High Wattage CPUs: Engineering Plausibility?

[Arqane]

I guess I will just explain the whole concept instead of dodging it, as it should help explain things much more clearly. I am looking to build a system to put processors into water heaters and use the output heat to do the job a normal heating element would. Water heaters use about 20% of the total electricity usage per month in homes. I am putting together a sustainable apartment business, and even with a single 400 unit apartment complex, that is a *lot* of computational power coming from 8,000 processors using the electricity usually just burned off every month.

Why do I want "inefficient" processors? Because nobody wants their water heater heating up too slowly. 4.5kW is the goal, but with current processors that's hard to fit in a 40-gallon tank effectively. Some people use lower power heating elements, but that would be a pretty big negative for most people.

Why is cost a factor? Because the initial cost of this system is by far the largest part (along with some maintenance). Other than that initial cost, it's basically a data center with $0 cost for electricity and $0 cost for cooling. The electricity is already being used for the water heater's original purpose and the water cycling through takes care of the cooling.

 

[berkeman]

Why do I want "inefficient" processors?

I'm skeptical about your overall concept, but one tip would be to think of using extra background processes on the CPUs (that aren't needed) to increase the power consumption, instead of thinking of finding "inefficient" processors. You can run them at full speed when it's not needed and execute loops that use all of the math processor, etc. when it's not really needed.

 

[russ_watters]

The overall concept is pretty sound (not without complications though), but the idea of using inefficient processors on purpose still isn't.

...but one tip would be to think of using extra background processes on the CPUs (that aren't needed) to increase the power consumption, instead of thinking of finding "inefficient" processors. You can run them at full speed when it's not needed and execute loops that use all of the math processor, etc. when it's not really needed.

Yes, I see no fundamental problems with that. From an economic standpoint though, the duty cycle of a water heater is terrible for this. You'd need to install your 5kW of processors and run them at an average of probably 200W or so. It's a lot of money to spend on processors that most of the time aren't running.

The OP gets 2/3 of the equation right: you want high wattage and cheap. But there's no need to specify low efficiency. You just take what you get for where cost and wattage intersect.

 

[Arqane]

The overall concept is pretty sound (not without complications though), but the idea of using inefficient processors on purpose still isn't.

Well the thought process is that I simply wanted higher power processors in the same space. 20 i9s, or especially 30+ i5s or below (to make it cheap) just doesn't fit in the space. They don't have to be inefficient. But at the same time, if I could find a cheap generic company that skipped corners and used substandard materials with a bit more resistance, but could still pull off a decent amount of efficiency compared to current processors, then that would be fine for this purpose.

I mean, it would be nice if these were the same size and just as replaceable as heating elements are in water heaters currently. But I was looking if a middle ground was possible: a bunch of cheap processors put in there as the heating element that could produce enough heat to work just as well as a water heater. It doesn't make much sense as a one-off for personal use in your own home. But when you have thousands of apartment units being powered by solar and you could network them together and use it for applications like bit mining or running physics/weather calculations, it makes a lot more sense. In fact, at the current rate that data is being rented out, it would actually have a very sizeable impact on revenue even compared to rent itself. Besides also being very environmentally friendly.

 

[russ_watters]

Well the thought process is that I simply wanted higher power processors in the same space. 20 i9s, or especially 30+ i5s or below (to make it cheap) just doesn't fit in the space. They don't have to be inefficient.

Yep. Looking back at your OP, most of what you described makes sense, it's just the scenario with the i9's that went sideways.

 

[Arqane]

Yep. Looking back at your OP, most of what you described makes sense, it's just the scenario with the i9's that went sideways.

Yes, well, I didn't want to give out my whole master plan, even though it's a fairly specific niche. But besides the sustainability benefit, it really does seem like a data center's dream come true. Most of their costs end up being the electric bill to power it all plus cooling it all, as well.

 

[Averagesupernova]

The issue here is attempting to make the processors less efficient. The electricity it takes to give off the extra heat is the same amount it takes to heat water the conventional way. Don't get me wrong, there is nothing wrong with heat recovery. But the best scenario is having so little heat to recover that it's pointless to utilize it.
-
The thing you can look at if you want to recover some heat is having a way to preheat the water going into a conventional water heater. The net energy used to heat the water to 120°F by the conventional heater will be less because it is starting with warmer water.

 

[berkeman]

But at the same time, if I could find a cheap generic company that skipped corners and used substandard materials with a bit more resistance, but could still pull off a decent amount of efficiency compared to current processors, then that would be fine for this purpose.

Sorry to be blunt, but that is laughable. Any company that produced a sub-standard IC/CPU would be out of business by now.

 

[DaveE]

I guess I will just explain the whole concept instead of dodging it, as it should help explain things much more clearly.

That would have been a really good idea initially. We aren't going to steal your IP. Others probably already have. We also aren't going to give you good answers if we don't know what the #$$ your asking about. Next time be more polite by not wasting your readers time.

If your idea works out really well, the big money isn't in running your apartment building better. It's in selling your appliance to everyone else that has apartments like yours.

Still I don't think you've really invested enough time working through the details. It's not a bad initial concept, but as others have pointed out, there are many parts of it that just won't work out in practice. Yet there are simple hybrid approaches using existing, relatively inexpensive components that can do what you want.

Try not to let the perfect solution preclude a good enough solution. Try to focus on your fundamental performance or customer requirements (including cost, reliability, serviceability, etc.) and not just one technological solution. Steve Jobs didn't get successful by focusing on building better hardware.

