drop down to a couple of watts per core, which you wont in a desktop.

But that's precisely the thing: you can't just pin things on one limitation - it's always a balance. I sincerely doubt we'd see 16 P cores even if area wasn't an issue, simply because the clock drop-off in high threaded workloads would be too large. If P cores at 3.9GHz are just barely more efficient than E cores at the same clock speed, with the P cores already being notably downclocked from their peak while the E cores being essentially at theirs, it seems logical to assume that the E cores would catch up or pass the P cores in efficiency at slightly lower clocks again.

There is a button called power limit

There is? Where? Pics plz.

 

[Tomorrow]

Does not look that impressive compared to 7950X actually. 7950X starts from 170W and goes to 230W. While 13900K starts from 241W and goes to the moon.
38k while OC'ed in R23 is in the same ballpark as 7950X.

My guess is that both of these will be very comptetitive when it comes to performance but 13900K will consume more power both at stock and OC'ed.
Atleast 7950X can be undervolted to reduce temps and improve MT performance. Well see about 13900K i guess.

 

[JustBenching]

No, but when you're talking per-core efficiency, that threshold is important when comparing architectures. It's one of the major advantages of E cores - they scale much, much lower than P cores while maintaining performance.

I'd love to see some data on that. Has anyone tested this in depth? I've seen Anandtech's E core testing (which only tested against stock P cores) and TPU's E core testing, which show a pretty complex image overall. AT saw their E cores peak at just 48W combined, while TPU sadly still does useless full system power testing, which is especially sad as they did iso-frequency testing. (Though that does tell us some interesting things even if the methodology is fundamentally flawed - like the P cores only configuration idling at measurably higher power levels. And if we go off of TPU's numbers - flawed as they are - it seems the P cores at 3.9GHz are just marginally more efficient than the E cores at the same frequency, with 43% more power for 52% more performance (using load power minus idle power as the power gauge). Of course that power number is just from a single nT test, so mapping it onto the whole of the test suite is rather problematic in and of itself, on top of the massive variability introduced by measuring full system power rather than CPU power only.

But that's precisely the thing: you can't just pin things on one limitation - it's always a balance. I sincerely doubt we'd see 16 P cores even if area wasn't an issue, simply because the clock drop-off in high threaded workloads would be too large. If P cores at 3.9GHz are just barely more efficient than E cores at the same clock speed, with the P cores already being notably downclocked from their peak while the E cores being essentially at theirs, it seems logical to assume that the E cores would catch up or pass the P cores in efficiency at slightly lower clocks again.

There is? Where? Pics plz.

There is a very in depth test done, dont remember where from. Ill link it when i find it, but generally speaking p cores >>> ecores in any wattage you would run a desktop cpu.

There is also a graph from intel that shows pcores being more efficient at every single wattage point but I dont think that's true for all applications.

The problem with these kinds of testing, like the one AT does, is that the moment you disable ecores, the cache clocks to 4.7ghz and vcore follows the voltage curve of the cache which is much higher than the cores. The only one that can really settle this without a shadow of a doubt is intel.

First thing youll be asked to do before you get into the bios is to set your power limits. If you dont like the bios there is xtu as well that has a button

Dont tell me again about people not knowing how to get into the bios, cause then that means that the upgradability that am4 offers is useless, since it requires a bios upgrade, something that is about 90 times more complicated than setting a power limit

 

[Valantar]

There is also a graph from intel that shows pcores being more efficient at every single wattage point but I dont think that's true for all applications.

I think you might be confusing Intel's E core vs. Skylake comparison graphs with this? I definitely can't find any slides doing a perf/power curve comparison for P and E cores in Intel's 2021 Archtecture Day slides or any other ADL coverage.

The problem with these kinds of testing, like the one AT does, is that the moment you disable ecores, the cache clocks to 4.7ghz and vcore follows the voltage curve of the cache which is much higher than the cores. The only one that can really settle this without a shadow of a doubt is intel.

... but wouldn't an all P core CPU also do exactly this?

First thing youll be asked to do before you get into the bios is to set your power limits. If you dont like the bios there is xtu as well that has a button

Really? I've never seen that, though I haven't used any newer generation Intel boards. I've seen vague, quasi-nonsensical "performance profile" or "cooling setup" selectors, though these are fundamentally non-transparent in what they do and which settings they change (and arguably serve to confuse novice users more than simplify things IMO). AFAIK these are also entirely dependent on the motherboard manufacturer and have nothing to do with Intel.

