I can finally make this thread.
Phoenix is indeed RDNA3. My advice to everyone: treat the old APU rumours as being out of date.
Phoenix is indeed RDNA3. My advice to everyone: treat the old APU rumours as being out of date.
If this is the case then there is a severe tdp bottleneck, with 12CU it should perform AT LEAST like a RX 6400 if there are no bottlenecks if the gpu clock is around 2300mhz. And the RX6400 trades blows with the GTX1650.
Considering it is a RDNA3 this this should already be matching a rx570 at the very minimum so no idea of what is going wrong here.
Regardless, i trust that youtuber, he has no reason to lie, and unless something is really really wrong there is no way RDR2 is running at 20fps at 1080p low, even a RX560 can get you consistent +40fps at 1080p low. Hell, even a 5700G can get you around 30.
Again as usual you're wrong. This is what happens when you play a game on the Iris Xe iGPU in my 11370H:As is usual with you that s the other way around, a 7800X3D require 60W to drive a 300W 4090, a 7800X a little more, it s no different for an AMD APU, CPU power is meaningless compared to the GPU power drain.
As is usual with you that s the other way around, a 7800X3D require 60W to drive a 300W 4090, a 7800X a little more, it s no different for an AMD APU, CPU power is meaningless compared to the GPU power drain.
With a 680M at 30W the GPU use 80% of the power; at 40W the GPU manage to work at full 2.2 frequency meaning that the CPU part need a much lower power, indeed CPU usage is very low.
AMD Radeon 680M benchmarks, review (RDNA2 AMD iGPU) vs Iris Xe, Intel
The AMD Radeon 680M is the much-awaited top-tier RDNA2-based integrated GPU bundled with the majority of the 2022 AMD Ryzen 6000 Rembrandt processors.www.ultrabookreview.com
That being said we still dont know why the "review" you linked display the 780M as lower performing than a Vega 8 in R2D2; you had to come with foolish explanations like the one in the post i quoted, anyway keep entertaining us with your phantasmogoric analysis, at least you bring some fun...
Again as usual you're wrong. This is what happens when you play a game on the Iris Xe iGPU in my 11370H:
View attachment 80481
The iGPU clearly consumes less power than the CPU cores. This is on Intel. I have no reason to believe that it's any different on AMD.
Completely irrelevant. AMD's own listing shows 2800MHz as well as third part tests of 7940HS show it clocks up to 2800MHz out of the box.Notebookcheck disagrees with you. AMD told them they have lowered it to 2700 Mhz on 7940HS.
Completely irrelevant. AMD's own listing shows 2800MHz as well as third part tests of 7940HS show it clocks up to 2800MHz out of the box.
Existing inventory? So basically Rembrandt?
Barcelo-R may still be existing Cezanne inventory rebadged and possibly with slightly updated firmware.Barcelo more like it. I took a fresh look at what Lenovo was claiming was "New Arrivals"... and while there's Rembrandt and Dragon Range models, they still clearly favor Barcelo.
It is mostly 7000 series now too so it's not that stale of product.
It isn't irrelevant because real clock speed matters and not a paper clock speed which could be wrong. And this is the question, is it wrong (not the first time) from AMD or is Notebookcheck wrong.
Barcelo-R may still be existing Cezanne inventory rebadged and possibly with slightly updated firmware.
HWiNFO for some reason does not accurately report the GPU power consumption in AMD APUs. Proof?No surprise, i was about to point from where your mistake come, that is , that you surely used your own laptop as reference.
Fact is that Intel for some reason,and contrary to AMD, prioritize the CPU over the GPU, when it comes to power delivery, hence you came back with even more clulessness.
Believe what you want, numbers are numbers and i linked a review that contradict your sayings, you can check elsewhere, there s other reviews that point this technical choice from AMD.
Edit : Check the CPU usage in the link i already posted, 2-3x the CPU usage for Intel vs AMD...
Evidence is here that AMD CPU part consume less that Intel CPU part in gamings, we dont even need HVinfo, the real debate would be to find the reason, either Intel has non optimised drivers or Radeon has, by design, much less driver ovehread or both.HWiNFO for some reason does not accurately report the GPU power consumption in AMD APUs. Proof?
Here:
This is running Prime95. As you can clearly see, package power is 45 W(as expected, because long duration power limit is 45 W). CPU Core Power = 36.2 W. CPU SoC Power = 3.9 W. Obviously CPU Core + SoC Power = 40.1 W, which is NOT GPU power, because this is Prime95.
