AMD Ryzen 5 7600

[big_glasses]

Yeah, so many people say that Ryzens consume ~50 W in idle. If that's so, how are there so many apps that consume less than 20 W in TPU testing?

Someone has to be right. Do some reviewers/users have power saving options disabled? Like C-states or whatever they have on AMD.

yea, seeing guru3d's review also, it falls basically in range of with most. Though they measure whole system (which of course is worse than TPU's testing, but it also shows how little importance the CPU can have vs rest of the system)

We show energy consumption based on the entire PC (motherboard / processor / graphics card / memory / SSD). This number depends and will vary per motherboard (added ICs / controllers / wifi / Bluetooth) and PSU (efficiency). Keep in mind that we measure the ENTIRE PC, not just the processor's power consumption

AMD Ryzen 5 7600 processor review (Page 6)

AMD has released the non-X version of their Ryzen 7000 series processor. The new update isn't intended for extreme performance but is tagged with a far friendlier 65W TDP. And that means better heat ...

www.guru3d.com

and here is one from https://www.ocinside.de/review/mainboard_asrock_z490_steel_legend/5/
to showcase difference between mobo's themself (there is more in the article, I just copied and marked a few of the same type)

or extremeTech (https://www.extremetech.com/computi...aws-so-much-power-its-warping-cpu-comparisons)

 

[Gica]

What do you determine as "consumption in real life"?
W1zzard already have 3 different power consumption test cases, that is actually testing the CPU power, and not relying on the software monitor (a combination of those is somewhat close to my real usage)

How?!
For video cards, it's simple: extensions to the auxiliary power and the PCIe slot. How does it only measure CPU consumption without VRM? Does it have an interposer between the motherboard socket and the processor?
It's a discussion for the sake of discussion.
I made a video (noob mode, but explicit) with the consumption of the ENTIRE SYSTEM in different stages, from idle to multitasking. If you want to contradict each other, let's do it with facts, so, contradict me with another video with AMD inside.

1. Idle
2. CPU-Z benchmark
3. Youtube 1080p@30FPS
4. Youtube 4K@60FPS
5. Youtube 8k@60FPS
6. Cinebench R23 multi
7. WoT 1080p Low
8. 4K@60FPS with Medial Player Classic
9. Idle

Filmed with 50 fps screen with 100 fps even with playback at 60 fps is not a very good idea. There are some desynchronizations, small, fortunately.

System:
CPU: i5-12500 Alder Lake (65W PL1, 117W PL2)
GPU: UHD 770 (igp)
MoBo: Gigabyte B660M DS3H
RAM: 2 x 8GB@3200MHz CL15 DDR4
Cooler: AQIRYS Uranus LS (custom backplate)
Sound: external Xonar U7 MKII
PSU: Corsair CX450, 450W, 80+, Bronze
2x NVME, 2x SATA SSD, 1x external HDD 2TB

 

[THU31]

How does it only measure CPU consumption without VRM? Does it have an interposer between the motherboard socket and the processor?

They measure the power coming through the EPS connector. How is VRM relevant here? VRM affects the efficiency of the power conversion, so the CPU will actually use less power than is flowing through the EPS connector, some of it is wasted during conversion from 12 V.
Just like the efficiency of your PSU affects how much power is drawn from the socket.

It is the same method used to measure the power draw of graphics cards, except there they also have to check the PCI-E slot. They have special devices for that.

 

[big_glasses]

How?!
For video cards, it's simple: extensions to the auxiliary power and the PCIe slot. How does it only measure CPU consumption without VRM? Does it have an interposer between the motherboard socket and the processor?

How -> go to read the review page on power consumption....

All power measurements on this page are based on a physical measurement of the voltage, current and power flowing through the 12-pin CPU power connector(s), which makes them "CPU only," not "full system." We're not using the software sensors inside the processor, as these can be quite inaccurate and will vary between manufacturers. All measurements are collected and processed at a rate of 30 data points per second, on a separate machine, so the power measurement does not affect the tested system in any way. Our new data processing pipeline allows us to link recorded data precisely with benchmark runs, so we can easily create the charts below.

ref page: https://www.techpowerup.com/review/amd-ryzen-5-7600-non-x/22.html
I'm sure W1zzard can answer if you have more questions on it.

If you want to contradict each other, let's do it with facts, so, contradict me with another video with AMD inside.

