F5F Stay Refreshed Hardware Desktop Power usage is restricted on CPU and GPU without clear justification.

Power usage is restricted on CPU and GPU without clear justification.

Power usage is restricted on CPU and GPU without clear justification.

Pages (2): 1 2 Next
H
Hot_Girl_Katja
Junior Member
37
04-28-2016, 01:50 PM
#1
I observed that my AMD Ryzen 5 2400g operates at just 20W of its 65W TDP, while my GTX 1650 draws only 56W from its 75W TDP. I’m unsure why this happens despite having a 400W power supply—it should be sufficient. Both the CPU and GPU are limited by their power constraints. Also, my BIOS is the standard version.
H
Hot_Girl_Katja
04-28-2016, 01:50 PM #1

I observed that my AMD Ryzen 5 2400g operates at just 20W of its 65W TDP, while my GTX 1650 draws only 56W from its 75W TDP. I’m unsure why this happens despite having a 400W power supply—it should be sufficient. Both the CPU and GPU are limited by their power constraints. Also, my BIOS is the standard version.

W
wesselboy11
Member
221
04-29-2016, 12:52 AM
#2
This situation seems recent, though I understand it could have been a known issue with Nvidia. They might have released a driver to limit performance of the 1650 chip, or they could be suggesting using an older version.
W
wesselboy11
04-29-2016, 12:52 AM #2

This situation seems recent, though I understand it could have been a known issue with Nvidia. They might have released a driver to limit performance of the 1650 chip, or they could be suggesting using an older version.

L
levoyageur92
Posting Freak
807
04-29-2016, 04:41 AM
#3
I believe I’ve experienced this problem consistently. I often noticed the CPU underperforming when compared to the results from YouTube, and I frequently encountered bottlenecks where I shouldn’t have been. It seems I might have overlooked checking the power supply.
L
levoyageur92
04-29-2016, 04:41 AM #3

I believe I’ve experienced this problem consistently. I often noticed the CPU underperforming when compared to the results from YouTube, and I frequently encountered bottlenecks where I shouldn’t have been. It seems I might have overlooked checking the power supply.

E
ElTorchoN
Member
161
04-30-2016, 07:28 AM
#4
It seems odd, perhaps you've acquired a subpar or underperforming device.
E
ElTorchoN
04-30-2016, 07:28 AM #4

It seems odd, perhaps you've acquired a subpar or underperforming device.

F
filipl02
Member
59
04-30-2016, 04:16 PM
#5
TDP is no longer a strict measure for CPUs or GPUs. It once gave good insight, but now mainly helps manufacturers decide what cooling is suitable for their products. If your processor and graphics card draw less than their TDP, it’s completely fine and usually not an issue. What matters most is whether the device can keep its boost speeds under the available power. If yes, it works properly, no matter how much current it uses. Online sources show that many systems with 2400W draw can run smoothly, especially if the components handle their peak loads. The Ryzen setup combines parts to reach around 38 watts, not just 20. If your RAM is also 2400, that could contribute, but a full performance check—like Cinebench results—is better. The 1650 model is built to stay under 75 watts in its PCIe slot.
F
filipl02
04-30-2016, 04:16 PM #5

TDP is no longer a strict measure for CPUs or GPUs. It once gave good insight, but now mainly helps manufacturers decide what cooling is suitable for their products. If your processor and graphics card draw less than their TDP, it’s completely fine and usually not an issue. What matters most is whether the device can keep its boost speeds under the available power. If yes, it works properly, no matter how much current it uses. Online sources show that many systems with 2400W draw can run smoothly, especially if the components handle their peak loads. The Ryzen setup combines parts to reach around 38 watts, not just 20. If your RAM is also 2400, that could contribute, but a full performance check—like Cinebench results—is better. The 1650 model is built to stay under 75 watts in its PCIe slot.

T
thebjmax1
Senior Member
395
04-30-2016, 08:42 PM
#6
it doesn't increase properly, the highest boost I've ever seen was 3.75 (though on the last shot I reached 3.9, but it usually doesn't go that high). In games the typical max is around 3.75GHz, though I only saw up to 3.9GHz once. The base is 3.6 and the upper limit is 3.9GHz. Also, the 2400g doesn't draw 65W but the issue here is that it only handles up to 38, which is more than half, causing boost problems. Temperatures stay under 65°C. If you need screenshots showing wattage in-game, let me know. Edited March 30, 2024 by splashblox
T
thebjmax1
04-30-2016, 08:42 PM #6

it doesn't increase properly, the highest boost I've ever seen was 3.75 (though on the last shot I reached 3.9, but it usually doesn't go that high). In games the typical max is around 3.75GHz, though I only saw up to 3.9GHz once. The base is 3.6 and the upper limit is 3.9GHz. Also, the 2400g doesn't draw 65W but the issue here is that it only handles up to 38, which is more than half, causing boost problems. Temperatures stay under 65°C. If you need screenshots showing wattage in-game, let me know. Edited March 30, 2024 by splashblox

I
iAmGabe
Junior Member
8
04-30-2016, 10:19 PM
#7
The full-core speed is around 3.75 to 3.9 GHz per core, which likely won’t reach those numbers. I think using Timespy (free version) will show a chart of the clock speed over time.
I
iAmGabe
04-30-2016, 10:19 PM #7

The full-core speed is around 3.75 to 3.9 GHz per core, which likely won’t reach those numbers. I think using Timespy (free version) will show a chart of the clock speed over time.

B
BrickMe352
Member
106
05-01-2016, 03:19 AM
#8
I couldn't download timespy due to my slow internet connection. I used Cinebench to check if my results were within the margin of error. Yes, 3.9GHz is for one core, but I'm having trouble reaching that speed even with extra thermal support. In Cinebench, I didn’t achieve 3.9GHz either—I’m not sure if it’s related to Cinebench or just a new benchmarking experience. The results came from the links provided.
B
BrickMe352
05-01-2016, 03:19 AM #8

I couldn't download timespy due to my slow internet connection. I used Cinebench to check if my results were within the margin of error. Yes, 3.9GHz is for one core, but I'm having trouble reaching that speed even with extra thermal support. In Cinebench, I didn’t achieve 3.9GHz either—I’m not sure if it’s related to Cinebench or just a new benchmarking experience. The results came from the links provided.

C
Chester09
Senior Member
491
05-01-2016, 08:19 PM
#9
I thought it would be simple to locate, but it proved difficult since the reviews, such as from @ Guru3D, rely on an outdated Cinebench version. It seems the reported score reflects overclocking, as the best result on hwbot for a 2400G is 1031. Not very useful. The downside is that your outcome is still acceptable, meaning any issues are minor.
C
Chester09
05-01-2016, 08:19 PM #9

I thought it would be simple to locate, but it proved difficult since the reviews, such as from @ Guru3D, rely on an outdated Cinebench version. It seems the reported score reflects overclocking, as the best result on hwbot for a 2400G is 1031. Not very useful. The downside is that your outcome is still acceptable, meaning any issues are minor.

T
tonylaflem
Member
218
05-01-2016, 08:31 PM
#10
Here are your top performance numbers.
T
tonylaflem
05-01-2016, 08:31 PM #10

Here are your top performance numbers.

Pages (2): 1 2 Next