Load peaks and capping
Let’s first look at the flowing currents at the 12VHPWR. I deliberately chose a distribution to four single 6+2 pin cables on the adapter here, because it still shows some peculiarity in the interaction with the power supply. First, let’s take a look at the flowing currents during gaming and just under 440 watts of power consumption. Measurements were taken at coarser 20-ms intervals, i.e. around 50 times per second, to simulate the load on the supervisor chip of the power supplies (shutdown). We see that ALL load peaks are brutally capped at 38 to 39 A at the latest. This puts you pretty much at 456 to 468 watts. So that’s all the power supply “sees”! This is exemplary and actually also completely harmless.
Nevertheless, we still need to take a look at the voltages, or the product of voltage and current flow. I already wrote that I measured at three different power supply connections, even though all three connections meet again on the graphics card’s PCB (or in the adapter). What we can now see here as much clearer fluctuations and peaks is due to the partially overvolting power supply and thus the voltage and not the currents. This is due to technical reasons, but not a big deal. However, we also see that the one peak at around 550 watts is not due to the current flowing (graphics card), but results from the power supply! Of course, reviewers should also take this into account.
In the Torture test, things look even better on the graphics card’s side. The card mercilessly caps the currents at around 35 amps already. Apart from the drops, which are harmless and serve to protect the card, there are no noticeable features.
If we now add the voltage again, we see a stronger ripple (m) which again results from the somewhat jittery operating voltage. However, in the power supply’s honor, it has to be said that this affects all current products of all manufacturers and is certainly hardly avoidable.
Because I would like to know it even more exact once more, I resolve the whole thing once more still more highly and take 20 ms as total running time. The intervals of 10 microseconds can just be measured sensibly and we also see the voltage here as a pink curve, whose average value is 12 volts, but which nevertheless alternates somewhat within the permissible range.
If you then convert that to the power consumption in watts, you get this picture:
I also did the whole thing again for the Torture loop, where we get to admire the regular drops. First again the currents:
And then again total wattage:
Power supply recommendation
Now we come to the point that completely reduces the expected sensation of exploding power supplies to absurdity.
Even IF you hopelessly overpower the card, no one really needs ATX 3.0 power supplies over 1000 watts unless the CPU eats more than 300 watts. This is really just a job creation measure for the struggling power supply industry and only satisfies the sick imagination of some standardization fetishists. You really have to put it so harshly. Well, the card doesn’t draw 600 watts at 12VHPWR even at full load, but almost. I could also only achieve this load with Furmark, so the transients are only very minimal here because there are also hardly any load changes. So you should always stay below 1000 watts even together with the CPU.
This is also the reason for my power supply recommendation, which for the 450-Watt models of the GeForce RTX 4090 is that a modern 850-Watt Gold or Platinum power supply should be sufficient. With a full power limit of 600 watts and Furmark, a 1000 watt model is also sufficient if the power supply isn’t absolute junk. Even though my list is short and I couldn’t test each power supply for more than 1 hour – I simply ran through what was still unassembled on the shelf upstairs in the lab and lasted.
be quiet! |
Up to 450 watts TBP: From 450 watts TBP: |
Seasonic | Prime 1300 Watt Titanium |
Corsair |
Up to 450 Watt TBP: From 450 Watt TBP: |
- 1 - Introduction, technical data and technology
- 2 - Test system in igor'sLAB MIFCOM-PC
- 3 - Teardown: PCB, components and cooler
- 4 - Gaming Performance WQHD (2560 x 1440 Pixels)
- 5 - Gaming Performance UHD (3840 x 2160 Pixels)
- 6 - Gaming Performance UHD + DLSS/FSR/XeSS (3840 x 2160 Pixels)
- 7 - DLSS 3.0 and the longest bars
- 8 - NVIDIA Reflex and Latency
- 9 - Workstation graphics and rendering
- 10 - Power consumption and load sharing
- 11 - Load peaks, capping and power supply recommendation
- 12 - Temperatures, clock rate, OC, fans and noise
- 13 - Summary and Conclusion
545 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Veteran
Urgestein
Veteran
Urgestein
Mitglied
Veteran
Urgestein
Veteran
Urgestein
Urgestein
Mitglied
Mitglied
Veteran
Alle Kommentare lesen unter igor´sLAB Community →