Discussion New computer parts as an enviromental disaster?

[Kocicak]

The current broad trend in all technology is moving toward increased efficiency in order to decrease energy consumption, which has various negative effects.

I just saw a rumour that an upcoming Intel CPU will have special performance mode allowing 350W energy draw. It also seems that upcoming AMD CPUs will have higher TDP than the current generation.

What is the sense of this in the point of view of the broad trend I mentioned in the beginning? State regulation in my opinion is sometimes a good thing and at this point it seems that power draw of consumer computers should be limited.

I think that limits of 50W for CPU, 50W for integrated GPU and 150W for discrete GPU would satisfy needs of most people.

Parts with higher power draw could be heavilly taxed so that they would be financially viable only for proffesionals who would use them productively.

I understand that some people may find my idea not beneficial for whatever reasons but what is happening now in this part of PC market is ridiculous and needs to stop.

(and production of virtual currencies should be banned completely as a huge energy waste, but that is a different topic.)

 

[DrMrLordX]

My 3900x system in total pulls around 170W @ idle. I win!

(that includes monitors, fans, pumps, the whole shebang)

 

[Kocicak]

I just spent some time measuring my two PCs, here are the results:



The efficiency of the two CPUs alone is very different, but when you add the power consumption of the rest of the PC, the differences get smaller. I believe that AMD system may have been disadvantaged by running 4 RAM sticks instead of two, but the Intel system runs 3 SSDs instead of 1, but they do not take part in the calculation.

 

[sdifox]

I am starting to feel that some of the posters here are from USA. Lack of effective state regulation made USA to have extremely poor or nonexistent public transportation systems and railways network, horribly high per capita energy consumption, problematic healthcare system with poor availability for some and overpriced for others, etc, etc, etc.

High per capita energy consumption and also greenhouse gas production is not just an internal problem of the USA, you are screwing things up for everybody on the planet.

USA serves as an example of disastruous impacts of a lack of state regulation and investment in many fields.

Did you somehow miss the fact that this forum is from USA?

 

[Spicedaddy]

US and Canada have 120V outlets x 15 amps = 1800W So that's probably where it'll stop.

Unless people start installing dryer/oven outlets for their computers.

 

[Kocicak]

I just spent some time measuring my two PCs, here are the results:

View attachment 67160

Looking at the numbers again, the difference performance per watt in single thread is HUGE.

I hope we cen get proper testing of this in reviews. Out of curiosity I would like to buy some AM5 system and see myself how it runs, but I have been burned twice recently buying brand new freshly introduced stuff. My Asus X570 board ended up not liking the 5700G at all and my Gigabyte Z690 board now freezes randomly with 12600K.

 

[igor_kavinski]

My Asus X570 board ended up not liking the 5700G at all and my Gigabyte Z690 board now freezes randomly with 12600K.

Seriously bad luck!

I don't envy you at all...

 

[Kocicak]

One way how to look at the efficiency is to observe time and energy used to compute a fixed task.

I found this graph in the Computerbase review:


From that I arrived at how much energy is needed to spare one second of the time of completing the task.



You can see that this CPU is allowed to run so far away from an efficient region so that it consumes up to QUADRUPLE of the energy needed to spare one second!

THIS IS RIDICULOUS AND NEEDS TO STOP !!!

 

[Markfw]

One way how to look at the efficiency is to observe time and energy used to compute a fixed task.

I found this graph in the Computerbase review:
View attachment 67229

From that I arrived at how much energy is needed to spare one second of the time of completing the task.

View attachment 67230

You can see that this CPU is allowed to run so far away from an efficient region so that it consumes up to QUADRUPLE of the energy needed to spare one second!

THIS IS RIDICULOUS AND NEEDS TO STOP !!!

Talk to Intel....

 

[Thunder 57]

One way how to look at the efficiency is to observe time and energy used to compute a fixed task.

I found this graph in the Computerbase review:
View attachment 67229

From that I arrived at how much energy is needed to spare one second of the time of completing the task.

View attachment 67230

You can see that this CPU is allowed to run so far away from an efficient region so that it consumes up to QUADRUPLE of the energy needed to spare one second!

THIS IS RIDICULOUS AND NEEDS TO STOP !!!

No one is going to tell me I can't drive a gas guzzler, even if I do not. Or a computer with a CPU that draws a ton of power. I live in an area that is warm year round. I like low TDP parts because of that. But, I very much believe in the the statement of "Live Free or Die".

 

[igor_kavinski]

This is what Intel does anytime they are faced with serious competition. Happened during the P4 days. Enjoyed their successes from Conroe to Cascade Lake. Now they are desperate to win again. It's like needing only a lighter to light a cigarette but instead, you figure you can make wayyyy more money if you sell people blowtorches to do the same.

 

[Kocicak]

Seeing what I did in the morning, it feels like a nice round trip from point A back to point very close to A, but it does not feel like a waste of time, walking is a good exercise and on the route you can look at things from a little bit different perspective...

 

[Markfw]

Seeing what I did in the morning, it feels like a nice round trip from point A back to point very close to A, but it does not feel like a waste of time, walking is a good exercise and on the route you can look at things from a little bit different perspective...

