Cubic boron arsenide semi conductors

[Wirko]

Yeah I assume TSMC, Intel, Qualcomm, Micron, Samsung, all the major chip makers have an R&D department that probably have dozens more semi conductor candidates lined up.

That mix of companies probably looks a little different. Like TSMC, Intel, Samsung, IBM, imec (woohoo, a European company), Lam Research, Applied Materials. So, chip makers and manufacturing equipment makers, but not fabless chip designers. I also suspect other companies on this list got at least some patents from doing semiconductor research.

 

[KapiteinKoek007]

The cooler attached has little influence in the capacity of a chip to heat it.
New smaller nodes give off more heat IN the circuit simply because of smaller transistors being more tightly arranged in a ever smaller package, and usually significantly more transistors each generation.
Dropping the volts used might help, but it cannot fully mitigate the heat saturation issues that abound now.

There's just too much happening in too small a space.

wasnt talking about the heatsink of a cooler, but the integrated heat spreader of the cpu itself. that design hasnt changed alot on intel especially the last few years when they were pushing everything out that 14nm node. increased voltages on a chip that hasnt had any real architectual changes in years, same goes for IHS on the cpu, because really, a 11th gen cpu is basicly a broadwell on steriods. broadwell released in 2014, only some i3 the cannon lake for example had a 10nm node. the rest of the i5/i7/9 from 2014 up untill about 2020, no real architectual changes (ihs included) since the 5th gen broadwell release.

 

[TheoneandonlyMrK]

wasnt talking about the heatsink of a cooler, but the integrated heat spreader of the cpu itself. that design hasnt changed alot on intel especially the last few years when they were pushing everything out that 14nm node. increased voltages on a chip that hasnt had any real architectual changes in years, same goes for IHS on the cpu, because really, a 11th gen cpu is basicly a broadwell on steriods. broadwell released in 2014, only some i3 the cannon lake for example had a 10nm node. the rest of the i5/i7/9 from 2014 up untill about 2020, no real architectual changes (heatsink included) since the 5th gen broadwell release.

So, that's still a thermal dissipation device and is irrelevant with regards to Why modern chip's are hotter.
My point stands, more transistors, less space = more heat.

What your talking about might be true, no change but it's not responsible for the core heat issues.

Yes it might mitigate the issues but it didn't cause them.

 

[KapiteinKoek007]

So, that's still a thermal dissipation device and is irrelevant with regards to Why modern chip's are hotter.
My point stands, more transistors, less space = more heat.

What your talking about might be true, no change but it's not responsible for the core heat issues.

Yes it might mitigate the issues but it didn't cause them.

Thermal dissapation is completely relevant for the IHS. if they are originally designed for 4 cores at 4 ghz, and now they are at 8 cores 5.3ghz (11th gen i9) a big bump in voltages as well on basicly the same chip and IHS. thats why intel had to make the switch to 10nm for alder lake because they were seeing they already pushed everything out of this already outdated 14nm node
more transistors on the die means it requires less voltage to do the same job which translates to less heat but they kept adding more cores and kept upping the default clockspeed and voltages on a chip that wasnt originally designed for those speeds and number of cores.

 

[TheoneandonlyMrK]

Thermal dissapation is completely relevant for the IHS. if they are originally designed for 4 cores at 4 ghz, and now they are at 8 cores 5.3ghz (11th gen i9) a big bump in voltages as well on basicly the same chip and IHS. thats why intel had to make the switch to 10nm for alder lake because they were seeing they already pushed everything out of this already outdated 14nm node
more transistors on the die means it requires less voltage to do the same job which translates to less heat but they kept adding more cores and kept upping the default clockspeed and voltages on a chip that wasnt originally designed for those speeds and number of cores.

Your believe what you want, but nothing you said counters what I said so far.

The Ihs also has changed as has the socket since quad core century just not enough for you apparently but it still isn't the Cause.

Look to delidded exploits on water blocks, while temperature is reduced it's still hard to keep cool because the die is Denser with transistors, simple.

But like I said you believe what you want.

 

[Assimilator]

All of the current alternatives to silicon are currently of nothing more than scientific interest due to the fact that silicon has been around for so long, is so well understood, manufacturing and using it in microprocessors is such a massive industry, and it's still currently Good Enough. To transition to something like cubic boron arsenide, or carbon nanotubes, or something else will require an incredible paradigm shift in both microprocessor design and fabrication. Guaranteed there are research departments in universities and companies figuring all this out, but much like alternatives to lithium ion batteries, we won't see them anytime soon.

 

[Shrek]

I believe Germanium (which I seem to recall came before Silicon) is still of use for high speed

Germanium Can Take Transistors Where Silicon Can’t - IEEE Spectrum

 

[DeathtoGnomes]

I believe Germanium (which I seem to recall came before Silicon) is still of use for high speed

Germanium Can Take Transistors Where Silicon Can’t - IEEE Spectrum

New Germanium-Based Material Could Replace Silicon for Electronics

A half-century after it was supplanted by silicon, germanium may push silicon—and graphene—aside

spectrum.ieee.org

talks about better stability than it had prior to 1960.

 

[lexluthermiester]

New Germanium-Based Material Could Replace Silicon for Electronics

A half-century after it was supplanted by silicon, germanium may push silicon—and graphene—aside

spectrum.ieee.org

talks about better stability than it had prior to 1960.

The compound in question, germanane, is the actual material in question proposed to be a replacement for silicon. It looks promising, but there are challenges to production. For one, Germanium is not as common as Silicon which makes mining and manufacuring less than ideal.

 

[R-T-B]

That is a myth. Quantum computers do some work very fast, but they will never replace traditional computers. Even if someone comes up with a room temperature QBit IC, they still will not replace transistors. They don't work the same way.

That, and you can generate quantum hardened encryption (that quantum computers can't deal with) on a classical computer too.

 

[Shrek]

That, and you can generate quantum hardened encryption (that quantum computers can't deal with) on a classical computer too.

How so? Quantum encryption is commercially available but needs more than a classical computer.

But now I'm off topic.