I remember a lot of noise when Loongson added support for their CPUs to Go. Go already had MIPS64, so if they had to do porting work, it might not be just a MIPS clone. I haven't looked at either architecture in detail for a long time, though.
Why the custom arch? If building a CPU is a challenge, then supporting an entire toolchain and ecosystem for it is a whole other level of complexity that isn't necessary or helpful. Moving from being pure MIPS only makes sense if there is some patent issue that you are trying to avoid because you want to sell to foreign markets to raise volume of production.
But then just go with RISC-V, I mean there are legions of ravenous fanboys out there _and_ it's already got support.
So the only thing that makes sense is that you have legacy MIPS clients and machines where either you want a drop-in replacement, or compatibility with them... but also want to piggy back on RISC-V and will end up with a CPU with a split personality.
"The release shows that domestic CPUs have reached a new height in self-sufficiency and performance. It also demonstrates that China is capable of producing first-class products based on a proprietary CPU structure,"
As this is also kind of one of the responses to the West trying to technology-boycot them (what can be considered now hugely failed and soon, if not already, backfiring..) could also be seen as a feature, and not a problem?! ;)
We'll have to wait until China's ASML stockpile is "old" tech (in 5 years or so), before deciding if the tech embargo has worked. I don't support the embargo, by the way.
It's possible that they looked at RISC-V and concluded that it was unlikely they could achieve their performance goals with that architecture.
As far as I understand it, the new Loongson CPU and the Asus mainboard are available commercially today. I don't think there are any RISC-V CPUs with compare single-core performance.
When LoongSon started the transition from MIPS to their own ISA, important parts of the RISC-V standard such as the Vector, Crypto and Bitmanip extensions weren't yet ratified.
Their long-term strategy is to support other ISAs through binary translation, thereby skirting around IP issues associated with them and making the hardware ISA matter less anyway.
It is exciting to have more competition in the CPU space. While it is currently possible to build a non-x86-64 desktop class computer, all the current options are really expensive (looking at you Talos). It would be nice to have more architectures at this level eventually. Even if only to see what the real costs of the 8086-legacy support is to modern performance (if any).
It'll be interesting to see the Snapdragon X Elite next year. It won't take the performance crown but it should finally give reasonably priced ARM outside of the Apple ecosystem something in the high performance class.
As for the real costs the upper bound seems to be no more than ~30% on average for the software layers to do it so probably less than that if the hardware is built from the ground up to run x86 code only. My guess is the real difference is <10%. Still appreciable, but not something that really drives the gaps.
Considering how most of the software we run is open source, I don't think legacy support is that important. I think lots of organizations would be happy to switch for 10% improvement in performance or power.
Open source could really facilitate a golden age in CPU design where backwards compatibility really doesn't matter.
> Considering how most of the software we run is open source, I don't think legacy support is that important
At the bottom of the open software pyramid are drivers, and they are often closed-source. Qualcomm is a major contributor to a lot of Android devices having short support periods. Even though Android is open source, Qualcomm's drivers are not, and Qualcomm would rather have consumers buy devices with the latest snapdragon than keep using an upgraded device from 5 years ago.
For the typical desktop user, windows drivers. This includes things like anti cheat in games. For the typical HN desktop user it should be good though.
8cx Gen 2 are ~$500 new. That's not exactly a bargain Chromebook but it's also not particularly expensive. Of course, we'll have to see what they do for pricing on this upcoming chip. Given the performance levels even just a desktop x86 CPU that competes may sell for that $500, not counting the entire rest of the computer.
> Snapdragon X Elite is faster than all but the M3 Max.
> From what I've seen it actually beats the M3 series in single-core performance.
Citations? Because from what I have found [1], Qualcomm only claims it’s 7% faster than a non-Pro, non-Max M2 in single-core. It’s 50% faster than the M2 in multi-core but that’s using 50% more cores
And all of this on a system with a TDP almost 4 times that of the M2 system, meaning it’s consuming a lot more power to get essentially the same performance.
