TSMC: First 7nm EUV Chips Taped Out, 5nm Risk Production in Q2 2019
anandtech.com
anandtech.com
One of the factors that prevents smaller companies from designing FinFET chips is development cost. Some estimates put the average cost to develop an SoC at around $150 million in labor and IP licenses. With N5 generation, these expenditures will rise to $200 – $250 million, according to EETAsia, which will limit the number of parties interested in using the tech.
Ouch. A $250M NRE cost for an SOC, really puts pressure on guaranteeing the volume of the part in order to recover any sort of margin on a part like that.
Pretty cool to see EUV seem to finally emerge out of the 'someday' status into the 'some now, more next year' status.
Apple sell roughly 60%+ Flagship iPhone every year, that is close to 140M unit, it would cost them roughly $2 per unit. As far as I am concern, I don't think Qualcomm sells anywhere close to that number of SD845, although the design cost can be utilise in their mid range phone at a later date.
And it basically means from 7nm onwards, Only a few companies can afford using leading node, namely Apple, Qualcomm, Huawei, Samsung, Nvidia and AMD. Other smaller players like Mediatek might have to wait longer before adopting it.
As in, "like everyone else, we haven't really figured out EUV shot noise yet".
Nice marketing though, lolz.
Octa patterning? ;)
So you still need the multi-megawatt death ray as a primary light source for a fab scale use.
>> "Risk Production means that a particular silicon wafer fabrication process has established baseline in terms of process recipes, device models, and design kits, and has passed standard wafer level reliability tests."
It sounds like a batch produced to certify the process for insurance and contract purposes.
At the start of risk production, by definition there have been no customer designs put through the fab. Actually this is not quite true, Sun explained. Before the start of risk production the company has already run a number of shuttles with test chips from customers, so foundry and clients are already starting to wring out the more critical structures in the first customer designs. But these test shuttles are not full chips either.
-- https://www.edn.com/electronics-blogs/practical-chip-design/...
So, it's a point in the process where the foundry is taking a chance, and the customer is also taking a chance.
I've also heard it applied to a customer design, e.g. the customer starts to volume ramp production before they finish fully testing the design.
You don't need to do anymore R&D or major NRE, but you may need to do a lot of tweaking and tuning. Risk production is generally low volume because stuff doesn't work so well yet.
Risk production just means the "customer" will accept the parts from the fab without the parts being fully tested.
In other words, the customer is accepting the "risk" of parts not working. When the fab settings and test flow are mature then the fab assumes responsibility for parts that do not pass the testing.
The mainland Chinese government disagrees.
I don't know about these particular tariffs, but I believe the U.S. would not consider tariffs aimed at (mainland) China to apply to Taiwan.
That's a minority/extremity opinion which isn't remotely prevalent even among those who feel intimate with China. It used to be the 'political-correct' belief held by the former authoritarian ruling party. But it's not considered a rational worldview anymore after its democratic modernization.
There are things that are de facto true, even if they're not de jure true.
The status of Taiwan is one. Israel's nuclear status is another famous example.
While the common sense reaction is "everybody knows it, why pretend?" those fictions generally exist because they help maintain delicate balances and order between parties that are generally at odds with one another.
What it comes down to is that it'd be lovely if we could all just say openly and outright that mainland China and Taiwan are separate entities, because, in practice, they are. But it's also pretty swimming for the global community for the PPRC and ROC to both claim that they're the real China while everyone does business with each other anyway. So while it's true that they're distinct and it's also true that everyone knows that to be the case, everyone also wins out to varying degrees by officially pretending that it isn't the case.
More information can be found in this position statement from the Department of State:
Taiwan is hurt indirectly. Taiwan has invested heavily on the mainland China and manufactures there.
Many Chinese products exported to US have very low Chinese value added. Chinese manufacturer buys Taiwanese, Japanese or US made components, assembles them and sell them to the US.
For many high-tech products, Chinese value added may be just 20% of the price of the product. Chinese tariffs hit everyone in the value chain. Even the US.
That is not strightly true because there isn't really an alternative to Foxconn in size and scale. But in other industry where there are lots of alternative around the world, the heat are quite intense.
