World's Largest Chip Foundry TSMC Sets 2050 Deadline To Go Carbon Neutral (arstechnica.com) 59
An anonymous reader quotes a report from Ars Technica: TSMC announced last week that it would flatten its emissions growth by 2025 and reach net-zero carbon by 2050. That'll be a tall order for a company that produced over 15 million tons of carbon pollution last year across the entire scope of its operations, about the same as the country of Ghana. Though the amount of carbon pollution per wafer produced by TSMC has declined in recent years, surging demand for semiconductors has driven overall emissions up, and years of rising energy use, likely from the introduction of EUV lithography, has slowed progress. "TSMC is deeply aware that climate change has a severe impact on the environment and humanity. As a world-leading semiconductor company, TSMC must shoulder its corporate responsibility to face the challenge of climate change," Chairman Mark Liu said in a statement.
Semiconductor manufacturing is both energy intensive and heavily reliant on potent greenhouse gases. Fabs use enormous amounts of power -- TSMC estimates that its 3 nm fabs will draw 7.7 billion kWh annually, or about the same as 723,000 American households. The company says that today about 62 percent of its total emissions come from energy. To achieve the net-zero goal, TSMC will have to work hand in hand with the Taiwanese government. The company uses 4.8 percent of the island's power today, a figure that's expected to rise to 7.2 percent by 2022 when the 3 nm fabs are turned on, according to Greenpeace Taiwan. Currently, Taiwan is heavily reliant on coal and natural gas, with less than 20 percent of electricity produced by nuclear and renewables. TSMC has started taking steps to address the issue, saying in July it was buying all of the power produced over 20 years by a 920 MW wind farm being built in the Taiwan Strait. It's expected to be completed by 2026. While significant, the company will have to make many more deals like that to eliminate its energy-related emissions.
Semiconductor manufacturing is both energy intensive and heavily reliant on potent greenhouse gases. Fabs use enormous amounts of power -- TSMC estimates that its 3 nm fabs will draw 7.7 billion kWh annually, or about the same as 723,000 American households. The company says that today about 62 percent of its total emissions come from energy. To achieve the net-zero goal, TSMC will have to work hand in hand with the Taiwanese government. The company uses 4.8 percent of the island's power today, a figure that's expected to rise to 7.2 percent by 2022 when the 3 nm fabs are turned on, according to Greenpeace Taiwan. Currently, Taiwan is heavily reliant on coal and natural gas, with less than 20 percent of electricity produced by nuclear and renewables. TSMC has started taking steps to address the issue, saying in July it was buying all of the power produced over 20 years by a 920 MW wind farm being built in the Taiwan Strait. It's expected to be completed by 2026. While significant, the company will have to make many more deals like that to eliminate its energy-related emissions.
To put that in perspective... (Score:2, Funny)
Kind of sounds like John Hammond insisting that Lex and Tim handle all the deaths on Isla Nublar once they're grown adults.
Beating swords into power plants. (Score:2)
TSMC has started taking steps to address the issue, saying in July it was buying all of the power produced over 20 years by a 920 MW wind farm being built in the Taiwan Strait. It's expected to be completed by 2026. While significant, the company will have to make many more deals like that to eliminate its energy-related emissions.
Just tap into all the warships hanging around the neighborhood. That'll solve that problem.
Re: (Score:2)
The company uses 4.8 percent of the island's power today, a figure that's expected to rise to 7.2 percent by 2022 when the 3 nm fabs are turned on, according to Greenpeace Taiwan
I especially loved the part where "Greenpeace" was the most authoritative source they could find for the raw engineering data.
Re: (Score:2)
And what is your problem with that - Greenpeace?
Boutique solutions are silly (Score:4, Interesting)
It is silly for each company to implement their own little boutique solutions to their carbon emissions.
It makes far more sense to implement large-scale solutions that reduce emissions on a national or even global basis.
Taiwan gets 45% of its electricity from imported coal. Offshore wind is likely the best solution, but even switching to gas would be a big win.
TSMC would do much more good by making their chips run cooler than by making the manufacturing use less energy.
Re: (Score:1)
Re: (Score:2)
One things annoys me about this promise though is that it is too far in the future. The people responsible for implementing it are probably dead by then and it might be too late too little anyway.