Now a GPU Load. Furmark:
CPU Package Power = 45.4 W, CPU Core Power = 35.8 W, CPU SoC Power = 4.7 W. CPU Core + SoC Power = 40.5 W
So regardless of a full CPU or a full GPU load, CPU Core Power remains similar, as far as reporting is concerned, and none of the other reported values show the ACTUAL GPU power consumption. Before you ask, what about GPU ASIC power, here's the author of HWiNFO clarifying what it is for APUs:
View attachment 80527
So, yeah - till now, neither you, nor anybody else, has actually given any evidence for me to believe that running a 3D load like a game on your AMD APU in laptops means that the GPU part of the APU consumes more power than the CPU part.
Both AMD and Valve engineers have stated the CPU always takes priority over the GPU on regular chips. Now that does not mean the GPU can’t use more power, that means that if the CPU needs more power, the GPU is going to throttle back.HWiNFO for some reason does not accurately report the GPU power consumption in AMD APUs. Proof?
Here:
This is running Prime95. As you can clearly see, package power is 45 W(as expected, because long duration power limit is 45 W). CPU Core Power = 36.2 W. CPU SoC Power = 3.9 W. Obviously CPU Core + SoC Power = 40.1 W, which is NOT GPU power, because this is Prime95.
Now a GPU Load. Furmark:
CPU Package Power = 45.4 W, CPU Core Power = 35.8 W, CPU SoC Power = 4.7 W. CPU Core + SoC Power = 40.5 W
So regardless of a full CPU or a full GPU load, CPU Core Power remains similar, as far as reporting is concerned, and none of the other reported values show the ACTUAL GPU power consumption. Before you ask, what about GPU ASIC power, here's the author of HWiNFO clarifying what it is for APUs:
View attachment 80527
So, yeah - till now, neither you, nor anybody else, has actually given any evidence for me to believe that running a 3D load like a game on your AMD APU in laptops means that the GPU part of the APU consumes more power than the CPU part.
Load it consume 56/63W, that show how Impressive Steam deck/Aerith is.Review of the Asus hand held device :
That actually depends on the power mode used. Quiet is most comparable to Steam Deck, though Ally has a bigger screen and the review cops out by stating "The Quiet power mode is not suited for gaming since the performance is just too low." 🤷Load it consume 56/63W, that show how Impressive Steam deck/Aerith is.
Can you post a quote/source for that detail? Don't think we've ever heard the specifics of the "tuning" AMD does.As I said, the gaming oriented CPUs like the Z1 are different.
"The Quiet power mode is not suited for gaming since the performance is just too low." 🤷
"Yazan noted on stream that the Deck's power management is tuned to emphasize GPU clock over almost anything else. According to Yazan, the system will throttle charging, downloads, and even SSD I/O to maintain GPU clock rates if necessary."Can you post a quote/source for that detail? Don't think we've ever heard the specifics of the "tuning" AMD does.
Ok, thanks. But then has anyone shown how that translates to real power numbers in gaming? Obviously they can't throttle the CPU down to nothing. And somewhat related, but I'd expect uncore power to correlate tightly with the GPU load, given bandwidth demands."Yazan noted on stream that the Deck's power management is tuned to emphasize GPU clock over almost anything else. According to Yazan, the system will throttle charging, downloads, and even SSD I/O to maintain GPU clock rates if necessary."
Valve Details Steam Deck’s AMD Aerith CPU, FSR, Game Load Time, And You Can Build One Too
A livestream targeted at developers reveals a few new juicy details on Valve's handheld game system.hothardware.com
Quote is from somewhere in this live stream:
When playing one can show the power numbers split between CPU and GPU on screen. I'd think there are some such recordings on youtube etc.Ok, thanks. But then has anyone shown how that translates to real power numbers in gaming? Obviously they can't throttle the CPU down to nothing. And somewhat related, but I'd expect uncore power to correlate tightly with the GPU load, given bandwidth demands.
Ok, thanks. But then has anyone shown how that translates to real power numbers in gaming? Obviously they can't throttle the CPU down to nothing. And somewhat related, but I'd expect uncore power to correlate tightly with the GPU load, given bandwidth demands.
Gaming CPU load does not scale 1:1 with graphics. The baseline game logic should be more than dominant. Do you have any actual data to suggest otherwise?If a 80 CUs 250W GPU need a 8C to output 80W then a 35W 12 CUs iGPU will mandate below 12W CPU power, GPU power has always been the main bottleneck in APUs.
Gaming CPU load does not scale 1:1 with graphics. The baseline game logic should be more than dominant. Do you have any actual data to suggest otherwise?