 

[Averagesupernova]

Try not to let the perfect solution preclude a good enough solution.

Very good advice. Yet I find myself going down that path at times.

 

[Arqane]

Try not to let the perfect solution preclude a good enough solution. Try to focus on your fundamental performance or customer requirements (including cost, reliability, serviceability, etc.) and not just one technological solution. Steve Jobs didn't get successful by focusing on building better hardware.

This was just one idea to use the very large amount of power the apartments would be producing. Since they're made to cover 100% of the usage in the apartments, they actually have to cover a bit extra to make sure the total used doesn't go above what is produced. I've been asking around, and figured this was a good place to check the viability of the system. I wasn't really looking for a perfect solution, just a workable one. The space requirement was the biggest hurdle. And since they would only be running 2-3 hours a day each, it does make the initial cost quite high for how much they'll be used. But on the other hand, with a 400-unit complex using these, it would save about $17,000/month worth of electricity (300 kWh/water heater, 400 water heaters, and $0.13/kWh) from what a data center would use just to power the same thing, and that doesn't count cooling.

It's just all about whether you can even fit the processors in the water heater in the first place, and if the initial cost for the CPU/heaters would be made up by the amount the processing could be sold for. If those two don't work, then it's not currently useful.

 

[Tom.G]

Please keep in mind that the duty cycle of hot water heaters is quite low. That's why there is a thermostat on them.

People want fast recovery after half the family takes a shower. But if you keep the 4.5kW heater on continuously either the over-temperature valve opens and dumps the hot water or the tank explodes; either way your water bill goes up.

Taking that in to account, please consider how much compute power you will really have available.

Cheers,
Tom

 

[Arqane]

Please keep in mind that the duty cycle of hot water heaters is quite low. That's why there is a thermostat on them.

People want fast recovery after half the family takes a shower. But if you keep the 4.5kW heater on continuously either the over-temperature valve opens and dumps the hot water or the tank explodes; either way your water bill goes up.

Taking that in to account, please consider how much compute power you will really have available.

Cheers,
Tom

Yes, I did post somewhere in this fairly long thread that the elements are usually only on for 2-3 hours per day in an average household. Very different style compared to a normal data center, but it would have its perks. But that was definitely a cost/value proposition I considered.

 

[f95toli]

This is already being done, but at a bigger scale where it most likely makes more economic sense.
There are several examples of waste heat from datacentres being used as input to district heating systems.
I am pretty sure there are some examples in Denmark and Sweden (I know of at least one Facebook/Meta datacentre) and I would be surprised if there aren't quite a few more.

Using "waste heat" in various forms to heat homes is certainly not a new idea; it much easier to implement in towns with district heating than in individual homes.

 

[phinds]

I have it on purely anecdotal authority, combined with an elderly memory, that there was a big, and successful, project to do bitcoin mining in Siberia with a non-trivial secondary income from the excess heat feeding a large apartment house's central heating system.

 

[Vanadium 50]

First, the IP idea "I won't tell you the details lest you guys steal it" is in tension with the idea "you guys tell me how to make this work, and then I'll commercialize it."

Next, everybody who says resistors are cheap and effective at heating, and CPUs are expensive and less efficient is correct. Further, there are two extremes: at one end you have as efficient a CPU as possible and at the other, only a resistor. If the system (costs, prices adn income) is linear, the optimal point is one or the other.

 

[Arqane]

First, the IP idea "I won't tell you the details lest you guys steal it" is in tension with the idea "you guys tell me how to make this work, and then I'll commercialize it."

Next, everybody who says resistors are cheap and effective at heating, and CPUs are expensive and less efficient is correct. Further, there are two extremes: at one end you have as efficient a CPU as possible and at the other, only a resistor. If the system (costs, prices adn income) is linear, the optimal point is one or the other.

I wasn't asking how to make it work. I was just asking if some of the basics of it broke the laws of physics, pretty much.

 

[phinds]

I wasn't asking how to make it work. I was just asking if some of the basics of it broke the laws of physics, pretty much.

It doesn't break the laws of physics, it just breaks common sense.

 

[Averagesupernova]

I will go a step farther and say your idea breaks laws of physics. You seem to believe that the heat generated in an inefficient processor somehow comes easier than heat coming off of a resistance. If you did not believe this to be true, then why the desire to find processors that produce more heat?

 

[DaveE]

I wasn't asking how to make it work. I was just asking if some of the basics of it broke the laws of physics, pretty much.

It doesn't break the laws of physics, it just breaks common sense.

As well as capitalist economics. High performance ICs are really, really expensive to develop, and only a few companies can do it. If you are the only one that wants to buy an intentionally inefficient uP, then they can't make it for you. Seriously, that might be a $1M dollar IC.

Plus, they probably don't even know how to do it. No semi manufacturer studies and develops the IP to make bad products. If they did do it, I'd bet they'd just use their standard cells and add a big resistor on the side. Which, as we have suggested, you can do yourself for much less money.

 

[Vanadium 50]

Plus, they probably don't even know how to do it.

Sure they do. They'd add some resistors to the chip, same as we've been saying.

 

[russ_watters]

I'm wondering if OP has looked up a bunch of processors to see what the cost per watt can be? That should tell the story on the economics.

 

[phinds]

@russ_watters it seems to me that this horse has been beaten to death over and over. Isn't it time to end this?

 

[berkeman]

What am I, chopped_liver? Sheesh.

Thread is done. A warm celebration of life for @Flika is being arranged...