Other nations play pretend to various degrees, trying to placate (mainland) China without too much swallowing their own pride in not being China's lapdogs. But they're effectively two separate countries.
So: The tariffs being applied to China do not apply to Taiwan. However, Trump seems to be throwing tariffs at several targets lately; he may or may not have applied some to Taiwan.
Sometimes you wonder how things can stay broken for so long.
Taiwan doesn't want the mainland's government, and the mainland doesn't want Taiwan's government.
If Hong Kong had been militarily defensible, it'd be in the same situation.
If the UK had really wanted to keep Hong Kong, could it have militarily defended it? No doubt the geography is less than favourable; but, keeping in mind that UK and PRC are nuclear powers, would regaining Hong Kong be worth the risk of war between two nuclear states? PRC might well have decided it wasn't worth that risk.
But in reality, the UK had no real interest in keeping Hong Kong, and were more than happy to hand it over to PRC. (Yes, there was a treaty saying they had to give back a big chunk after 99 years, but if the UK really wanted to get out of that, they could have found a solution – e.g. if they hadn't switched recognition from ROC to PRC, they could have asked ROC for a lease extension. Or, they could have made their switch of recognition from ROC to PRC conditional on PRC granting them perpetual sovereignty over the whole of Hong Kong.)
The Chinese would not have to have fired a shot to make HK surrender. HK was then a city with more than 6 million citizens, which was entirely dependent on water imported from the mainland. The treaty that allowed the importation of all that water was only set until the end of the lease, and the Chinese very clearly stated that they had no interest in extending it.
Also, earlier when HK expressed some interest in investing into desalinization, the Chinese threatened to immediately cut off the water supply if it seemed like the HK local government would have actually tried to build up enough water resources to make them not dependent on the mainland.
The result of shutting down the taps would have been millions dead within a week. Keeping HK British would have required a massive invasion of China by the UK, just to secure water. I don't see that as very likely.
Hong Kong also had its own desalination plant 1975-1981. Why was it shut down – because of pressure from the PRC government? Or because it was significantly more expensive than importing water? I don't know for sure, but my impression is the later was a much bigger factor than the former.
If the UK had thought this through, they would have setup real democracy in Hong Kong a generation before the lease ended, instead of doing the colonial "high commissioner" thing. A legitimate Hong Kong government would have survived a lot longer with their people's support.
A democratic HK would squarely falsify this claim.
Indeed, I'd argue that the single most urgent reason (but not the only one) for China 's bullying of Taiwan is precisely this, Taiwan is a thriving democratic and Chinese society.
But if the UK had done the same thing in 1980 or 1970 or 1960 or 1950? I don't think the PRC could have done anything about it.
The UK never really cared about Hong Kong democracy. If they had, they would have implemented democratic reforms decades ago, instead of leaving it until the very last minute.
Both for their colonies and for themselves.
China may be able to force import tariffs on goods imported by Taiwan (idk not an expert) but I bet Taiwan imports relatively little directly from the US?
Heh. Exactly how close an ally is yet to be tested in any meaningful way. No U.S. aircraft carrier group has transited the Taiwan Strait since 2007. Some would forecast that no U.S. carrier group will transit - or finish transiting - that strait ever again.
It was rumored TSMC's 7nm wouldn't be the same as Intel's 10nm. But if they reach ~5nm then they'll likely be knocking on Intel's 10nm door. Combined with the double whammy of 7nm Eypc servers from AMD it seems like Intel's technical offerings are rapidly getting commoditized by the rest of the market.
Maybe it's easy to find out in google but cant find the correct combination of keywords...
But when we hit 20/18/16/14nm that went out the window and it became a marketing term, not so much a literal description. A lot of this was driven by moving to FINfet's which are really MOSfets, as they have lower leakage at smaller sizes, but they also aren't square which makes generalizing a singular node to density a bit wonky.
Wouldn't 5 nm allow shorter wire traces, which would reduce another limiting factor in computational speed?
The fundamental limit of clock speed is power draw. As clocks increase the wattage ~ frequency relation goes from
frequency = Constant * Power Draw
frequency = Power Draw * Power Draw
Packing things in more tightly lets you spend less time in transit, which might let you squeeze more gates into a cycle, or do the same things a little faster.
