Liquid Reservoir

[britechguy]

which is crazy to think, when comparing CPUs that were built on a 130nm process and modern ones are down to 5nm now.

Among other things. I've never understood why earlier computer chips were so "fragile" when it came to heat as opposed to automotive ECUs that operate in MUCH harsher environments.

One of the last things to fail, most of the time, in cars is electronics (and I separate that from electrics - light bulbs in particularly, and wiring to a lesser extent - fail far, far more frequently than ECUs do).

I think we've reached that point with computers, because they now handle thermal management "all by themselves" to an extent that was utterly unthinkable when I first entered the field, and would have been nearly unthinkable even 15 years ago.

 

[frase]

Heat disperses from the CPU to other areas of the board as well. It comes down to the three principles which have advanced over the years Conduction, Convention and Radiation. There are more efficient methods to apply conduction now though the principle in the same as in specific heat pads on GFX cards memory banks for example. Convention in the way heatsinks are created to assist in dispertion of heat . I add heatsinks now to my 1980's Commodore 64 CPU, SID, PLA, VIC-II chips to dissipate heat, as these heat up a lot the same as Pi Devices. This is where it is critical to have the correct airflow setup in a system or one will have an oven. Radiation is in part the same as conduction and convection in part.

Users of LQ on the other hand go through a mind numbing process that to me I could not be bothered with. As I said these are the realms of overclockers who want to push their hardware to the limits without the system going into a singularity. If one is not an overclocker I simply do not see the point of LQ vs AC.

 

[Blues]

Not to mention that modern processors are designed to have operational temperatures that are entirely within normal limits that many old timers insist are "hot."

Yes, in absolute terms, if you touch 'em, they're hot. But when you look at the spec sheets for what Tmax is before any throttling occurs, they're nowhere close.

Times and technology change and evolve.

Yeah I remember I would set the BIOS temp threshold to 65C today I would have a system shutting down just from running anything that began to tax the system.

 

[ComputerDave]

And you shouldn't do what the client wants when it goes against your professional opinion, formed from years of experience.

We are hired for our expertise, and if a client rejects that, and it's on something we personally consider important, then the relationship is best terminated, promptly. I have yet to hear anyone happy after they've gone against their better judgment.

Well said. Thank you!

 

[britechguy]

Thank you!

You're welcome. We, as a culture, seem to have entirely forgotten that it is our personal right, in all cases, to decline to do something we don't want to do. It's OK to say, "No," and stick to it when saying yes is a violation of your personal morals, professional judgment, or similar. You do NOT have to agree because somebody else wants something.

Refusal often does come with consequences, but because people so seldom seem to do it, they also don't realize that just as often, it doesn't. You should not let yourself be bullied into doing anything that you know, in your heart of hearts, you will not like yourself for having done later. And that's even if that's the only negative outcome of having done it.

In the immortal words of Bartleby, the Scrivener: "I would prefer not to."