I refer to my own post above, about how unviable it is to measure whole system (especially from SW), and how difference in other HW can impact even idle.
And I'm not even sure what that would accomplish (I'll also admit I can't be bothered, especially since Im posting from work, and I can't run HW Monitor or equivalent on the work laptops)? My system WILL use more power than yours, but it will also finish the tasks faster and have higher framerate. My system isn't even tuned for power efficiency, I'm not gonna deny it's power usage is high.

AMD Ryzen 5 7600

Yeah, so many people say that Ryzens consume ~50 W in idle. If that's so, how are there so many apps that consume less than 20 W in TPU testing? Someone has to be right. Do some reviewers/users have power saving options disabled? Like C-states or whatever they have on AMD. yea, seeing guru3d's...

www.techpowerup.com

Even power plan and OS will impact power usage (and efficiency)

AMD Ryzen 9 3900X Power Usage Is Running Measurably Higher On Linux Than Windows - Phoronix

www.phoronix.com

let's see the consumption in real life.

Is those cases what 8,5 cases you consider "real life"? (2nd idle being half a case)
AND how are those better than W1zzards cases?

 

[Gica]

They measure the power coming through the EPS connector. How is VRM relevant here?

So, we have:
1. CPU + VRM
2. "The 12V power connector mainly supplies power to the CPU" (you can find it in the manual of all Gigabyte motherboards).
So it's not just ONLY processor. In this case, a wattmeter offers much more accuracy if the software information doesn't seem right when the Intel comes out fine.

Is those cases what 8,5 cases you consider "real life"? (2nd idle being half a case)
AND how are those better than W1zzards cases?

From idle to Cinebench R23 (exceeds R20 in CPU stress), a benchmark with which you can include the maximum consumption of the entire encoding/rendering software package. Beyond it you can only find Prime95 - Small FFTs, but not to "play" Prime I spent money for the system. Maybe you.
Do you contradict me with a video proof, or do you continue to invoke the spirits?

Video
50 watts, minimum, in idle. One of the reasons why I avoided AMD until now for my system for office, www, etc (second: the weaknesses of the outdated Vega in hardware acceleration). The one in the video consumes almost as much while doing nothing as I do playing WoT.

 

[THU31]

But you have no graphics card in your system, so how can you possibly compare this to a Ryzen system using a graphics card?

 

[Gica]

Does it matter? A powerful video card makes it unnecessary to talk about CPU consumption. With a weak one (6-7W in idle) or igp I can see much better the consumption of the processor.
Reading the comments, that minimum 50W seems like a record for the r5 3600. Many complain about much higher consumption with that processor.
Regarding the consumption of my system, I consider the HWinfo information to be correct. The difference between 144W in Cine R23 (wattmeter) and ~81W CPU Package (HWinfo) can be easily explained with the performance of the power supply and the VRM + the consumption of the other components. If the "up" consumption of an Intel is ok for AMD fans, why wouldn't the "down" consumption be ok as well?

 

[THU31]

OK, so can you please explain these power consumption results that you ignored earlier - https://www.guru3d.com/articles_pages/amd_ryzen_5_7600_processor_review,6.html

Why is their total system power consumption (idle and single-threaded) so high on Intel? Sabotage? Or is it because they are using a 3090 in the system?

Look at all the APUs at the top of the list, which were tested without a graphics card.

 

[Gica]

I can't read those graphs rationally. I have no way.
So, 6900HX with 3090 (is it possible???) and only 11W???
Ryzen 3100 (4 cores) at the same idle consumption as 3800XT (8 cores) and higher than a 3600 (6 cores) ???

I found another one. 70W in idle with the 5700X, but this aspect did not catch my eye. With just a simple movement of the mouse it went up to 83W. Excuse me, I have ~25W in idle (by the way, 10500 drops to 21W with igp) and if you move the mouse it's like an airplane propeller and it doesn't jump by 30W. As an idea, 147 W fixed in the same test, with AVX CPU. Idle at 5700XT means 8W. Correlating his data with mine, I am sure that the measurements displayed by HWinfo are correct on my system.

One more thing, if you still like graphics, see that intel 12 and 13th is more efficient in single-threaded.

 

[big_glasses]

From idle to Cinebench R23 (exceeds R20 in CPU stress), a benchmark with which you can include the maximum consumption of the entire encoding/rendering software package. Beyond it you can only find Prime95 - Small FFTs, but not to "play" Prime I spent money for the system. Maybe you.
Do you contradict me with a video proof, or do you continue to invoke the spirits?

