AMD Ryzen 7000 Series CPUs may use up to 170W, according to rumors.

Published by

Click here to post a comment for AMD Ryzen 7000 Series CPUs may use up to 170W, according to rumors. on our message forum
https://forums.guru3d.com/data/avatars/m/260/260103.jpg
Will be interesting to see how these perform, but some already complain that the 5950 is already to hot even with good aftermarket cooling. Still i'm way overdue for an upgrade, so i'll see how reviews go.
https://forums.guru3d.com/data/avatars/m/72/72830.jpg
170w is OK, unlike modern GPUs this is avoidable because most have absolutely no need for 16 cores at all.
https://forums.guru3d.com/data/avatars/m/90/90667.jpg
Performance going to be top-notch probably, so CO with PBO2 and undervoltage could help a lot.
data/avatar/default/avatar16.webp
Only for 16c/32t and its a big maybe so its all fine. I wonder about Raptor Lake TDP..
https://forums.guru3d.com/data/avatars/m/164/164033.jpg
Maddness:

Will be interesting to see how these perform, but some already complain that the 5950 is already to hot even with good aftermarket cooling. Still i'm way overdue for an upgrade, so i'll see how reviews go.
If they manage to increase the size of the heatspreader and what not it should be better. Yes the 5xxx series runs hot even with quite low wattage. But also the surface area is small.
data/avatar/default/avatar29.webp
Now, Correct me if I am wrong but: 170W, 5nm? Lets consider that 5 nm means more transistor density, so heat built up would be higher. Then, 65W more? And efficiency benefits from going to 5nm? To me that sounds like its either flat out wrong OR (and this is my speculation) AMD needs that much power to get ZEN4 to a good place because their first gen ZEN4 is not particularly good. I got a 5900X on a 560 Radiator and that thing still reaches 68C max on stock (hotest core, short burst, good waterblock, Arctic silver 5, D5 pump, only CPU in that loop, Cinebench R15, 18C ambient). At 170W? Yeah.... no. Bare in mind maybe my setup is not perfect, but you cant expect people to use custom watercooling for a CPU as default.... right?
data/avatar/default/avatar23.webp
"May use up to 170w." ----> play CB23, and using 250w 😀 My old 5900xx said 105w?, but used 142w stock.... Who cares anyway. If the performance is there, bring it on 😎
https://forums.guru3d.com/data/avatars/m/180/180081.jpg
TDP is quite different from actual power draw though. It might have 105w TDP, but how much power does it actually use? 😛
https://forums.guru3d.com/data/avatars/m/229/229509.jpg
It's still half of what current intel chips are using, so I'd call that a win!
https://forums.guru3d.com/data/avatars/m/250/250418.jpg
"May" and "up to" are to be taken with a grain of salt. We know TDP means nothing and actual power draw is certainly higher. I hope they keep being efficient, even if we have to tune them ourselves. As someone said, very few of us need 16 cores so 6 and 8 cores should still draw the usual.
https://forums.guru3d.com/data/avatars/m/256/256969.jpg
5900X with slight overclock are already at 190W+ for multi thread workload
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
If this is how much the CPU uses under certain bursts of heavy loads and with boost clocks then I'd say that's totally reasonable, especially compared to Intel. If 170W was the new TDP then I'd be concerned.
https://forums.guru3d.com/data/avatars/m/126/126739.jpg
nizzen:

"May use up to 170w." ----> play CB23, and using 250w 😀 My old 5900xx said 105w?, but used 142w stock.... Who cares anyway. If the performance is there, bring it on 😎
My 850 watt PSU with a new 600W Nvidia GPU and a 170W CPU
this-is-fine.0.jpg
https://forums.guru3d.com/data/avatars/m/270/270008.jpg
Catspaw:

Now, Correct me if I am wrong but: 170W, 5nm? Lets consider that 5 nm means more transistor density, so heat built up would be higher. Then, 65W more? And efficiency benefits from going to 5nm? To me that sounds like its either flat out wrong OR (and this is my speculation) AMD needs that much power to get ZEN4 to a good place because their first gen ZEN4 is not particularly good. I got a 5900X on a 560 Radiator and that thing still reaches 68C max on stock (hotest core, short burst, good waterblock, Arctic silver 5, D5 pump, only CPU in that loop, Cinebench R15, 18C ambient). At 170W? Yeah.... no. Bare in mind maybe my setup is not perfect, but you cant expect people to use custom watercooling for a CPU as default.... right?
I suspect based off what we know already that 5nm custom node created for AMD can hit some nice frequencies over 5Ghz. This is a departure from the current 7nm process. I expect this 170w part to be a high clocked version of a 16 core part. If you look at some of the leaks from those who know the 8-core parts are all 65 watt parts.
https://forums.guru3d.com/data/avatars/m/282/282473.jpg
very low imo
data/avatar/default/avatar18.webp
Well I can already make my 5950X consume 160W with an all core load at 4.4Ghz 1.1V, so not bad if the performance goes up a good notch.
https://forums.guru3d.com/data/avatars/m/271/271560.jpg
the deal with node shrinkage is a smaller node equals higher performance or lower TDP. as in either/or when AMD introduced ryzen they could have both higher performance and lower TDP because they were three nodes smaller which continued until AL introduced e cores which dropped their (still high) TDP Ryzen 4 will be two nodes smaller than AL and the performance differential should be similar ( 7n vs 5n ) but the power envelope has increased because this gen AMD's uArch has stiff competition unlike last.
data/avatar/default/avatar34.webp
Catspaw:

Lets consider that 5 nm means more transistor density, so heat built up would be higher. Then, 65W more? And efficiency benefits from going to 5nm?
The chip sounds like it is still parted up in 8 core dies, so the higher density will not scale 1-1 and things like double cache and other design changes, will maybe spread out the cores more and thereby spread out the heat generation. A thicker heatspreader could also spread the heat out better, just like Intel previously has experimented with different heatspreaders and thermal transfer paste and solder methods with different results. Guessing the cooling requirements already now, is very difficult, the 170W could also be full speed AVX performance or maybe AMD now has a short allcore boost that needs more then normal power for a set time, like Intel does with the 241W timed boost TDP (ignored by most motherboards).
data/avatar/default/avatar16.webp
170W was a typical FX with a slight OC.
https://forums.guru3d.com/data/avatars/m/271/271560.jpg
TLD LARS:

The chip sounds like it is still parted up in 8 core dies, so the higher density will not scale 1-1 and things like double cache and other design changes, will maybe spread out the cores more and thereby spread out the heat generation. A thicker heatspreader could also spread the heat out better, just like Intel previously has experimented with different heatspreaders and thermal transfer paste and solder methods with different results. Guessing the cooling requirements already now, is very difficult, the 170W could also be full speed AVX performance or maybe AMD now has a short allcore boost that needs more then normal power for a set time, like Intel does with the 241W timed boost TDP (ignored by most motherboards).
exactly the fact is not only is the die smaller the over all soc is smaller - but the heatspreader has conventional dimensions of width and length. a thicker (depth) would take more heat before saturation enabling a loop (vs air) to have an slightly easier time to keep near ambient temps. but for air cooling not so much. you can do it but "big air" means big cases or noisy systems. and this is a case where i would expect a larger delta between air and liquid cooling vs. an OLD chip @ same power rating. i'm not confident at all of SFF air coolers with a Ryzen 9 - but they will easily work well with all the other SKUs