MSI First with Motherboard BIOS that Supports Ryzen 9000 "Zen 5" Processors

[Steevo]

No, you're wrong. Your CPU is a resistive heater. A CPU that consumes 100W of power will produce 100W of heat.

Physics says I'm right, if All, 100%, of every watt sent into a CPU was converted into heat and no energy left it would be a resistive heater, however, the CPU writes to the RAM, Pcie bus and devices, hard disk with signals that leave so it cannot be 100% resistive.

 

[Chrispy_]

Physics says I'm right, if All, 100%, of every watt sent into a CPU was converted into heat and no energy left it would be a resistive heater, however, the CPU writes to the RAM, Pcie bus and devices, hard disk with signals that leave so it cannot be 100% resistive.

The RAM is powered by the CPU's IMC. so the RAM power draw comes directly from the CPU. Yes it's true that some of that heat is dissipated by resistance of the gates in the RAM modules themselves, but the source of electricity to create that heat is the CPU's IMC which is pushing the 1.1V at around 2A to the DIMM slots.

As for the other devices, those are powered separately by the motherboard and the signalling between them is state-monitoring. The signals don't really "leave" because current barely flows. Voltage can be measured across a very large input resistance, so using P=V^2/R, you get this:

Power use (of voltage signalling) in Watts = ~1.25V/10KΩ,
So about 0.0001W

People like to use water pipe analogies for electrical power through wires, and in this case, the signalling pipe is sealed at the receiving end with a pressure dial. Because the pipe is completely sealed, no fluid can flow, which means no power can be transmitted, but the pressure of the water can be read by the receiver on the pressure dial.

 

[Guwapo77]

MSI first proves once again that MSI has the best support, and probably the best boards out of them all there.
Maybe ASRock is also good, the others no - asus, gigabyte...

My next board will be MSI, I haven't had one of their boards in almost two decades. They use to suck bad, but it looks like they really have turned things around. I made a build for someone else and loved their BIOS. Additionally, everything just worked, it was a blessing. I was a diehard Asus motherboard user, but I've had more issues with their MB than anyone else. Gigabyte has been ok overall, just don't like their BIOS.

 

[sLowEnd]

Physics says I'm right, if All, 100%, of every watt sent into a CPU was converted into heat and no energy left it would be a resistive heater, however, the CPU writes to the RAM, Pcie bus and devices, hard disk with signals that leave so it cannot be 100% resistive.

That's the thing. If you think you are right, then you have a fundamental misunderstanding of the laws of thermodynamics. All of the energy used by your CPU ends up as heat.

This should answer most or all of your questions

UCSB Science Line

It is also demonstrable in practical experimentation

Gaming PC vs. Space Heater Efficiency

Winter is coming, and with the power bills stacking up it may be hard to justify gaming on your ulta-powerful gaming rig. But what if we were to tell you that, watt for watt, your gaming PC produces exactly the same amount of heat as a space heater?

www.pugetsystems.com

 

[Steevo]

Yes, all of the energy used by the CPU ends up as heat, just not all of it is dissipated by the CPU die and thus the TDP is lower than the total power consumed by some amount. Some is dissipated by the RAM, 28 lanes of PCIe https://www.synopsys.com/designware-ip/technical-bulletin/reduce-power-consumption.html newer PCIe5 is around 120mW per lane 3.5W not a lot, but still not part of the CPU's actual TDP, and truly driven by the CPU directly into the restive load of the GPU/NVME/Chipset. Combine all the connections and you will find that 5-15% of the power is NOT part of the TDP.

My 7800X3D shows 30ish W during RDR2 and if 3W of that is just for communicating with the GPU & NVME there is 10% of the power not as TDP. Its not absurd to think that with a 120W TDP it might show power consumption of 130 or 140W but only be dissipating 120W of heat.


EDIT** You are showing that a whole computer, NOT just the CPU, and they even mentioned that its the GPU that consumes the most power.

 

[A Computer Guy]

Now I think it's going to be worth migrating to AM5; the difference is going to be too significant to ignore.

I've been diligently holding back to save for a 9950x. Hopefully in 2ys time it will be available on discount around $550.