Not an answer. I didn't ask for Cinebcnh info, I didn't ask for Prime95 info. I do NOT see the point of me showing you that my system is more powerful and consumes more power than yours.
W1zzard have not used Prime95. He did use Blender (instead of Cinebench, but that's same-same). He used mp3 encode for single thread, blender for multi, avg. of 45 applications and 12 games.

Is those 8.5 cases what you consider IRL cases. YES or NO?

also will (mid?)face by saying I ain't watching some YT video. If it's in a video and not an article, it's not something I consider worth my time.

Above is a look at the CPU power consumption during the entire duration of ~24 hours worth of benchmarking, across the wide range of workloads carried out.

AMD Ryzen 5 7600 / Ryzen 7 7700 / Ryzen 9 7900 Linux Performance Review - Phoronix

www.phoronix.com

or this (similar to guru3d)

AMD Ryzen 5/7/9 7000 65W Tested: Low Power, High Performance - Page 3

AMD revealed its line-up of 65W Ryzen processors, the Ryzen 5 7600, Ryzen 7 7700, and Ryzen 9 7900, and we've put them to the test. - Page 3

hothardware.com

I'm also seeing how you pointedly ignored that given all the differences our systems will have (and mine not even being a low consumption "version" AMD CPU), that there is 0 point of getting. Mine WILL use more, unless you want to gloat that your APU consumes less than my system, there is 0 point in it.
I will also fully admit, that I don't want to use my personal at-home time on benchmarking useless cases. I have way too many things I'd rather be doing. And, I will l say again, the systems and settings is too different to be worth anything. Would be a shocker, a 5900X+GPU uses more than an APU, big shock. Call the goddamn Nobel committee for this shocking revelation
Dude, there is a reason people link to reviews for comparison, it's because the systems is disclosed, and often highly similar (RAM/Mobo/CPU being only difference)

Do you contradict me with a video proof, or do you continue to invoke the spirits?

Pardon me, but what the fuck does this even mean? "invoke the spirits"?

 

[Gica]

Not an answer. I didn't ask for Cinebcnh info, I didn't ask for Prime95 info. I do NOT see the point of me showing you that my system is more powerful and consumes more power than yours.

Man, try to learn how to do 1+1, because with you it breaks down in floating point and it comes out stupid.
Idle: 24.5W. The whole system.
Cinebench R23 (rendering, yes?): ~146W. The whole system.
1+1 = minimum and maximum consumption on that system. Everything I will run, from music, games, encodings, etc. will fall between these values. What is so hard to understand?!
I have had the processor and the motherboard since June 6, last year, when Zen 4 was just a rumor. With 330 euros, how much they cost together, I buy the cheapest Zen4 and 30% of the motherboard. For the rest... you calculate how much you need for zen 4, motherboard and memories.
Question: did you have the processor at 100% when you wrote these mileage posts? Do you think you are faster in the office with a zen4, or do the movies look different? Does he know how to cook and is he an expert in massage?
If we are arguing about consumption, correct me with facts, gentlemen. I threw down the gauntlet with this video recording, do you pick it up or hide behind other people's reviews?
P.S. Reviews also have small letters. I notice that you miss it.
P.P.S. It's true, a powerful video card skyrockets consumption. It makes the discussion about processor consumption pointless, but we are happy if our processor consumes 5 W less in an application that we don't even run. Maybe it consumes more in the applications we use (or is less efficient), but what does it matter if the reviewers decided otherwise.

 

[IntelSnowflake]

Awesome review, just want to give some inputs regarding stock cooler temperature.
The default voltage(1.24v as I remembered) is too high for 3.8Ghz, that's why you got 97 degree on max loading with the cooler at default speed.

You can manually set voltage to 1.0v and it still keeps the same performance. I've only got ~80 degree on max loading after doing so.
I can also use another PBO profile, which gives me 4.8Ghz/1.0v roughly. Still beyond the default clock performance and the temperature is completely fine.

As for its rival 13500, the stock cooler is bad too. But 13500 doesn't have much overclocking/voltage controls as 7600. Makes 13500 competitive in the benchmark only.

As for platform price, AM5 is not that expensive as everybody saying at all. The price starts to drop already.
Not to mention it has better longevity.