If not dead in 20 years then certainly retired. It will not be too little too late by then because the people being honest about the rate global warming is happening know that this is a very slow process, and that "too late" is impossible to define. Global warming sucks for a lot of people because they live where the climate is very suitable for human civilization. There will be places on Earth that benefit from global warming but few people live there, and these places are sparsely populated because it
Re: (Score:2)
Taiwan used to have a nuclear power program. Politics got in the way though and the government in power does not believe in building infrastructure.
Re: (Score:1)
Nuclear power is an existential threat to Taiwan. The island sits on fault lines and if it had a disaster similar to Fukushima it would displace huge numbers of people and absolutely destroy their economy. The price of fixing it and the loss of that electricity would be crippling, as well as the cost of compensating all the victims.
As such for energy and economy security Taiwan is looking to move away from nuclear.
More anti-technology FUD from the 1970s. (Score:2)
Nuclear power is an existential threat to Taiwan.
Bullshit. We know how to build nuclear power plants that can hold up to earthquakes, storms, or whatever 1970s FUD you can bring up. One example is to put the reactor on a floating platform. We've been putting nuclear power plants on submarines and ships for something like 70 years now. If we can put windmills on a floating platform that can hold up to an earthquake or hurricane then we can use that same kind of platform for a nuclear power plant.
Fukushima was a 2nd generation nuclear reactor design, a
Re: (Score:2)
Nuclear power is quite safe and Fukushima proves this. ...
You lost me somewhere there
Re: (Score:2)
You are the only person that seems lost so maybe the problem is on your end.
Re: (Score:2)
It makes far more sense to implement large-scale solutions that reduce emissions on a national or even global basis.
Yeah, we should all sit around waiting for the American public to vote for a president who'll sign a globally agreed policy to limit the USA's CO2 emissions.
That'll work.
Re: Boutique solutions are silly (Score:2)
They're a foundry, they don't do chip design. The best they can do in that respect is make the process technology available to make more energy efficient designs. From what I know, they do, or used to, when I still did tapeouts, they had a performance and a low power version on a process node, alongside a regular in-between flavour.
And I will go vegan by 2050, promise! (Score:3)
Right now I prefer to buy more offsets for my rib eye consumption.
2050? (Score:1)
China will have invaded Taiwan by then and there will have been or not have been a full scale war by then. Either TSMC will not exist and be controlled by the CCP or we will have the first nuclear world war and global warming might be our least concern. Carl Sagan bet on nuclear winter.of course. In the 1970's they predicted A/C would not exist by the turn of the decade.
Great (Score:3)
Re: (Score:2)
Probably, since they'll have to change materials eventually.
Water Consumption (Score:2)
There are all sorts of entirely legitimate reasons that TSMC set up their operations on the island of Taiwan - they’re a Taiwanese company, after all.
But to a certain extent the electricity and water consumption of this company are known, predictable quantities yet so far what they appear to have done is set a questionable target to address one and
Re: (Score:2)
The real issue is likely more about wastewater than process needs; they need to find a way to strip out all the chemicals after washing the wafers. The rinse process (not sure of the technical name) requires near-zero concentrations of the chemicals to be washed away.
Find a way to pull the chemicals out and you can recycle the water. Much easier said than done. Added benefit though is that the (very diluted) chemicals don’t go into the water supply.
Distillation (Score:2)
I'm not sure about RO (Reverse Osmosis) though. It works just fine for extracting fresh water from brine (sea water)... I'm just not sure if the specific contaminants in this case can be trapped by current membrane technology... or what the resultant purity would be.
Re: (Score:2)
The government requires them to clean the water before it is released from the plant. Basically they have a waste-water processing facility on-site. Even small factories are required to.
Check out some of Gamers Nexus' factory tour videos on YouTube, they often cover waste water treatment.
Re: (Score:2)
Yes, they could use extensive off-shore wind and solar for power, and divert some of that to reverse-osmosis for their fresh water but is that really less impactful than picking a better location in the first place?