Dont tell me again about people not knowing how to get into the bios, cause then that means that the upgradability that am4 offers is useless, since it requires a bios upgrade, something that is about 90 times more complicated than setting a power limit

Have I talked about AM4 upgradeability here? I mean, I know you're rather fond of creating straw men to argue against, but that really doesn't help you here. Also, a BIOS update these days is typically a simple downloadable executable. Is performing such an update trivial? No. Is it more accessible to users than entering the BIOS? Yes.

As for setting a power limit being trivial: you're going to have to provide some details on this. Do the motherboards give you presets, or do you have to enter PL1/PL2/Tau limits manually? If there are presets, are these universal, or motherboard brand dependent? AMD has what I would consider a simple solution: "Eco mode" profiles at 65/45W TDP levels (with matching PPT, EDC and TDC settings baked in) - but this is also hidden in BIOS submenus that no novice user would ever find, nor is understanding what this option does especially easy. It's at least a simple toggle, and it's universal, but it's still not accessible.

Heck, you can set power limits in XTU, but to do so you need to know enough to install XTU, and set safe and sensible settings there manually. Point being: this is non-trivial. Period.

 

[JustBenching]

I think you might be confusing Intel's E core vs. Skylake comparison graphs with this? I definitely can't find any slides doing a perf/power curve comparison for P and E cores in Intel's 2021 Archtecture Day slides or any other ADL coverage.

No, there is an actual graph with p and e cores. Not very scientific, so I'm not sure how accurate it is

... but wouldn't an all P core CPU also do exactly this?

No, not necessarily. Its up to intels stock settings. I mean im not even sure its intels settings or mobo manafacturers doing that kind of crap.

Really? I've never seen that, though I haven't used any newer generation Intel boards. I've seen vague, quasi-nonsensical "performance profile" or "cooling setup" selectors, though these are fundamentally non-transparent in what they do and which settings they change (and arguably serve to confuse novice users more than simplify things IMO). AFAIK these are also entirely dependent on the motherboard manufacturer and have nothing to do with Intel.

All boards ive used recently (z490 b560 and a bunch of z690s), dont even allow you to get into the bios without choosing a power limit. Basically they ask what cooler you have and a description of the wattage limits with that option.

Have I talked about AM4 upgradeability here? I mean, I know you're rather fond of creating straw men to argue against, but that really doesn't help you here. Also, a BIOS update these days is typically a simple downloadable executable. Is performing such an update trivial? No. Is it more accessible to users than entering the BIOS? Yes.

As for setting a power limit being trivial: you're going to have to provide some details on this. Do the motherboards give you presets, or do you have to enter PL1/PL2/Tau limits manually? If there are presets, are these universal, or motherboard brand dependent? AMD has what I would consider a simple solution: "Eco mode" profiles at 65/45W TDP levels (with matching PPT, EDC and TDC settings baked in) - but this is also hidden in BIOS submenus that no novice user would ever find, nor is understanding what this option does especially easy. It's at least a simple toggle, and it's universal, but it's still not accessible.

Heck, you can set power limits in XTU, but to do so you need to know enough to install XTU, and set safe and sensible settings there manually. Point being: this is non-trivial. Period.

No you havent but other people have. Even today a bios update is harder than setting a power limit. But regardless, the point is i assume you consider the am4 upgradability a very important part of the am4 platform, but on the other hand you consider power limit a niche use case for the intel platform, which to me doesnt make sense

 

[Gica]

You explain and show how CPUs with 16 cores are effective in multimedia and browsing and only idiot should complain for the power consumption under full load? Apparently you are not aware that this 16 cores are not for 100% of the time to be spend on gaming or browsing, this 16 cores are for doing work, and if you don't reach 100% of the potential of this cpu then you are fine with lower tier, because you get 95% of the performance for half money.

Video editing. Testers are only used for encoding, which is incorrect because video editing involves several steps. That is why it is correct to compare the real consumption through a benchmark of all stages and not only during encoding. You can work days, weeks and even months until encoding.

Adobe Premiere Pro: 12th Gen Intel Core vs AMD Ryzen 5000 Series

Intel has launched their new 12th Gen Intel Core desktop processors (code-named "Alder Lake") with support for DDR5, PCIe 5.0, as well as a completely new hybrid architecture featuring a mix of Performance and Efficient-cores. This is a lot of new technologies in one product, so we are excited...

www.pugetsystems.com

For home users, idle, gaming, multimedia, online browsing and many tasks that do not demand 100% from the processor are the order of the day, but everyone chooses their processor according to their requirements or ... whim. Some want the fastest gaming processor, others want it just because it has many cores, it's their choice. The idea is that they will not reach the maximum consumption of this processor unless they run Prime95 and other torture tools and not in the daily use of the programs.