Posting things like having the government tell us what we should have and deciding our lives is not going to cut it with most of us here. If you leave the politics out of it, and then complain about power usage, then you have half the people saying "we want to use all that power, since we want the fastest", so now you have half the people hating you. Now trying to say "but x mfg has the the fastest and most efficient cpu", then you only have a small minority to fight with.

Just some advice to not be attacked by the whole forum here.

 

[sdifox]

US and Canada have 120V outlets x 15 amps = 1800W So that's probably where it'll stop.

Unless people start installing dryer/oven outlets for their computers.

My NEMA L6/30R installation begs to differ But this is actually more efficient than 120VAC.

https://www.amazon.ca/gp/product/B00002NAT9/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

Yes the pictures are old.

 

[Kocicak]

Kitguru tested performance per watt of the new Ryzen 7950X cpu.



at 215W - 179 points per watt
at 142W - 250 points per watt
at 88W - 335 points per watt

It is really sad seeing such a GREAT power efficient CPU being sold running so fast, that the performance per watt gets almost HALVED compared to moderate power usage setting and is hitting 95°C while doing that.

 

[In2Photos]

Kitguru tested performance per watt of the new Ryzen 7950X cpu.

View attachment 68144

at 215W - 179 points per watt
at 142W - 250 points per watt
at 88W - 335 points per watt

It is really sad seeing such a GREAT power efficient CPU being sold running so fast, that the performance per watt gets almost HALVED compared to moderate power usage setting and is hitting 95°C while doing that.

It literally takes 2 seconds to switch to ECO mode to achieve these results. Why is it such a problem to give the consumer options. If they want brute performance they have it. If they want performance with efficiency, they have it. If YOU don't want it, don't get it! It really is that simple!!!

 

[Kocicak]

... If they want brute performance they have it. ...

They? Everybody gets this since it is a default setting. I have no idea how many people will be asking about their coolers being mounted improperly hitting super high temperatures at default settings!

 

[In2Photos]

They? Everybody gets this since it is a default setting. I have no idea how many people will be asking about their coolers being mounted improperly hitting super high temperatures at default settings!

Are they going to be asking you if their coolers are mounted properly? Does it affect YOU?

If the consumer is familiar enough with software to check temperatures, they are familiar enough to use a Google search and decide if they want to enable ECO mode to lower said temps. Those temps also won't be reached with everyday tasks like internet browsing, streaming, and even gaming. It is only the most demanding tasks that would even hit those temps.

 

[DrMrLordX]

I mean let's be honest

If people want a 142W 7950X, all they have to do is put a ~150W HSF on it and call it a day. It'll hit 95C and dial back the volts/clocks. It will do better undervolted because, you know, reasons, but if you want simple, there you go.

 

[RTX]

If the OP cared that much about power draw, does the OP also plan on delidding and buying derbauer's delid tool + the direct die frame? Lower temps = lower voltage required to hit the same frequencies.

 

[Markfw]

I mean let's be honest

If people want a 142W 7950X, all they have to do is put a ~150W HSF on it and call it a day. It'll hit 95C and dial back the volts/clocks. It will do better undervolted because, you know, reasons, but if you want simple, there you go.

I plan on going 105 ECO mode.

 

[Kocicak]

This is finally the kind of output I was hoping to see, from a japanese blog.



It is visible that there is an optimal frequency for AMD 7950X with highest performance per watt (here it is the minimum of the curve) at around 2 GHz and that there is little point in pushing it beyond 5 GHz.

The reasonable frequency limit for Intel 12900K seems to be around 4700 MHz.

All this is valid for maximal multithreaded load.

I have not studied the blog post if any voltage optimisation took place, because that could change the curves.

It would be ideal if this sort of testing could be done for some "reasonable load" as 4-8 threads of Cinebench.

Seeing this picture we can debate what should be the limit where the processors should perform. Should we accept only 2 times worse performance per watt? Then the limit for 7950X would be cca 4300 MHz. At three times worse performance per watt the limit is 4800 MHz. At 4800 MHz 7950X consumes JUST around 100W according to another graph in the blog post.

Such a nice cool and quiet the system would be at that power draw and frequency. No wasteful, 95°C, "water cooling needed" redacted






No profanity allowed in the tech fprums.


esquared
Anandtech Forum Director

 

[Kocicak]

I just measured Intel 13900K. At low load the efficiency seems to be similar as that of 12600K, but at higher load it is noticeably less efficient.

 

[igor_kavinski]

I just measured Intel 13900K. At low load the efficiency seems to be similar as that of 12600K, but at higher load it is noticeably less efficient.

View attachment 69063

With which mobo and RAM kit? Can you do a Geekbench 5 / "Rapydmark High" run?

 

[Harry_Wild]

Depending on how the national media spins it; food that you eat can be an environmental disaster too!

 

[Kocicak]

I have put together, how performance and efficiency at different level of load (number of running computing threads), is affected when the CPU is limited by the total power it can draw.

The improvements in efficiency of the CPU while running heavy load are impressive. Unfortunatelly poor efficiency of the CPU remains poor at lower load.

Limiting power is useful to match available cooling solution or while running this CPU at maximal load to significantly improve its efficiency.

Unfortunatelly it cannot improve overall efficiency of the CPU, which most of the time will run at below maximal load.

A positive takeaway is that the performance of the CPU below maximal load is unaffected or just slightly affected by lowering the maximal energy draw.