Soldato
According to Intel's website on the 12900k, stock is 125/241 56 TAU. Clearly he is not using that, he is using unlimited turbo.
-
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.
Raptor Lake Leaks + Intel 4 developments
- Thread starter g67575
- Start date
Caporegime
30,000 @ 65 Watts vs 38,000 @ 240 watts
For 27% more performance its using 3.5X more power, its well outside of its efficiency curve at stock, they have pushed it to squeeze every last point out of it, Intel have been doing that for years and if that's how the game is going to be played then so be it.
I can't blame them, AMD could lock it to 125 watts and in doing that claim to be responsible, no one cares.... along come Intel with a 300 watts chip that beats it and all the Youtube headlines all read "Intel CRUSHES AMD"
If you can't beat them, join them, meh...
Last edited:
Ahh so everybody's testing is flawed. You see I am more on the side of credit where credit is due, regardless of the brand loyalty you have to look objectively at numbers across a bunch of trusted reviewers. If you do that you can see that places like tech power up put the efficiency roughly the same as gamers nexus and also test it with unlimited turbo as do many others. Either way it's not winning any efficiency battles. Multithreaded it got absolutely dominated in efficiency and I mean dominated by 5950x.
I do think 13900k is shaping up better and should skew efficiency more towards the chip in multi threaded applications due to the addition of more of those gracemont efficiency cores in relation to the number of P cores. 13900k will be more efficient than 12900k (I think) for that reason. Again the 12900k is a bliteringly fast/good chip but efficient it is not.
I do think 13900k is shaping up better and should skew efficiency more towards the chip in multi threaded applications due to the addition of more of those gracemont efficiency cores in relation to the number of P cores. 13900k will be more efficient than 12900k (I think) for that reason. Again the 12900k is a bliteringly fast/good chip but efficient it is not.
Last edited:
Soldato
Obviously 13900k will be more efficient than the 12900k, not because of extra E cores, but because of extra cores. Even if they added P cores the efficiency would go up.
Regardless, Intel's spec sheet is on their webpage, the 12900k has a 125/241 power limit, so after 56 seconds that the TAU expires, it is more efficient than the 7950x, at least in CBR23. That's just a fact. The only thing you can argue is whether 125/241 is stock settings or not, and im all open to talking about that (since it will apply for the 13900k), but if it is, then it is an absolute fact that stock vs stock the 12900k is more efficient not just in ST workloads (see picture below from gnexus) but also in MT workloads (of course, im talking about stock, not tuning)
Last edited:
Caporegime
Obviously 13900k will be more efficient than the 12900k, not because of extra E cores, but because of extra cores. Even if they added P cores the efficiency would go up.
Regardless, Intel's spec sheet is on their webpage, the 12900k has a 125/241 power limit, so after 56 seconds that the TAU expires, it is more efficient than the 7950x, at least in CBR23. That's just a fact. The only thing you can argue is whether 125/241 is stock settings or not, and im all open to talking about that (since it will apply for the 13900k), but if it is, then it is an absolute fact that stock vs stock the 12900k is more efficient not just in ST workloads (see picture below from gnexus) but also in MT workloads (of course, im talking about stock, not tuning)
Intel only have themselves to blame for adding a high wattage boost to their chips that just so happens to last just long enough to complete a Cinebench run. Those are your shenanigan's Intel.
The FACT is its scoring 27,000 at 240 watts, the 7950X score 38,000 at about 240 watts, the latter is 40% higher at the same power.
More over the 7950X scores higher at 65 watts than the 12900K does at 240 watts, if at 240 watts the 12900K cannot keep pace with a 7950X at 65 watts it has no chance of winning the efficiency argument at any sane reasoning.
Last edited:
Permabanned
- Joined
- 28 Sep 2018
- Posts
- 0
Raptor lake will be more efficient because of process optimizations in voltage domain.
Raptor lake wouldn’t even exist at current specs if the adl node process carried over as is.
Raptor lake wouldn’t even exist at current specs if the adl node process carried over as is.
Soldato
But who cares, whos running rendering for...56 seconds? LOL.
An actual professional user who cares about efficiency out of the box will realize that the 7950x is less efficient than the 12900k. Unless, of course, he finishes his workload in ....56 seconds
Obviously 13900k will be more efficient than the 12900k, not because of extra E cores, but because of extra cores. Even if they added P cores the efficiency would go up.
Regardless, Intel's spec sheet is on their webpage, the 12900k has a 125/241 power limit, so after 56 seconds that the TAU expires, it is more efficient than the 7950x, at least in CBR23. That's just a fact. The only thing you can argue is whether 125/241 is stock settings or not, and im all open to talking about that (since it will apply for the 13900k), but if it is, then it is an absolute fact that stock vs stock the 12900k is more efficient not just in ST workloads (see picture below from gnexus) but also in MT workloads (of course, im talking about stock, not tuning)
I'm out - Extra cores does not mean more efficient, not at all, not now not ever. The ratio of P to E will effect efficiency at least where multi threaded efficiency is concerned. Again this just shows that you don't understand how to calculate efficiency or what efficiency is, as I said previously power draw which is the chart you have shown has no bearing on efficiency. Work completed / Time / Power draw etc all feed into a calculation of efficiency sure but as you stated to me previously and I agreed with high power draw does not necessarily mean its not efficient when the other factors are considered rather than just energy consumption in a silo which is what the chart above is.
Soldato
Didn't you know only Bencher is qualified to test anything at all, and what they say is fact and true, just like Trump!
Caporegime
Last edited:
But who cares, whos running rendering for...56 seconds? LOL.
An actual professional user who cares about efficiency out of the box will realize that the 7950x is less efficient than the 12900k. Unless, of course, he finishes his workload in ....56 seconds
What is that meant to mean ? Jay's upload he kept it looping to let it saturate with heat and the score didn't go down . How it works with temps is different
He plans to do another one where he directs heat on it to force it to drop
Go to the channel and watch almost feels like you have the chip and been testing it yourself
Last edited:
Soldato
Yes, it actually does. More cores at the same wattage means each core will be in a more ideal voltage / frequency curve. That's like knowledge 101. Also P cores are generally more effective than E cores at ISO wattage of anything above 3 watts per core.
Soldato
Wasnt talking about the 7950x. Was talking about the 12900k (and probably the 13900k). After 56 seconds TAU expires and they run at 125w by Intel specs. And when the 12900k runs at 125w (so after 56 sec) it's more efficient than the 7950x at stock. Or at least in CBR23, dunno about other stuff
Soldato
Today, Bencher learnt what PL1, PL2 AND TAU mean.....
Soldato
Soldato
Personal attacks are for people that know they are wrong so they don't have any arguments. So yeah, keep it up 
Personal attacks are for people that know they are wrong so they don't have any arguments. So yeah, keep it up
Dude, everyone in the world is reporting otherwise - Let the people decide based on proper testing and real results.. If you think i'm wrong then thats fine but please stop insulting the intelligence of members with what is effectively just misinformation.
Edit: Also quite how you read that as a personal attack is beyond me.
Last edited:
Lets all accept something - that it's fantastic that both AMD and Intel are releasing exciting new CPU's that have us all talking. Gone are the days of 5% IPC, same core count. Now we have exciting V-cache, E cores, CCX's, etc etc and each generation seems much more substantial. What a time to be alive!