I really don't think so. AM5 offers lower performance per watt.

View attachment 343766

AMD Ryzen 9 5900X Benchmark

AMD Ryzen 7 7700X Benchmark

This is accounting for multi-core output (16 vs 24 threads). Gaming performance shows a completely different picture.

lol

I think that guy who wants to build a 10 year PC is gonna find that parts become obsolete what feels much quicker than before.

Don't forget If you never use those faster parts you will never feel it. In an age of 16+ thread CPU's and 32+ GB ram it's getting hard to justify more frequent upgrades outside of specific requirements and keeping up with high end gaming.

 

[AusWolf]

Yes, all of the energy used by the CPU ends up as heat, just not all of it is dissipated by the CPU die and thus the TDP is lower than the total power consumed by some amount. Some is dissipated by the RAM, 28 lanes of PCIe https://www.synopsys.com/designware-ip/technical-bulletin/reduce-power-consumption.html newer PCIe5 is around 120mW per lane 3.5W not a lot, but still not part of the CPU's actual TDP, and truly driven by the CPU directly into the restive load of the GPU/NVME/Chipset. Combine all the connections and you will find that 5-15% of the power is NOT part of the TDP.

My 7800X3D shows 30ish W during RDR2 and if 3W of that is just for communicating with the GPU & NVME there is 10% of the power not as TDP. Its not absurd to think that with a 120W TDP it might show power consumption of 130 or 140W but only be dissipating 120W of heat.


EDIT** You are showing that a whole computer, NOT just the CPU, and they even mentioned that its the GPU that consumes the most power.

As far as I understand, you have to dissipate all of your heat to stay at a constant temperature. If your CPU consumes 100 W, but it only dissipates 90, then its temperature will keep rising. I'm not an expert on thermodynamics, so someone tell me if I'm wrong.

The 7800X3D is a special case, though, as it only consumes 80-90 W in a full load. The 120 W TDP is just a recommendation for cooling.

 

[Steevo]

As far as I understand, you have to dissipate all of your heat to stay at a constant temperature. If your CPU consumes 100 W, but it only dissipates 90, then its temperature will keep rising. I'm not an expert on thermodynamics, so someone tell me if I'm wrong.

The 7800X3D is a special case, though, as it only consumes 80-90 W in a full load. The 120 W TDP is just a recommendation for cooling.

So a light switch powering a 100W light also emits 100W of heat???


I'm saying the CPU transmits power to other components, just like a light switch and these loads through the CPU don't contribute 1:1 to the heat load.

 

[A Computer Guy]

So a light switch powering a 100W light also emits 100W of heat???

A light switch normally does not consume 100W however 100W may be conducted through a light switch. However if that light switch has poor connectivity in the circuit creating excessive resistance or arcing then it will heat up consuming power along the line that it is conducting.

I'm saying the CPU transmits power to other components, just like a light switch and these loads through the CPU don't contribute 1:1 to the heat load.

After some reading I'm inclined to believe the following per the conversation. Someone correct me if I'm wrong.

If your CPU is consuming 100 watts of heat it will output 100 watts of heat. Most of that heat will be radiated from the CPU to the heat sink while some of the heat will be conducted through other materials connected to the CPU.

Because the CPU isn't consuming 100% of the power you give it other signals connected to the CPU continue working. (obviously)

 

[Steevo]

A light switch normally does not consume 100W however 100W may be conducted through a light switch. However if that light switch has poor connectivity in the circuit creating excessive resistance or arcing then it will heat up consuming power along the line that it is conducting.

After some reading I'm inclined to believe the following per the conversation. Someone correct me if I'm wrong.

If your CPU is consuming 100 watts of heat it will output 100 watts of heat. Most of that heat will be radiated from the CPU to the heat sink while some of the heat will be conducted through other materials connected to the CPU.

Because the CPU isn't consuming 100% of the power you give it other signals connected to the CPU continue working. (obviously)

Yes, and if the CPU power measurement includes all its direct connections and the software reporting its power can't differentiate or hardware isn't capable of reporting its included on the total power for the CPU.