 

[Valantar]

No, there is an actual graph with p and e cores. Not very scientific, so I'm not sure how accurate it is

Well, if you can find it, feel free to link it. It's not in any of Intel's ADL launch or architecture day slide decks at least.

No, not necessarily. Its up to intels stock settings. I mean im not even sure its intels settings or mobo manafacturers doing that kind of crap.

But if cache clocks are bound to E core clocks currently, that means that cache clocks would likely be bound to P cores if those were the only ones around. Not necessarily a 1:1 match, but if this limitation wasn't directly due to the presence of E cores it wouldn't make sense for the CPU to change its cache clocks when these cores are disabled. Which renders this argument pretty much moot. Yes, Intel can set whatever settings they want, but it seems clear that they would clock their caches higher if it wasn't for the E cores.

All boards ive used recently (z490 b560 and a bunch of z690s), dont even allow you to get into the bios without choosing a power limit. Basically they ask what cooler you have and a description of the wattage limits with that option.

Yes, as I said, they provide a non-transparent set of settings that affect this with bare-bones explanations of what is done or how this affects anything. And, crucially, this is not an Intel thing, it's a motherboard maker thing - in part a response to criticism of previously just enabling MCE by default and causing most run-of-the-mill builds to reach throttling territory for no reason.

No you havent but other people have.

... and how exactly is that is a reason to bring it up when discussing things with me?

Even today a bios update is harder than setting a power limit.

Again: no. A BIOS update is by no means trivial, but setting a power limit, one that's actually a conscious choice of how to configure your PC, requires a lot more knowledge and understanding than applying an executable update from Windows.

But regardless, the point is i assume you consider the am4 upgradability a very important part of the am4 platform, but on the other hand you consider power limit a niche use case for the intel platform, which to me doesnt make sense

This is some excellent, grade-A whataboutism. I mean, you're using arguments other people are making, about mostly unrelated things, projecting them onto me, and then drawing an extremely tenuous line between this argument that I haven't made and what you're trying to argue. You see that what you're saying here is essentially "Hey, other things with using a computer are complicated too!", right? That isn't an argument for or against anything. It's a vague, broad, directionless statement, that in no way affects the truth or relevance of the fact that manually setting a lower power limit for a CPU is a complicated thing that requires quite a bit of knowledge before it can even be considered, let alone done by a user. Just as with the simple fact that most people building their own PCs don't even enable XMP, most people building or buying desktop PCs have no idea that there is such a thing as a CPU power limit, that it can be configured, that this can be done without sacrificing a ton of performance, nor do they know how this might be done if they were to gain this knowledge. That other things are also complicated is not an argument against this in any way, shape or form.

 

[ratirt]

Just as with the simple fact that most people building their own PCs don't even enable XMP, most people building or buying desktop PCs have no idea that there is such a thing as a CPU power limit, that it can be configured, that this can be done without sacrificing a ton of performance, nor do they know how this might be done if they were to gain this knowledge. That other things are also complicated is not an argument against this in any way, shape or form.

I think you have described 90% of users. Not to mention, Bios update in fact is relatively easy. You can see it clearly with people updating their GPU Bios since they want to OC something with it instead manual OC on current Bios. Most problems people have with their computers starts with "upgrade bios or what bios you have". OC or undervolt etc. requires more knowledge and most people wanna go for the easy thing which mostly ends up in bricks (gpu related) but that kinda proves people do know about Bios updates for mobos and are doing it but rarely know, how to OC or Tweak power/voltage or any other settings.

 

[JustBenching]

I think you have described 90% of users. Not to mention, Bios update in fact is relatively easy. You can see it clearly with people updating their GPU Bios since they want to OC something with it instead manual OC on current Bios. Most problems people have with their computers starts with "upgrade bios or what bios you have". OC or undervolt etc. requires more knowledge and most people wanna go for the easy thing which mostly ends up in bricks (gpu related) but that kinda proves people do know about Bios updates for mobos and are doing it but rarely know, how to OC or Tweak power/voltage or any other settings.

Are you saying that power limiting is harder than bios updates? Im outta here, bye

 

[Valantar]

Are you saying that power limiting is harder than bios updates? Im outta here, bye

Lol, if you're actually arguing that having the knowledge to understand the concept of adjusting a CPU power limit, and having the skill to do so, is less complex than knowing of, downloading, and running an executable BIOS update, then bye!

