Researchers uncover the fastest semiconductor yet
spectrum.ieee.org
spectrum.ieee.org
I call BS. It is rare, true. But it is found in reasonable concentrations in some minerals (alongside molybdenum) that are already being mined and we just need to learn to extract rhenium out of it. For high perf semiconductors you only need microscopic amounts of it anyway.
As to costs, majority of the costs are not in substrates anyway, however costly they are. The costs are in IP and in processing. I suspect even if silicon was as expensive as gold it would not meaningfully change the cost of current high-end CPUs that go into our phones.
(edited, I mistakenly wrote ruthenium where I meant rhenium)
Most chemical elements with atomic numbers equal or greater than that of lutetium are very slightly radioactive (with a few exceptions like iridium, gold, mercury and lead). Rhenium is one of the most radioactive of them, but it is several orders of magnitude less radioactive than thorium and uranium.
Nevertheless, its radioactivity is so weak that it can be handled safely with bare hands. Even so, making some jewelry kept permanently on the skin is unlikely to be a good idea.
Rhenium does have one stable isotope, which comprises 37,4% of naturally ocurring Rhenium. 62,6% is radioactive Rh-187 with a half-life of 10^10 years. That's a bit longer than U-235 or U-238.
But unlike uranium, Rh-187 decays into stable osmium.
I would worry more about radioactivity as a failure mode for these transistors. Given how fast these are and how few atoms are supposed to be there, the question is if a single atom in the lattice undergoing fission can
a) blow up the lattice enough to disrupt it
b) make it no longer perform the function because the atom that is there has different chemical properties
The yearly production per human is a barely visible grain with a volume of about a quarter of a cubic millimeter. If each human would want to have a ring of rhenium, that would need all the global production accumulated during one thousand years.
All of it is heavily contended for various applications like gas turbine blades, so the price is in the same range as for platinum-group elements, which are in fact much more abundant.
If the semiconductor layers would have thicknesses in the tens of nanometer range, i.e. they would be deposited on some crystallographically compatible substrate, then such a semiconductor might be usable for expensive devices.
Nevertheless, developing a deposition method that can recycle all the non-deposited rhenium may be not easy. The deposition methods normally used frequently deposit far more substance on the equipment walls than on the semiconductor wafer. When depositing pure inert metals like gold or platinum, there are relatively easy methods to recover them, e.g. by dissolving in acid everything else. Such a complex substance containing rhenium might be obtained from a chemical reaction during the deposition, but it may be difficult to find precursors that are also easy to recycle.
Then why does it only cost $1 per gram?
[1] https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-rhenium.pd...
The price of rhenium fluctuates wildly and a few years ago there was a maximum at $10 per gram.
It is unlikely that it will ever return to the COVID-19 price.
The problem with rhenium is not the current price, but the minuscule production, for which there is very little hope that it could be increased much.
If instead of being used in a few military and commercial airplanes and in a few industrial applications, like thermocouples or catalysts, there would be demand to use it in something used by everybody, the price would increase quickly.
Wait, above you said it's heavily contended...
> The price of rhenium fluctuates wildly and a few years ago there was a maximum at $10 per gram.
You lost sight of the simple fact I mentioned, you only need a very small amount of it and you add it to a CPU that costs hundreds of dollars already. So the impact on the cost of the CPU would be negligible.
Even at the highest cost of rhenium you have mentioned, even if you needed a full gram of it (which I don't believe, this is going to be more like milligrams or even micrograms), you still only change the cost of the CPU slightly, at most.
This before you account for inevitable improvements in obtaining the Rhenium itself.
It all depends on a lot of details that haven't yet been quantified here. And the recycling the GP talks about is a huge one.
People don't have understanding of what expensive is. Even at hundreds or thousands of dollars per gram it could be pretty much irrelevant to the end result cost if you only need a layer that is couple nanometers in width.
The entire industry is a game of "value-add". The blank silicon wafers start out with effectively zero (or negative) value. The raw materials are worthless to the business. Only once validated features (process layers) begin to accumulate do those raw materials begin to inherit some sense of value.
Imagine if even basic electrical systems needed wires made of rhodium and semiconductors only worked at cryogenic temperatures. We'd never have gotten over the "technological activation energy" in the 1800s.
IBM was the first in 1997 and then others followed.
https://en.wikipedia.org/wiki/Copper_interconnects
That's how technology is. Things start out hard and with a lot of work and time we invent new methods. Aluminum was once worth more than gold. Then we invented new processes of extraction and refinement.
https://clintonaluminum.com/aluminum-was-once-worth-more-tha...
I can't speak to cpus with thousands of pins, but with significantly smaller chips that often have N>1 vcc or ground, it's typically due to layout convenience. The multiple vcc may be on e.g. opposite sides of the chip, and that makes it easier to route a pure vcc signal to a spot that needs it. It's easier to route outside the chip than inside the chip, because there's more space.
