• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

14th Gen "Raptor Lake Refresh"

Bencher said:
I dont get the meaning of your post. Never argued that at those out of the box clockspeeds the 12900 is efficient. Of course its not, and no cpu is running 5ghz all core. If you care about efficiency you would power limit it just like i do, and suddenly the 12900k is the most efficient cpu on planet earth. So the question is, why do you care about out of the box settings? Do you run your cpus at stock or something?
Click to expand...

You still haven't answered my question, if ADL is so efficient then does is scale really badly given that its pulling twice as much power as a 5950X for similar performance?

Also, record a video of your settings with HWInfo open of an R23 loop 20 minute run.
 
humbug said:
You still haven't answered my question, if ADL is so efficient then does is scale really badly given that its pulling twice as much power as a 5950X for similar performance?

Also, record a video of your settings with HWInfo open of an R23 loop 20 minute run.
Click to expand...
It doesnt draw twice as much. I ready told you i can get 170w for 28k score and 100w for 25k.

Let me put it this way. How much watts does the 5950x draw to run cbr23 at 4.9 ghz all core? Then you'll figure out why the 12900k is inefficient out of the box.

Cbr23 loop is not a stress test, its pointless running that. I can run hours of it with unstable settings. I mean ill do what you asked, im just pointing out that any instability wont show.
 
Bencher said:
It doesnt draw twice as much. I ready told you i can get 170w for 28k score and 100w for 25k.

Let me put it this way. How much watts does the 5950x draw to run cbr23 at 4.9 ghz all core? Then you'll figure out why the 12900k is inefficient out of the box.

Cbr23 loop is not a stress test, its pointless running that. I can run hours of it with unstable settings. I mean ill do what you asked, im just pointing out that any instability wont show.
Click to expand...

Because you can tune it the power consumption is wrong? You can do the same for any CPU, ADL is not unique with this, there are many reasons for it, silicon quality, ambient temperatures, cooling effectiveness, PSU quality, motherboard quality, the quality of the wiring in your house, CPU vendors, be it Intel or AMD have to assume the worst when applying power values to the silicon intended for retail, if you have a good quality montherboard, PSU, ecte.... you can custom tune it to get more efficiency out of it.

The 5950X runs at about 250 watts at 4.7Ghz, which is a much higher clock from stock and with that will also perform that much higher.
 
humbug said:
Because you can tune it the power consumption is wrong? You can do the same for any CPU, ADL is not unique with this, there are many reasons for it, silicon quality, ambient temperatures, cooling effectiveness, PSU quality, motherboard quality, the quality of the wiring in your house, CPU vendors, be it Intel or AMD have to assume the worst when applying power values to the silicon intended for retail, if you have a good quality montherboard, PSU, ecte.... you can custom tune it to get more efficiency out of it.

The 5950X runs at about 250 watts at 4.7Ghz, which is a much higher clock from stock and with that will also perform that much higher.
Click to expand...
I never said the 240w is wrong. I said it makes absolute sense given it runs at insane high clockspeeds. If you care about efficiency then you wont run at 4.9 ghz, so what exactly is the problem? Besides, we are only talking about heavy all core Workloads. Igorslab tested efficiency on other productivity applications and the 12900 absolutely wiped the floor with the 5950x both in performance and efficiency. But i guess you dont care about those, do you?

As youve mentioned, the 5950x also needs a watt galore in those clockspeeds, which is my point all along. Imagine how much it needs for 4.9 ghz, probably way over 300 watts!! Most cpus hit a wall after 4 ghz, thats just a fact.

Sure you can tune any cpu, but no cpu will reach 15k at 35w besides the 12900. Its the most efficient cpu hands down unequivocally
 
Bencher said:
I never said the 240w is wrong. I said it makes absolute sense given it runs at insane high clockspeeds. If you care about efficiency then you wont run at 4.9 ghz, so what exactly is the problem? Besides, we are only talking about heavy all core Workloads. Igorslab tested efficiency on other productivity applications and the 12900 absolutely wiped the floor with the 5950x both in performance and efficiency. But i guess you dont care about those, do you?

