I know that these CPUs get faster and faster, but I can't see the energy efficiency improvement. It's not there! If we divided compute by energy used, perhaps it wouldn't look that bad. And I'm more than ok with short-term boost. It's awesome! The compute demand nature is spikes. Why wouldn't CPU design address that? But high-end CPUs suck a lot of power. E cores are just a name to show them as "less power hungry" furnaces which P cores are. GPUs get more and more power-hungry too, if not more, just to be 1st. And the rest follows. One can see that even by temperature increase in their home. Having said that, I want games to look good on my rig, and run butter smooth without hiccups. Unless game and hardware designers address transient frame render time spikes, I feel like even despite said all that, I'll fuel this madness

Ven0m:

I know that these CPUs get faster and faster, but I can't see the energy efficiency improvement. It's not there! If we divided compute by energy used, perhaps it wouldn't look that bad. And I'm more than ok with short-term boost. It's awesome! The compute demand nature is spikes. Why wouldn't CPU design address that? But high-end CPUs suck a lot of power. E cores are just a name to show them as "less power hungry" furnaces which P cores are. GPUs get more and more power-hungry too, if not more, just to be 1st. And the rest follows. One can see that even by temperature increase in their home. Having said that, I want games to look good on my rig, and run butter smooth without hiccups. Unless game and hardware designers address transient frame render time spikes, I feel like even despite said all that, I'll fuel this madness

Well, my 12900k have 40-50% more fps per watt in cpu bound games than my 5950x. Both running max tuned memory.... just saying.... Looks like I'll throw out my 5950x to put in a 13600k 😛

the real reason this works is because they get rid of HT which has been a bit of a double edged sword for AMD and Intel alike for some years. Hopefully the price is right on this gen.

nizzen:

Well, my 12900k have 40-50% more fps per watt in cpu bound games than my 5950x. Both running max tuned memory.... just saying.... Looks like I'll throw out my 5950x to put in a 13600k 😛

The 5950x is a wasted CPU for gaming to begin with, of course a 13600k is better at gaming, because half of the 5950x is just waiting for work anyway.

TLD LARS:

The 5950x is a wasted CPU for gaming to begin with, of course a 13600k is better at gaming, because half of the 5950x is just waiting for work anyway.

They're both a waste for gaming. Unless you're into pissing contests of who can get the highest possible frame rate, a 12700F is plenty good enough and not overpriced.

schmidtbag:

They're both a waste for gaming. Unless you're into pissing contests of who can get the highest possible frame rate, a 12700F is plenty good enough and not overpriced.

to be fair my sisters i7 980 is still "good enough" for gaming. I gave it to her with a modest 4ghz OC after replacing it with a 1600X. I actually miss the machine myself. I love X58.

on Intel, it bothers me that I have to change the motherboard every six months. This minimizes the lifespan of motherboards and CPUs, which I think is neither ecological nor economical. intel does not show any dizzying progress and constantly forces motherboard manufacturers to make new changes. the overall performance of ddr5 vs ddr4 is nothing dizzying due to the lower bus width compared to ddr4. Compared to 10 gen eg 10600kf do I have to invest in ddr5 and motherboard? A lot of customers who bought 10 gen will surely wait for Intel to arrive with better price and power per watt. intel cpu temps are dizzying compared to amd. and consumption as well. I see no reason to invest in another motherboard and ddr5 again after a year. everything runs smoothly on my i5 10600 kf

TLD LARS:

The 5950x is a wasted CPU for gaming to begin with, of course a 13600k is better at gaming, because half of the 5950x is just waiting for work anyway.

Wrong! 5950x was the best AMD cpu for gaming before 5800x3d came. Best binned cores for higher boost clock. Best oc potential. Often good imc for high memory oc.

Zorgan.cz:

on Intel, it bothers me that I have to change the motherboard every six months. This minimizes the lifespan of motherboards and CPUs, which I think is neither ecological nor economical. intel does not show any dizzying progress and constantly forces motherboard manufacturers to make new changes.

I doubt Intel has needed to force the mobo manufacturers into anything, in that particular regard. They ought to be happy to be able to sell increasingly more expensive motherboards to consumers year after year. Sometimes twice a year. It's their business ecosystem. If you really could use the same mobo for multiple generations of CPUs without really gaining anything significant by switching to a new mobo, it would ruin the mobo manufacturers' financies. The CPU manufacturers need the mobo manufacturers as well for their own business, so they won't backstab them. AM4 was probably as far as it could go, with the last gen offering PCIe 4.0 as an incentive.

Zorgan.cz:

on Intel, it bothers me that I have to change the motherboard every six months. This minimizes the lifespan of motherboards and CPUs, which I think is neither ecological nor economical. intel does not show any dizzying progress and constantly forces motherboard manufacturers to make new changes. the overall performance of ddr5 vs ddr4 is nothing dizzying due to the lower bus width compared to ddr4. Compared to 10 gen eg 10600kf do I have to invest in ddr5 and motherboard? A lot of customers who bought 10 gen will surely wait for Intel to arrive with better price and power per watt. intel cpu temps are dizzying compared to amd. and consumption as well. I see no reason to invest in another motherboard and ddr5 again after a year. everything runs smoothly on my i5 10600 kf

Why do you need to upgrade cpu every 6 months? Most people with 4770k/4790k had cpu for 5+ years. Raptor Lake will use ddr4 too 🙂

tty8k:

And you're gaming at 720p on a 3090 ....

3440x1440 mostly 1080p 360hz on some fps shooters. We all know you only play CandyCrush on a old phone. Sorry about that 😛

So Intel 12600k is less power used and P-core 4.9 GHz E-core 3.6 GHz .. but the 13600k has more boost speed P-core 5.1 GHz E-core 4 GHz and use more power, sounds to me like Intel just oc the 12600k

nizzen:

Wrong! 5950x was the best AMD cpu for gaming before 5800x3d came. Best binned cores for higher boost clock. Best oc potential. Often good imc for high memory oc.

