Can you build a sleeper PC with almost NO cooling?

[Euchre]

18 hours ago, Luscious said:

Long gone are the days of passive CPU cooling

I don't agree in the sense that it wouldn't be impossible to create a fanless and waterless cooling solution, just very much harder. With a heat pipe or vapor chamber setup, you could create one massive enough to transfer away heat to a very large sink with a lot of surface area for dissipation. In fact, I'd love to see Linus attempt exactly that - make a completely fanless and waterless PC using normal components, not laptop or ITX embedded stuff.

[Intergalacticbits]

22 hours ago, Spotty said:

Check out part 1 of the pyramid PC. 

I already watched that one.

(janky) 

 

I probably shouldn't have said that but it did look like two sloppy teenagers had made it.  

No offense intended.

To be fair though it was an unusual project.

It is just my impression that men who know how to use machine shop equipment can usually fabricate stuff way better.

Maybe Linus should consider hiring someone who lives up there in Canada who is good with his hands in that way.

I'll bet the projects would look tight and super cool then.

[Spotty]

13 minutes ago, Intergalacticbits said:

I already watched that one.

(janky) 

 

I probably shouldn't have said that but it did look like two sloppy teenagers had made it.  

I'm not going to tell Alex you said that...

 

It's not finished yet. Stay tuned for part 2. Though there is more jankyness in part 2...

[SolarNova]

6 minutes ago, Spotty said:

I'm not going to tell Alex you said that...

 

It's not finished yet. Stay tuned for part 2. Though there is more jankyness in part 2...

Lol.

 

Im sure he was implying 'experience' ..rather than 'know how'. Alex is relatively young in that regard, im sure he has the theory, but he doesnt have like 20 years experience in machining

Doing a good job none the less. cant expect any more, if LTT lasts another 10 years and Alex stays with them, im sure we'll see better stuff in the years to come.

[Intergalacticbits]

11 minutes ago, Spotty said:

I'm not going to tell Alex you said that...

 

It's not finished yet. Stay tuned for part 2. Though there is more jankyness in part 2...

 

Thanks lol.

I enjoy LTT and build videos so I will definitely watch it.

[pd.]

3 hours ago, Euchre said:

I don't agree in the sense that it wouldn't be impossible to create a fanless and waterless cooling solution, just very much harder. With a heat pipe or vapor chamber setup, you could create one massive enough to transfer away heat to a very large sink with a lot of surface area for dissipation. In fact, I'd love to see Linus attempt exactly that - make a completely fanless and waterless PC using normal components, not laptop or ITX embedded stuff.

If memory serves me right it was already done and that thing was massive. IMO it is much better to use the most power efficient components, downvolt and then allow them to turbo up to certain radiator temp.

edit:

 

[Luscious]

6 hours ago, Euchre said:

I don't agree in the sense that it wouldn't be impossible to create a fanless and waterless cooling solution, just very much harder. With a heat pipe or vapor chamber setup, you could create one massive enough to transfer away heat to a very large sink with a lot of surface area for dissipation. In fact, I'd love to see Linus attempt exactly that - make a completely fanless and waterless PC using normal components, not laptop or ITX embedded stuff.

That's been tried before. The size of the heatsink would need to be MASSIVE to account for the surface area needed to transfer that much heat passively and not involve any airflow. Even for a system with just a single CPU/GPU it would be unrealistic and impractical. Keep in mind as well that vapor chambers have a limit to how large/long they can be before they too become inefficient.

[randyval]

my sleeper

 

[ColinLTT]

On 10/24/2020 at 1:36 PM, SolarNova said:

Hold up.

 

Did u really put 5100mhz RAM on a Ryzen system ?

cont....

Yep, we did. We were limited to 4666mhz by the 3900X, we actually could have used a 3600X and gotten much better results on that overclock due to the actual physical location of the CCX dies on the chip, but this is a sleeper after all, so more cores and more thermals to deal with seemed like the right move. We did NOT give this system the best chance at working but in the end, it did pretty well. 

 

On 10/25/2020 at 3:12 AM, idb said:

Not a great job in this video, no graphs, couldn't recognise an ISA socket, plenty of mistakes and ... (cont) 

Sorry I cant recognize a socket from a standard that was made literally a decade before I was born? https://en.wikipedia.org/wiki/Industry_Standard_Architecture