I think you have described 90% of users. Not to mention, Bios update in fact is relatively easy. You can see it clearly with people updating their GPU Bios since they want to OC something with it instead manual OC on current Bios. Most problems people have with their computers starts with "upgrade bios or what bios you have". OC or undervolt etc. requires more knowledge and most people wanna go for the easy thing which mostly ends up in bricks (gpu related) but that kinda proves people do know about Bios updates for mobos and are doing it but rarely know, how to OC or Tweak power/voltage or any other settings.

Idk, IMO GPU BIOS updates are an example of an in-between phenomenon, more like people being in the "know enough to be dangerous" zone. Definitely requires skill and knowledge above applying a motherboard BIOS update, but on the other hand people doing this are also wildly misinformed, so ... yeah. There's plain ignorance, and then there's ignorance that thinks that it's actually smart.

Though I don't really see this as particularly relevant to the topic, beyond the simple facts that manually adjusting a power limit is non-trivial, and that understanding that CPUs can be tuned through power limiting without losing a lot of performance is anything but common knowledge among PC users. The ability to set lower power limits doesn't make high stock power limits any less problematic - but the real question is how much power is actually drawn in real-world use cases.

 

[ratirt]

Are you saying that power limiting is harder than bios updates? Im outta here, bye

That is exactly what I'm saying. Take care.

Idk, IMO GPU BIOS updates are an example of an in-between phenomenon, more like people being in the "know enough to be dangerous" zone. Definitely requires skill and knowledge above applying a motherboard BIOS update, but on the other hand people doing this are also wildly misinformed, so ... yeah. There's plain ignorance, and then there's ignorance that thinks that it's actually smart.

Though I don't really see this as particularly relevant to the topic, beyond the simple facts that manually adjusting a power limit is non-trivial, and that understanding that CPUs can be tuned through power limiting without losing a lot of performance is anything but common knowledge among PC users. The ability to set lower power limits doesn't make high stock power limits any less problematic - but the real question is how much power is actually drawn in real-world use cases.

What I'm trying to say is, people are more keen to update bios than fiddle with settings or configuration of a motherboard etc. since it is easier and requires less time. Bios update requires looking for your card model or motherboard model to update and downloading it. The power limits and stuff is harder nut to crack and requires more knowledge. That is why it is quite a problematic subject and not as easy as some people claim it to be. Which is what has been in discussion here for quite a while and some people bringing arguments to the table in my eyes are either detached from reality or are playing dumb in a way you can't have a meaningful conversation with them even though you try.

 

[mahirzukic2]

I think you have described 90% of users. Not to mention, Bios update in fact is relatively easy. You can see it clearly with people updating their GPU Bios since they want to OC something with it instead manual OC on current Bios. Most problems people have with their computers starts with "upgrade bios or what bios you have". OC or undervolt etc. requires more knowledge and most people wanna go for the easy thing which mostly ends up in bricks (gpu related) but that kinda proves people do know about Bios updates for mobos and are doing it but rarely know, how to OC or Tweak power/voltage or any other settings.

Several times in the span of 3 years that I have worked for my previous employer, I have had a firmware update (BIOS) as a part of other ubuntu updates being run at some point.
It does warn to keep the laptop powered by cable while the operation is underway just to be safe, even though the battery was mostly full.

My point being, that BIOS updates are so easily done, that they were in my case done automatically for a Dell ubuntu mobile workstation, whichever model it was. I'm pretty sure it was like that for the whole family of those products.
Many motherboards let you do a bios upgrade without even having a processor enabled, so it's pretty easy to do.

 

[JustBenching]

Lol, if you're actually arguing that having the knowledge to understand the concept of adjusting a CPU power limit, and having the skill to do so, is less complex than knowing of, downloading, and running an executable BIOS update, then bye!

Idk, IMO GPU BIOS updates are an example of an in-between phenomenon, more like people being in the "know enough to be dangerous" zone. Definitely requires skill and knowledge above applying a motherboard BIOS update, but on the other hand people doing this are also wildly misinformed, so ... yeah. There's plain ignorance, and then there's ignorance that thinks that it's actually smart.

Though I don't really see this as particularly relevant to the topic, beyond the simple facts that manually adjusting a power limit is non-trivial, and that understanding that CPUs can be tuned through power limiting without losing a lot of performance is anything but common knowledge among PC users. The ability to set lower power limits doesn't make high stock power limits any less problematic - but the real question is how much power is actually drawn in real-world use cases.

It requires intense skill to type 125 in a box, i get it.