How do you route that current in one pin to the two locations the two pins handled? You've just moved the point where the split happens from on the motherboard to on the chip package.
https://semiengineering.com/wirebond-technology-rolls-on/
But all the high performance chips in leading edge process nodes like 3nm are flip chips. The last time I worked on a wire bond chip was in 130nm in 2004.
With a wire bond chip you can only have IO for signals and power/ground around the periphery. Some wire bond chips have 2 rings of IO pads but it makes the wire bond angles complicated. It's difficult to jump over other wire bonds to get closer to the center.
https://en.wikipedia.org/wiki/Flip_chip
Flip chips have a series of bumps above the top layer of the chip. These bumps are then connected to a small PCB inside a package or some other kind of interposer.
The transistors are on the bottom of the die and then up to 18 layers of metal are built on top. In a wire bond chip the heat has to go up through all that metal stackup which is usually encased in a glob top which isn't great for heat transfer.
https://www.gluespec.com/blog/glob-top-encapsulation
In contrast a flip chip has the die mounted upside with the top layer mounted to the PCB and the side with transistors is on top and can be directly mounted to a heat sink. Intel and AMD used to have bare die around 2000 but then mounted heat spreaders on top because sometimes people would mount the heat sink incorrectly and crack the corner of the die when tightening down non-uniformly.
http://mantravlsi.blogspot.com/2014/10/flip-chip-and-wire-bo...
With a flip chip we can have over 15,000 IO in the chip. The flip chip bumps can be all over the die not just the periphery. Not only can we put IO in the center but the density of the bumps can be much higher compared to the pad points where a wire bonder would attach.
As for your original question about some kind of continuous bar shaped contact we have a power grid underneath on every layer to distribute the power across the chip. This has to go from the top layer Metal18 down through vias to the transistors below Metal1.
Modern chips have multiple voltages in multiple voltage domains. The DDR and PCIE sections have their own voltage requirements. The standard cells that are the combination logic within a CPU operate on much lower voltages. We have dynamic voltage control where the voltage is lowered to save power. We have voltage islands where the USB port can be shut off if nothing is plugged in or CPU core 1 is active while cores 2-4 are off saving power. This requires dedicated power / ground bumps and head switches to disconnect power to sections of the chip.
I don't think we could manufacture your concept of a "bar shaped contact" because the process DRC (Design Rule Check) stuff is very rigid about what can be manufactured. Certain shapes, widths, and turns decrease the yield so they aren't allowed.
You can see a diagram here of the layer stack up. You can see that the solder bump on the top that connects to the outside world is huge compared to the internal metal layers.
https://en.wikipedia.org/wiki/Back_end_of_line
A standard cell in 5nm is around 200nm in height. The width can vary but this way all the cells go together like Lego bricks. Each metal layer going up from Metal1 to the top gets wider and thicker. By the time you get up to Metal18 or so it is probably 20 times wider and thicker than M1. This is useful because those upper layer metals distribute the power and ground and a chip wide clocks. The large top layer metal is also required for the huge bumps above that connect to the outside world.
I think the current minimum bump pitch is around 130 microns. You can see these with your eye and don't need a microscope. Do the math on a chip that is 25mm by 25mm and a pitch of 130 microns and you can determine how many bumps you could fit on the chip.
https://en.wikipedia.org/wiki/Physical_design_(electronics)
This is the program I use everyday. It has a list price of over $1 million for a single license. My company has about 800 licenses but we probably get discounts of 60%
https://www.cadence.com/en_US/home/tools/digital-design-and-...
Power distribution via hydraulics instead?
I think the germanium point contact transistor was sold, but eventually got us via practical methods to the silicon chips we have now. We have had more expensive exotic materials like gallium arsenide.
So I think his comment in the article will happen:
“Now that we know what structural and electronic properties are needed to achieve the new transport regime discovered in this work, there is good likelihood that we will find earth-abundant alternatives to this rhenium-based material that also show impressive transport properties,” Delor says.
Have a gander at https://en.wikipedia.org/wiki/Health_hazards_in_semiconducto... for more info.
I feel like the only person I've heard talking about graphene and actually demonstrating any practical application is Robert Murray-Smith. This guy is like he fell down the popular science rabbit hole 20 years ago and emerged naked but for a towel screaming "Eureka".
I work with a company that has a couple of patents on a carbon nanotube based sensing system for hydrocarbons. This company is making and selling units every day, like this shit is real man.
These units are installed out in the field all around the world and are tied into a dashboard for customers to see if any of their projects are experiencing leaks.
I agree that the carbon nano-tech revolution that we were promised by pop-sci writers hasn't materialized but that doesn't mean that it isn't a genuinely useful substance.
I guess they were civil engineers, not mechanical engineers.
http://en.wikipedia.org/wiki/History_of_the_steam_engine#Pre...
