ryzen CPU and GPU power limited for no apparent reason.

[splashblox]

I have a AMD Ryzen 5 2400g and i noticed that it only uses 20w of the 65w TDP and my GPU (GTX 1650) only uses 56w of the 75w TDP.
i don't have any idea why, my power supply is 400w so it should provide de wattage. 

both GPU and CPU are power limited. 

btw my bios is stock.

[Sanedish]

Did this just happen or did you always have this issue and only noticed it now?

As i know Nvidia, they will probably have done an driver to prevent the 1650 to be used at full power, maybe try older drivers?

[splashblox]

2 hours ago, Sanedish said:

Did this just happen or did you always have this issue and only noticed it now?

As i know Nvidia, they will probably have done an driver to prevent the 1650 to be used at full power, maybe try older drivers?

i think i always had this issue, i always saw that cpu unperformed compared with the results from yt, i always had bottlenecks where i shouldn't.  

i guess i never checked the power.

[Sanedish]

8 hours ago, splashblox said:

i think i always had this issue, i always saw that cpu unperformed compared with the results from yt, i always had bottlenecks where i shouldn't.  

i guess i never checked the power.

Hm, thats weird then, maybe you have just got a bad lot/bought underperforming hardware?

[Tetras]

13 hours ago, splashblox said:

I have a AMD Ryzen 5 2400g and i noticed that it only uses 20w of the 65w TDP

First off, TDP is not what anything (CPU or GPU) actually uses anymore. It used to be fairly accurate, but it is now more of a guide for OEMs in terms of what is an acceptable level of cooling for the PC they're selling.

 

If your CPU and graphics card are using less, or more than their TDP, this is totally normal and not necessarily a problem.

 

The important thing is: with whatever power draw you have, can the CPU or GPU maintain their boost clocks? If the answer is: YES, then it is functioning correctly, regardless of the power draw.

 

From a quick look on the Internet, there are many circumstances where the power draw of the 2400G is very low and this is not unusual, since the G CPUs are a different design to the non-G CPUs and often have a different profile with their power usage.

 

The Ryzen CPUs are two chips together, the CPU and the I/O die, which together put your power usage at 38 watts, not 20 watts.

 

10 hours ago, splashblox said:

 i always saw that cpu unperformed compared with the results from yt, i always had bottlenecks where i shouldn't.  

Is your RAM 2400? It looks like it is, if so, that is probably one of the reasons why, but we'd really need to compare something like Cinebench to reviews, since YouTube videos don't always declare their whole system, overclocks and game settings.

 

13 hours ago, splashblox said:

my GPU (GTX 1650) only uses 56w of the 75w TDP.

I'd say the same thing here, that's likely to be normal if the card can successfully maintain the boost clocks. The 1650 is designed so that it stays under the 75 watt limit of the PCI-Express slot.

[splashblox]

13 hours ago, Tetras said:

First off, TDP is not what anything (CPU or GPU) actually uses anymore. It used to be fairly accurate, but it is now more of a guide for OEMs in terms of what is an acceptable level of cooling for the PC they're selling.

 

If your CPU and graphics card are using less, or more than their TDP, this is totally normal and not necessarily a problem.

 

The important thing is: with whatever power draw you have, can the CPU or GPU maintain their boost clocks? If the answer is: YES, then it is functioning correctly, regardless of the power draw.

 

From a quick look on the Internet, there are many circumstances where the power draw of the 2400G is very low and this is not unusual, since the G CPUs are a different design to the non-G CPUs and often have a different profile with their power usage.

 

The Ryzen CPUs are two chips together, the CPU and the I/O die, which together put your power usage at 38 watts, not 20 watts.

 

Is your RAM 2400? It looks like it is, if so, that is probably one of the reasons why, but we'd really need to compare something like Cinebench to reviews, since YouTube videos don't always declare their whole system, overclocks and game settings.

 

I'd say the same thing here, that's likely to be normal if the card can successfully maintain the boost clocks. The 1650 is designed so that it stays under the 75 watt limit of the PCI-Express slot.

it doesn't boost correctly, max boost i ever saw was 3.75 (i dont know why, but on the last image i did boost to 3.9, but i doesnt boost like that normally, in games the most i saw was 3.75ghz), (it was a one time thing), base is 3.6 and the max boost is 3.9ghz. Also, ok the 2400g doesn't use 65w but the problem here is that it only uses a max of 38, a bit more than half, and because of that it doesn't boost correctly. temps always are good, max 65°C. if you want screenshots from the wattage in game. 

Edited March 30 by splashblox

[Tetras]

51 minutes ago, splashblox said:

it doesn't boost correctly, max boost i ever saw was 3.75 (i dont know why, but on the last image i did boost to 3.9, but i doesnt boost like that normally, in games the most i saw was 3.75ghz)

The stock all-core boost is 3.75 Ghz and 3.9 Ghz for a single core, so games would probably not hit that high.

 

I believe if you run Timespy (free version) it produces a graph where you can see what the clock was over time.

[splashblox]

12 hours ago, Tetras said:

The stock all-core boost is 3.75 Ghz and 3.9 Ghz for a single core, so games would probably not hit that high.

 

I believe if you run Timespy (free version) it produces a graph where you can see what the clock was over time.

i couldn't download timespy in time because my internet is slow as f, so i used Cinebench to see if my points where in the margin of error, yes i know that 3.9ghz is for 1 core, my issue is that i doesn't boost to 3.9 like ever, even if i have thermals to spare. Btw in Cinebench i didn't hit 3.9ghz, i don't know if it's a Cinebench thing, I'm kinda new to benchmarking.

 

i got the benchmarks from here.  

 

[Tetras]

1 hour ago, splashblox said:

i couldn't download timespy in time because my internet is slow as f, so i used Cinebench to see if my points where in the margin of error, yes i know that 3.9ghz is for 1 core, my issue is that i doesn't boost to 3.9 like ever, even if i have thermals to spare. Btw in Cinebench i didn't hit 3.9ghz, i don't know if it's a Cinebench thing, I'm kinda new to benchmarking.

Hmm. I thought it would be easy to find, but it is actually not, because the reviews like @ Guru3D use a much older version of Cinebench.

 

From what I could gather, the quoted number is overclocking, because the record on hwbot for a 2400G is 1031. Not very helpful I'm afraid.

 

On the plus side, your result is reasonable, so if something is bad, it is not very bad.

[splashblox]

7 hours ago, Tetras said:

Hmm. I thought it would be easy to find, but it is actually not, because the reviews like @ Guru3D use a much older version of Cinebench.

 

From what I could gather, the quoted number is overclocking, because the record on hwbot for a 2400G is 1031. Not very helpful I'm afraid.

 

On the plus side, your result is reasonable, so if something is bad, it is not very bad.

Got the Cinebench R15 results. here are the guru results.

[Tetras]

9 hours ago, splashblox said:

Got the Cinebench R15 results. here are the guru results.

Hmm, so as before, they're not too bad, but could be better.

 

Could you take a picture of your BIOS settings for the CPU (e.g. PBO, PB2, XFR)?

 

 

[splashblox]

5 hours ago, Tetras said:

Hmm, so as before, they're not too bad, but could be better.

 

Could you take a picture of your BIOS settings for the CPU (e.g. PBO, PB2, XFR)?

 

 

i should've mentioned before, my mobo is a ASRock a320m, it doesnt have PBO but the core performance boost on auto, the options are auto and disabled.

if it helps, the site for the mobo is here.