As youve mentioned, the 5950x also needs a watt galore in those clockspeeds, which is my point all along. Imagine how much it needs for 4.9 ghz, probably way over 300 watts!! Most cpus hit a wall after 4 ghz, thats just a fact.

Sure you can tune any cpu, but no cpu will reach 15k at 35w besides the 12900. Its the most efficient cpu hands down unequivocally
Click to expand...

Your use of angry hyperbolic language is very telling, this is not a rational debate for you, if you want to know why i'm not taking anything you say as read, that's why.

I'd like to see that slide where IgoresLab; "the 12900 absolutely wiped the floor with the 5950x both in performance and efficiency" And yes i want to see that video of your amazing efficiency 12900K in action.
 
Here you

https://www.igorslab.de/en/intel-ma...ion-einsatz-und-eine-niederlage-fuer-amd-2/9/

And quoting him, the 12900 is 71% more efficient than the 5950x in autocad. The 12900 is the most efficienct cpu

Those are quotes, from the reviewer...

71 freaking percent, so inefficient right? But as ive said numerous times, no matter what anyone says or does is going to change your mind cause facts are meaningless. You just want to go around talking **** about intel and proping amd, so the floor is yours.
 
Bencher said:
Here you

https://www.igorslab.de/en/intel-ma...ion-einsatz-und-eine-niederlage-fuer-amd-2/9/

And quoting him, the 12900 is 71% more efficient than the 5950x in autocad. The 12900 is the most efficienct cpu

Those are quotes, from the reviewer...

71 freaking percent, so inefficient right? But as ive said numerous times, no matter what anyone says or does is going to change your mind cause facts are meaningless. You just want to go around talking **** about intel and proping amd, so the floor is yours.
Click to expand...

I can certainly see the 12900K is pulling around 85 watts vs around 110 watts for the 5950X while also being much faster.

For sure that one use case, is Autocad indicative of everything? Even you know its not, so how do you expect this edge case to change my mind? your hyperbole isn't going to do it either.
 
humbug said:
I can certainly see the 12900K is pulling around 85 watts vs around 110 watts for the 5950X while also being much faster.

For sure that one use case, is Autocad indicative of everything? Even you know its not, so how do you expect this edge case to change my mind? your hyperbole isn't going to do it either.
Click to expand...
The same applies to every single lightly threaded task.. Photoshop premiere etc. But NOBODY is going around in amd threads spamming nonstop about how terrible efficiency the 5950x has, right? Now check intel threads and tell me you dont see people spamming about how terrible efficiency the 12900 has based on cb and blender. Do you see a problem there?
 
Bencher said:
The same applies to every single lightly threaded task.. Photoshop premiere etc. But NOBODY is going around in amd threads spamming nonstop about how terrible efficiency the 5950x has, right? Now check intel threads and tell me you dont see people spamming about how terrible efficiency the 12900 has based on cb and blender. Do you see a problem there?
Click to expand...

Again i can see its a lighter threaded workload, given that the 8 core 5800X scores near the 16 core 5950X, it may be using 10 cores? 110 watts is about the limit for an AM4 CPU, its usually around 120, what Zen 3 does is boost until it either reaches those power limits or the cores have reached their maximum clock limits, usually that's around 5Ghz.

The 12900K sure is using very little power, i don't know anything about the workings of Autocad so i don't know what its doing, and of course no one else, so far as i know uses it for benchmarking, so there is nothing to compare it to, one thing that i did find interesting is that the PL1 125 watt and PL2 241 watt power draw and performance are pretty much the same, so those 8 P-Cores are not locked to a lower power state, they are also not drawing as much power in Autocad as they usually do when they are fully loaded up and stressed, the E-Cores use hardly anything at all.
Again its apparently using 10 cores, there are 8 P-Cores.
 
Seems to me that different architectures are more or less efficient depending on the workload. Seems to be more of the thing recently as Intel and AMD designs have diverged. I'd always advise checking what your workload is and seeing which best suits your use case.
 