 

[daddelbud]

Is either going for this or the 7900

 

[Why_Me]

Is either going for this or the 7900

This setup was posted in two builds today on another site for peeps who were on a budget but wanting an AM5 build.

https://pcpartpicker.com/product/CFzhP6/msi-pro-b650-p-wifi-atx-am5-motherboard-pro-b650-p-wifi
MSI PRO B650-P WIFI $199.99

https://www.msi.com/Motherboard/PRO-B650-P-WIFI

https://pcpartpicker.com/product/yXmmP6/amd-ryzen-5-7600-38-ghz-6-core-processor-100-100001015box
AMD Ryzen 5 7600 $229.00

https://www.amd.com/en/products/apu/amd-ryzen-5-7600

https://www.newegg.com/g-skill-32gb/p/N82E16820374417
G.SKILL Flare X5 Series AMD EXPO DDR5 5600 32GB (2x16GB) CL36 $119.99

 

[daddelbud]

Thanks
Im leaning towards the 7900 as its easier to cool, but really waiting for the techpowerup review.

 

[SpAwNtoHell]

I recently got a b650 gygabyte motherboard and a 7600 with ddr5 6000 cl36, first install ram even tho on qvl did not boot with expo and boot times very long. Updated the bios and now even without expo boot times are under pretty fast to the point i strugle to get in bios at post a i end up very fast in windows login after a second or so of mb post.

So i assume the problem was solved? I think @W1zzard will report this in the upcoming 78003d review if still the case.

 

[Caring1]

Why isn't the integrated graphics tested by itself in the review and in comparison to the older G processors for those that don't want the expense of a graphics card?
I signed in just to look for that.

 

[THU31]

I signed in just to look for that.

I think you answered your own question. I doubt anyone else has ever asked for that.

What would you like to be tested? Newest AAA games?

This is a 0.5 teraflop iGPU. It's meant to be used just to drive the display and have video decoding capabilities. It's not meant for gaming or productivity acceleration.

 

[Chrispy_]

I think you answered your own question. I doubt anyone else has ever asked for that.

What would you like to be tested? Newest AAA games?

This is a 0.5 teraflop iGPU. It's meant to be used just to drive the display and have video decoding capabilities. It's not meant for gaming or productivity acceleration.

Yep. AMD discussed it when Zen4 launched; It's there for testing/backup and 2D applications. The fact it has gaming architecture is just a side effect of them requiring 2 compute units worth of unified shaders to handle parts of the encode/decode pipeline that aren't run on the fixed-function media engine.

Yes, you can technically execute games on it because the RDNA2 drivers and hardware fully-support current gaming APIs, but it only has 4 ROPs so you're going to be getting slideshow framerates in anything remotely demanding. Take the weakest GPU released in the last 3 years, cut away 80% of it, underclock it, and give it system RAM instead of GDDR6.

 

[Deleted member 232534]

somebody used Ryzen 7k series iGPU in the CAD applications in a virtual machine passthrough mode?

 

[Chrispy_]

somebody used Ryzen 7k series iGPU in the CAD applications in a virtual machine passthrough mode?

I haven't tried it but I could possibly try it if I'm at the London site on Friday. What hypervisor, ESXi?

 

[Deleted member 232534]

that would be amazing. I hope you can test hardware accelerated CAD app like Solidworks or Ironcad (free trial). Any type 1 hypervisor is perfect.

 

[Caring1]

I think you answered your own question. I doubt anyone else has ever asked for that.

What would you like to be tested? Newest AAA games?

This is a 0.5 teraflop iGPU. It's meant to be used just to drive the display and have video decoding capabilities. It's not meant for gaming or productivity acceleration.

I never mentioned gaming, Gamers need to get over themselves thinking everything with graphics capability is about them. It still needs to be tested against prior standards in iGPUs.
Hopefully reviewers take note and include details and not add graphics cards when testing CPU's, after all it is not GPU review.

 

[THU31]

I never mentioned gaming, Gamers need to get over themselves thinking everything with graphics capability is about them. It still needs to be tested against prior standards in iGPUs.
Hopefully reviewers take note and include details and not add graphics cards when testing CPU's, after all it is not GPU review.

But you still haven't mentioned what kind of tests you'd want to see. Should they guess?

And you kind of answered your question again. It's a CPU review, not a GPU review, nor an APU review.

Intel CPUs have had an iGPU present for over a decade, yet they're never tested in the CPU review. Why?

 

[Rabit]

somebody used Ryzen 7k series iGPU in the CAD applications in a virtual machine passthrough mode?

I do daily in hybrid GPU mode * option available in bios + Sapphire GPRO X080 mining edition RX 6700 10GB

Sapphire GPRO X080 Specs

AMD Navi 22, 2450 MHz, 2304 Cores, 144 TMUs, 64 ROPs, 10240 MB GDDR6, 2000 MHz, 160 bit

www.techpowerup.com