That's not wise because wind and solar cost more than other energy sources, especially offshore wind and rooftop solar. They'd need considerable amounts of energy storage to cover the times when the wind and sun are not sufficient. The most viable energy storage we have is pumped hydro. You mentioned hydro power and if there is ample hydro power available then adding some wind and solar means relying less on rainfall.
There's been many people that looked into our energy and water needs in many nations all
Re: (Score:2)
Against popular believe, there are regions on the earth where consumption of water - drinkable water - is not a problem at all.
That's quite a bit of confidence in his abilities (Score:2)
as a CEO. I mean, setting a goal almost 30 years down the road pretty much means that he'd expect to stay another quarter century or so, before his successor has to deal with the crap.
Any normal CEO would have promised that for 5 years, knowing that he won't have to pull through with it anyway.
Re: (Score:2)
In the US. ... if you wonder.
But the company in question is in Taiwan.
That is in Asia
Hence: different culture.
Nuclear Power is 20% part the solution (Score:2)
Re: (Score:2)
Nuclear power is the solution, unless radioactivity is an issue close to fab plants.
Even the best maintained coal power plants release more radiation than any properly operating nuclear power plant. The radiation standards are so tight on nuclear power plants today that if the control room of a nuclear power plant had used the same granite in the floors as in Grand Central in NYC then they'd fail any inspection. People get more radiation at a train station than at a nuclear power plant.
Nuclear power is the solution if radiation in the fab is a concern. That's because all the radiation
Re: (Score:2)
Even the best maintained coal power plants release more radiation than any properly operating nuclear power plant.
That is triple wrong.
1) most coal does not contain any Thorium at all. Hence it is not radioactive
2) German coal plants are not allowed to burn coal that contains radioactive isotopes
3) on most places on the planet, coal plant exhaust is scrubbed
So, if at all, the radioactive elements are in the ash. And: the concentration is so low, it does not matter at all.
On the other hand: even in German
Re: (Score:2)
Are you allergic to providing sources? Where is Germany finding coal free of radioactive isotopes after Chernobyl? You do know that there are radioactive isotopes besides thorium in the dirt, no? Some of them are artificial and spread widely over Europe. There's no radiation free coal in Germany because a trusted source (that would be you) tells me that the wild boar in Germany glow in the dark from all the radiation in the dirt, put there from Chernobyl.
Then the radiation that isn't there ends up in th
Re: (Score:2)
Where is Germany finding coal free of radioactive isotopes after Chernobyl?
Chernobyl did not contaminate coal mines 100ds of meters under the ground.
You do know that there are radioactive isotopes besides thorium in the dirt, no?
You mean after Chernobyl? Did you not just claim coal plants produce more radioactivity than nuclear plants? Now you want to claim the reason is: because of a nuclear plant?
There's no radiation free coal in Germany because a trusted source (that would be you) tells me that the wild
Re: (Score:2)
Claiming we have now morw radioactive coal because of Chernobyl is pretty stupid.
No, it's quite brilliant. I just got you to see how people safely dispose of radioactive waste. We don't have a radioactive waste problem, you just demonstrated how we dispose of it. We don't have a radiation problem from Chernobyl, you just demonstrated how to avoid the radiation. I finally shamed you into providing a link to back up your claims, and it is a link that proves my point. There is no radiation problem with nuclear power, and we know this because you demonstrated it.
Thank you so very much
Re: (Score:2)
We have a "nuclear power" radiation problem, as most of Germany is still polluted by Chernobyl.
I demonstrated nothing about Chernobyl.
I posted links about COAL. And I perfectly clearly pointed out: coal does not put radiation into the environment.
So: no idea what your reading or comprehension problem is.
Perhaps, first, they could come up with... (Score:3)
the technology required to fill all the orders for semiconductors that are currently on the books. The world needs its critical suppliers to be able to meet market demands for components that are required by a huge percent of products, before it needs them to risk reducing capacity as they attempt to achieve boutique political goals.
Perhaps, If they cannot even supply demand reliably now, they might completely destroy the global economy in the future by failing to do so even more spectacularly.
Alternatively, perhaps their carbon neutrality will not matter because they'll no longer be a critical supplier.
Re: (Score:2)
the technology required to fill all the orders for semiconductors that are currently on the books. The world needs its critical suppliers to be able to meet market demands...