Size is small enough for anything consumers care about, even watches.
Intel's instruction set architecture dominance will keep it going for a long time but ultimately it would probably make sense for Intel to spin off its fabs into the US equivalent of TSMC and capture more margin revenue from their designs versus their process.
[1] http://www.tsmc.com/download/ir/annualReports/2017/english/p...
[2] https://s21.q4cdn.com/600692695/files/doc_financials/2017/an...
TSMC ~ $11B Intel ~ $9.5B
TSMC has become more profitable than Intel since 2016.
If you have ever been in an investment club (basically friends who get together and exchange ideas on what stocks to invest in) you have probably had the 'revenue/income' discussion.
Amazon is the poster child for 'net income doesn't matter' as it has reinvested most of its margin in itself over the years to grow.
So it can go either way.
TSMC does pay a dividend which Intel doesn't so from a financial trading/investing point of view you could argue that TSMC is a better stock to hold long as it will generate income.
From the 'future of the business' point of view re-investing more would seem to be the wiser strategy. Time will tell.
In terms of roadmap TSMC has been extremely open about their progress all they way down to 3nm and has been executing to perfection. They take pride to be Apple's pure play Fab.
I do believe in re-investing for the future, the problem is I have trouble seeing that future from Intel. And their management has been lying for far too long doesn't bring me confidence.
Time will Tell.
It turns out that Intel's architecture cheated by skipping privilege checks during speculation, while AMD designs did the correct thing. SMT turns out to be a security nightmare, but in any event AMD now offers the same capability. Either way SMT is another lost advantage.
That leaves Intel's process technology and vertical integration. If it outsources manufacturing it will have effectively ceded these completely, meaning Intel will have lost all of its competitive advantages.
I suppose Intel's human capital might be a competitive advantage, but their missteps cast serious doubt on that.
node ratings are a marketing joke at this point. Actually most likely it became a joke more than 10 years ago.
That's not true. Imaging ICs (or other structures) down to sub-nm resolution is much much easier than making them. Most physics labs will have no problem doing that..
The various imaging techniques (AFM, STM, etc.) used can routinely achieve sufficient resolution to verify these claims with fairly moderate amounts of tweaking and cost. STM is often used to image things down to atomic resolution.
For example, see Chipworks's analysis of Intel's 14nm process:
https://www.chipworks.com/about-chipworks/overview/blog/inte...
"As yet we don’t have any detailed TEM imaging to look at the transistors or fins in close-up"
And that's 14nm from years ago, not the 7nm or 5nm being claimed these days.
> ... imaging techniques (AFM, STM, etc.) ... with ... tweaking ...
I don't see the 'not true' part here. (Also as I responded to the other commenter who shared a third party link, even 14nm could not be verified in 2014: "As yet we don’t have any detailed TEM imaging to look at the transistors or fins in close-up").
The wiki page on EUV is uncharacteristically excellent.
https://en.wikipedia.org/wiki/Extreme_ultraviolet_lithograph...
WTF!? Why? Is there any hope for better optics?
That's why you have to build fabs around a multi-megawatt laser source.
The only known hope going forward are photonic crystal mirrors
Even if you account for feature size marketing and nomenclature differences, it doesn’t look good.
And the $250MM mentioned, pardon my language but they crap that much for breakfast. I don’t mean to overstate it, look at the financials, the net profit available to them after existing R&D investments is staggering.
Sometimes I think the difficulties between Jobs and salary don't make any sense at all.
Intel spent 13 billion on R&D last year, and had $26 billion left over ebitda profit still. You can argue a different profit number than ebitda should be used, but pick any one you want, they’re all massive.
Say your friends make 50k salary (just for arguments sake and let’s not calculate a fully burdened salary for simplicity).
If 1,000 Intel engineers were averaging 150k in the US, the difference would only be 100MM.
That is so insignificant to them, they couule use it as fire wood just for fun. So it can’t be higher labor costs.
My best guess is they made a few “big bets” that went wrong and cost them dearly. Or there’s always the old standby thst the wrong leadership is in place at the EVP/VP levels and it’s had a negative ripple effect.