Show me The average oc on HWbot is a little higher on a 5800x then 5950x, but practically the same. Splave has run both a 5800x and 5950x, he topped out at 6ghz LN2 on both of them. Why would a 5950x have better memory overclocking when the memory controller is on the IO chip? The max frequency should be locked to the fabric 1:1 if you want max performance, so why does it matter?

TLD LARS:

Show me The average oc on HWbot is a little higher on a 5800x then 5950x, but practically the same. Splave has run both a 5800x and 5950x, he topped out at 6ghz LN2 on both of them. Why would a 5950x have better memory overclocking when the memory controller is on the IO chip? The max frequency should be locked to the fabric 1:1 if you want max performance, so why does it matter?

Pretty much every 5950x has better binned cores than lower models. No need to show it, because everyone knows it (exept you aparently) First rev og zen 3 tend to have better binned "infinity fabric" than lower models. The trend was there. Why on earth do you think ln2 results has anything to do with what I talk about? I overclocked with ln2, an I can say that ln2 is changing the rules. You want the samples with the most power leak. They often scales better with "voltage" on ln2. Read; the worst samples for air cooling.

nizzen:

Pretty much every 5950x has better binned cores than lower models. No need to show it, because everyone knows it (exept you aparently) First rev og zen 3 tend to have better binned "infinity fabric" than lower models. The trend was there. Why on earth do you think ln2 results has anything to do with what I talk about? I overclocked with ln2, an I can say that ln2 is changing the rules. You want the samples with the most power leak. They often scales better with "voltage" on ln2. Read; the worst samples for air cooling.

The average overclock of the 5800x and 5950x is still the same on HWbot, they include a good deal of air and watercooling too. Guru3d overclock had practically identical frequencies and voltages in 5800x,5900x and 5950x launch day review. LN2 overclocking is not about getting the most leaking CPU, It as about getting the best scaling one as you also say, this is what the ASUS score is all about, higher score = better scaling (mostly)

Kool64:

to be fair my sisters i7 980 is still "good enough" for gaming. I gave it to her with a modest 4ghz OC after replacing it with a 1600X. I actually miss the machine myself. I love X58.

That's why I roll my eyes when I hear people focusing so much in gaming performance for the CPUs ,if you do not target 240hz or more . It does not matter , if someone running a 3070 or 3060ti ... Again it matters a lot less because the cards become the bottleneck way before a 3090 .... Literally almost any random CPU from the shelf paired with a semi decent 120-144 Hz freesync monitor will give you nice fluid gaming . Everyone is a l33t mlg pro 360 sniper no scope nowdays 😛

Venix:

That's why I roll my eyes when I hear people focusing so much in gaming performance for the CPUs ,if you do not target 240hz or more . It does not matter , if someone running a 3070 or 3060ti ... Again it matters a lot less because the cards become the bottleneck way before a 3090 .... Literally almost any random CPU from the shelf paired with a semi decent 120-144 Hz freesync monitor will give you nice fluid gaming . Everyone is a l33t mlg pro 360 sniper no scope nowdays 😛

Everyy guru is like 60hz v-sync 58fps locked nowdays... Imagine under 144hz monitor in 2022! Hello 2010 🙄 144hz 1080p came ~2012 Even the mouse pointer is lagging like crazy on a 60hz monitor. Don't even talk about scrolling up and down in a browser. Cringe 😕

TLD LARS:

The average overclock of the 5800x and 5950x is still the same on HWbot, they include a good deal of air and watercooling too. Guru3d overclock had practically identical frequencies and voltages in 5800x,5900x and 5950x launch day review. LN2 overclocking is not about getting the most leaking CPU, It as about getting the best scaling one as you also say, this is what the ASUS score is all about, higher score = better scaling (mostly)

It's very easy to see you didn't try any Zen3 cpu.

Too bad that the increase in multicore score is driven by adding alot of E-waste cores, that are worthless for gaming.

nizzen:

Everyy guru is like 60hz v-sync 58fps locked nowdays... Imagine under 144hz monitor in 2022! Hello 2010 🙄 144hz 1080p came ~2012 Even the mouse pointer is lagging like crazy on a 60hz monitor. Don't even talk about scrolling up and down in a browser. Cringe 😕

First of all I am not talking about resolution at all , resolution is not affecting the fps output because of the CPU .... If you can do 500 fps on 720p you can do it on 4k too assuming you have the GPU power to pull it of. I never said 60hz is ideal as a matter of fact I did not even mention 60hz at all, I directly mention 144hz regardless of resolution , 120hz are fine too but in general they seem to be gone or almost gone and the standard to be 144hz that's why I went with 144hz , while 240 Hz+ are better it is common knowledge the higher you go you get deminishing returns. Especially if you factor the price you have to pay to actually get that extra tiny bit of smoothness. You can belittling people for not spending the amounts of money you spend and adding stuff that where never said as much as you like really. But please go ahead and post me some cs:go benchmarks showing how the glorious 12900 and 58003d can pull 600 fps with a 3090 vs a 5600 or 12600 that can do only 500 fps and somehow present it like the mid range CPUs can not game . You mock people that they play cinebench ... You do the exact same thing with numbers from games what changes is the program you use to measure performance. Anyway cheers I have no issue with what you buy or how you use it cheers and power to you . It is just a cringefest when numbers like 20% more fps in this CPU when you put on a 360 rad custom water cooling spend 500 extra on ram and you present it like it is the only way to somehow. And to do a car analogy is like you try to convince everyone they need a Bugatti Veyron as their daily driver and Their Audi a8 is so old it practically can not move !