I'm pretty sure we know they are the next asbestos. So we won't do things with them that tend to produce a bunch of dust for people to inhale.
That seems like quite the discovery.
Unfortunately the rarest of rare materials, but the above quote does point out that it shows us what’s possible even if it isn’t viable yet.
[1] https://en.m.wikipedia.org/wiki/Rhenium
[2] https://www.statista.com/statistics/1312513/rhenium-producti...
Most probably you'd see this tech in highly specialized interfacing circuits, like current GaAs chips.
I can also imagine Juniper using them in an ASIC and charging a ton.
The price of rhenium would skyrocket to million dollars per kg, and as long as production would stay the same (assuming it's limited by raw resource availability rather than just extraction methods) it would keep getting higher.
The current relatively low price of rhenium (relative to its rarity) is simply due to low demand.
Sure, assuming (1) we can actually produce a rhenium CPU, and (2) we are unable to increase the production of rhenium then, yes, the price of rhenium will increase.
But it will increase precisely because we have a working rhenium CPU in production in the first place, which is what the article disputes is possible (due to its current rarity).
Furthermore, the price would skyrocket only if people are actually willing to pay a high price for these rhenium CPUs, which again means they’re worth the money.
For stuff previously needed in low quantities it's usually the reverse - price goes down as more is needed. Initially prices are high because manufacturing equipment has to be maintained even if idle, wages have to be paid, and because of logisticical overhead for the small quanitites. A second price drop occurs as we get into mass-manufacturing and better processes are found.
Rhenium is a byproduct of mining and refinement, but I suspect it is often not captured because the small quanitites needed don't make it economically interesting - you would invest in infrastructure to extract it and immediately crash the price. That would change if there was a stable demand of higher quantities.
The 60g figure is for the packaging, not the chip. The actual chips are tiny and weight much less than that, probably under 1g for most processors.
This new semiconductor isn't pure rhenium, it's a compound, and hence less than 1g would be needed, or about $10 per chip, maximum.
Realistically, this new semiconductor would be deposited as an extremely thin layer on top of something cheaper like silicon or quartz. The material cost per chip would be measured in cents.
Depending on the kind of deposition method used, for depositing a certain amount on the wafer, a much greater quantity is used, which ends deposited on the equipment, or as chemical precursors mixed and reacted or unreacted.
Due to the rhenium cost and scarcity, all the rhenium compounds that are not deposited on the wafer must be recycled. That can raise the cost a lot.
Finding a compatible substrate for deposition, with an appropriate crystal structure, can be very difficult.
I had been thinking that viability of yield could be an issue too as it would have the same wafer fabrication yield, wafer sort yield, and packaging yield that silicon does. And as a new and profoundly expensive material there is going to be an appreciable learning curve.
Is that what you’re referring to? Or is there more to it even than that?
If desirable CPUs are made of Rhenium with an effective production line and that the raw supply is so tight, the price certainly won't stay at $10000/Kg, and probably will align with equivalent computation power/watt of traditional designs. Any investor will see that it will only make miner rich without that much return miles aways, and walk away.
No, your source clearly says that was its peak price in 2008/2009, after which it fell to ~$3000 in 2018, and in 2020 it was $1000/kg [1].
[1] https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-rhenium.pd...
It is unlikely that the rhenium price will ever return to that value.
The problem is that there are no chances of greatly increasing the production, so any new application will increase the price.
> “[…] they are not necessarily compatible with current hardware used in the semiconductor industry,” […] the applications for these semiconductors “would likely be different than those for traditional semiconductors.”
processing speeds in devices based on them could reach femtoseconds
> Although the new material is made using one of the rarest elements on Earth, the researchers suggest counterparts made from more abundant materials may be discovered that operate comparably fast.
you're probably reading this on a device that uses motherfucking hafnia for its dram dielectric
i mean admittedly hafnium is still a thousand times as abundant as rhenium
Something doesn't add up.
Rhenium price seems to be $2k/kg. Gold price is $62k/kg. Silver price is $0.8k/kg. So Rhenium is about 3 times as expensive as silver and 30 times cheaper than gold.
Instead they are extracted from ores in the tailings of large scale mining operations like coal or copper and refined by specialty processors.
It is not commercially viable to mine the rare earth elements on their own. Because the price is only $2k/kg, it makes more sense to build a business around all the coal (or whatever) that has to be moved to get a kg of a rare earth.
Because you still have to move all that dirt.
Basically, the price reflects only the costs of refining, marketing, and distribution. Extraction is sunk cost.
Based on what?
Manganese has significantly different properties (due to smaller atomic size and greater electronegativity), so just substituting it in the same chemical formula would not create the same crystal structure and any properties would be different.
Nevertheless, it is likely that other substances with similar properties will be found, but it remains to be seen if any of them are stable enough and cheap enough to be used in practical devices.