Troezar said:
Seems to me that different architectures are more or less efficient depending on the workload. Seems to be more of the thing recently as Intel and AMD designs have diverged. I'd always advise checking what your workload is and seeing which best suits your use case.
Click to expand...

Its a curious oddity that we sometimes find in software, if you look at it overall the 12600K is as near in performance to the 12900K that it make no meaningful difference, so to get anything higher end than that for this is wasted money and energy.

Its usual IgoresLab stuff, pretty much everything he does is a bit wired leaving you uniformed with nothing but questions.
 
humbug said:
Again i can see its a lighter threaded workload, given that the 8 core 5800X scores near the 16 core 5950X, it may be using 10 cores? 110 watts is about the limit for an AM4 CPU, its usually around 120, what Zen 3 does is boost until it either reaches those power limits or the cores have reached their maximum clock limits, usually that's around 5Ghz.

The 12900K sure is using very little power, i don't know anything about the workings of Autocad so i don't know what its doing, and of course no one else, so far as i know uses it for benchmarking, so there is nothing to compare it to, one thing that i did find interesting is that the PL1 125 watt and PL2 241 watt power draw and performance are pretty much the same, so those 8 P-Cores are not locked to a lower power state, they are also not drawing as much power in Autocad as they usually do when they are fully loaded up and stressed, the E-Cores use hardly anything at all.
Again its apparently using 10 cores, there are 8 P-Cores.
Click to expand...
If you look through all the tests, he runs lots and lots of them, it's not just autocad that the 12900k is by far superior. He runs like 20+ benches and in most of them the 12900k is far far ahead of any competition.


If i remember correctly phoronix run a test suite of over 150 applications, and the 12900k of course easily walked away with a win in both performance and efficiency throughout the whole test suite.
 
Troezar said:
Seems to me that different architectures are more or less efficient depending on the workload. Seems to be more of the thing recently as Intel and AMD designs have diverged. I'd always advise checking what your workload is and seeing which best suits your use case.
Click to expand...
Yeah, out of the box the 5950x is way better than the 12900k in all core heavy workloads, while the 12900k is way better in more mixed workloads (gaming / lightly threaded productivity apps etc.). The thing is, you can actually "fix" the 12900k and make it extremely efficient in all core workloads as well, but you cannot do the same for the 5950x in lightly threaded workloads. That's why, in my opinion, alderlake reigns supreme.
 
Bencher said:
If you look through all the tests, he runs lots and lots of them, it's not just autocad that the 12900k is by far superior. He runs like 20+ benches and in most of them the 12900k is far far ahead of any competition.


If i remember correctly phoronix run a test suite of over 150 applications, and the 12900k of course easily walked away with a win in both performance and efficiency throughout the whole test suite.
Click to expand...

Again with the hyperble, he runs about 8 applications. Not 20+ and AMD wins in about half of them.
 
Bencher said:
Yeah, out of the box the 5950x is way better than the 12900k in all core heavy workloads, while the 12900k is way better in more mixed workloads (gaming / lightly threaded productivity apps etc.). The thing is, you can actually "fix" the 12900k and make it extremely efficient in all core workloads as well, but you cannot do the same for the 5950x in lightly threaded workloads. That's why, in my opinion, alderlake reigns supreme.
Click to expand...
At least for my uses I find the hardware is so close these days I buy on platform features and/or cost. I jumped in on first gen X570 so just upgrading the CPU was a no-brainer.
 
humbug said:
Again with the hyperble, he runs about 8 applications. Not 20+ and AMD wins in about half of them.
Click to expand...
And you are doing it again. At this point Ill assume you are just lying. Out of all the tests he ran, Intel won 29 while amd won 9. That's a difference of over 300%....but yeah, for sure AMD won half of them. It says a lot about your affiliation when you think going 29 to 9 is a tie ;)

I mean I've known it beforehand, no amount of facts or benchmarks will change your mind, fanboyism is a strong mental disorder sadly. Even if Intel won 99.999999% of tests, you would find that 1 it doesn't and call it a tie, so I guess it's pointless to even argue at this point. I give up
 
You must log in or register to reply here.
Back
Top Bottom