Let's just stop right here for a moment.
If you're looking to solve the actual problem here, it's market demand.
Foundries and society in general should be figuring out how to make components that last. Thanks to incessant and runaway Greed, how many chips are installed in disposable hardware barely designed to last 2-3 years?
THAT is the insanity that needs to stop crushing the demand side of this. We should be taking a hard look at disposable electronics that drive chip demand, and look to not merely pena
Re: Perhaps, first, they could come up with... (Score:2)
Re: (Score:2)
People want new shiny.
Humans also need a place to live. This quickly becomes an I-don't-give-a-shit "problem."
If you build them to last, with an increase in resources needed to do that, and they get thrown out as outmoded, then the problem has been made worse.
Plenty of products ARE still built to last. And they do. Those companies aren't going out of business, nor is society revolting against their longevity, price, or not "new shiny". Haven't even when everyone was making hardware that was built to last half a century ago. Greed insists we can't make things more reliable or at least less disposable, which is becoming a piss-poor excuse.
Only if people don't want to upgrade or if products can be refurbished and sold to the less demanding, does it make sense to make products more robust.
Yeah, or when humans find the plan
Re: Perhaps, first, they could come up with... (Score:2)
Re: (Score:2)
If an item is of a sort that does not become outmoded quickly, then building it to last makes sense. My favourite set of parallel-nosed pliers I inherited and are 75 years old and going strong, and my favourite pair of secateurs forty years old. My hi-fi amp is twenty-five years old, the second one I've had, and I see no particular need to change it. The speakers are fine too. But for things like phones (more relevant to a chip maker), rapid progress means obsolescence and replacement. I'd actually be happy with more robust phones that could be passed to new owners, but in a free market, there is no particular profit in offering something that will definitely last more than maybe three years, even if someone (e. g., my mother) would still find it useful. Yes, things can be resold on eBay, but the manufacturer doesn't get any benefit, and in fact it can eat away against sales of current low end models. From the perspective of the environment I'd agree it's a problem, but it's hard to see any easy fixes.
If there is a sound, solid, and proven second-hand market today for used smartphones, then this "rapid progress" you speak of would have destroyed that market by now. Rather obvious that obsolescence doesn't happen anywhere near as fast as you're suggesting, and therefore even the speed of technology would be able to support making and marketing a phone with a 5-7 year useful life instead of 2-3. And given the sheer numbers we're talking about here, that's a significant drop in chip demand.
The problem her
Re: (Score:2)
Well,
my iPhone 4S lasted over 10 years. However the network providers are disableing 3G. And I wanted a dual SIM phone.
Funnily I ended up with a new iPhone and a Samsung.
I expect both to last more than ten years (my old iPhone works just fine on wifi)
Re: (Score:2)
Outlaw people from buying new electronics? That's not going to happen. Getting people to vote on this isn't going to happen, that's your first problem.
Assume we could get people to agree to outlaw new purchases of phones, wristband devices (calling them a "watch" is almost an insult), tablet, or whatever. What of the health and safety benefits of a new device? Maybe I have a Fitbit that monitors heart rate, but I see an Apple Watch that has an EKG. Later on I see one with a blood oxygen sensor. Then t
Chip manufacturing is not green (Score:2)
When you really think about it you're trying to make something so pure (a microscopic silicon circuit) that even a few atoms out of place will break it. Anything in our Universe where you're trying to make something super pure is going to result in massive amounts of energy and resources burned. Chip making was never considered "clean" by any means.
Taiwan wont exist as a state in 2050. (Score:1)
As soon as fabs in the EU and USA are operational, the west has no more reason to protect Taiwan from the Chinese. It will be a very short war.
After that; no western company will order chips from TSMC again.
Re: (Score:2)
You say that as if TSMC isn't building fabs outside of Taiwan.
Re: (Score:2)
Far out (Score:2)
Set a deadline for reduced CO2 emissions in 2030 and you'll see them scurry for safety, since they know they'll have to make hard choices to meet it.
Re: Far out (Score:2)
'bout 20 years too late to avoid the apocalypse... (Score:1)
... but ain't we virtuous! Look at us!