All of this on a chip that’s coming out in 2024 at the earliest, when Apple will be introducing the M4 series — two generations down the road from what the Elite X is performing at.
Looks like the numbers I was looking at were for the high-power configuration, and differed a little between sources, nontheless the low-power configuration, which has a significantly lower TDP (23W) than the M3 Pro (~30W), while having the same core count, is still very close to M3 Pro performance even if M3 Pro is a bit higher.
Your graph says the single core performance of the high tdp option is lower than any of the M3 models and the multi core performance is lower than both the pro and max. It is close though, but it's also comparing CPUs out now to CPUs to be released in the future.
Not to mention, Qualcomm has a pretty bad history of meeting its promised dates.
In 2021, it claimed to be on track to shipping M-series-level chips in 2023. [1]
In 2022, it claimed to be on track to shipping M-series-level chips in 2023. [2]
Now, in 2023, it claims to be on track to shipping M-series-level chips in 2024. At least that date is still in the future, but I think it should be looked at skeptically in light of their past unfulfilled promises.
Be careful what you wish for. I don't want this kind of competition. You'd have to be crazy to use these given the sheer amount of tech theft and hybrid warfare coming out of China.
> Semiconductor Manufacturing International Corporation (SMIC) is a partially state-owned publicly listed Chinese pure-play semiconductor foundry company.
Not strange. Chinese government control all the industries through party delegates that injects in the companies that become relevant, and plans strategically their movements as the state needs, so TSMC has little chance with the sues?
Other different thing is that thankfully for us, most of those political delegates kills part of the initiative of those companies due corruption.
If they are so self-destructive then why are we so worried about Chinese technology to the point we need to throw around ever-stronger sanctions? Something is not right about this narrative. Sounds like a bunch of huge assumptions to me.
Nobody is afraid of tech competition. The US is acting on the assumption (intelligence?) that China will start a war of aggression in the near future. If war is a given then every bit of tech will translate into more dead. An ounce of prevention is worth a pound of cure.
It is not self-destruction, its more like it mostly induce a slow development and quality of in-house tech dependent products. Not always, just mostly, but you're right this is an assumption based by my mere external observations, I have never lived in that country.
Also I'm underestimating that there are countless Chinese citizens studding overseas, and the government have nasty resources for to make them return back if their knowledge becomes relevant.
The matter is the sanctions. As I see it, comes more from other facts,
The first one is the Chinese government has been pursuing a coordinated long term dumping plan towards independent companies allocated in the rest of world. Independent companies are hopeless when a whole developed country that is able to manipulate freely the prices target them and run against them; the companies enters in economical problems and are acquired by the Chinese government for keeping the influence expansion. Chips manufacturers are between their top targets.
By other way, such government want to invade Taiwan, one of the main focal points were the chips are manufactured for the rest of the world, and this would be done with armament guided through chips manufactured ironically by those same factories.
What worries to many countries also, and affects the world in general, if Taiwan were invaded, the first consequence of losing the ability/independence in chips manufacturing through some years of prices dumping, is that -among other things- the armament would require of Chinese's government approval. Right now is being subsidized the creation of chip factories from Taiwanese companies across the world, whom -for not lose protection- are not exporting their top tech into those factories.
They are also worried with telecommunications, because Chinese government is involved in companies commercializing products where they are able to introduce backdoors within the designs, and is promoting their use through prices dumping.
More or less I understand it like this, but Im sure it does exist even more factors that keep affecting all of us.
This is like the Hyundai (automobile) of CPUs. When Hyundai started competing in the US, it's cars were slow, cheap and kinda crappy. They got some foothold by being cheap above all... But now, their cars are formidable competition to some of the best out there, and are leading the way in a lot of areas. Intel better not get overconfident again, these guys could catch up before they know it.
Hyundai didn't have the problem of requiring people to build roads differently. Loongson however will have that problem, as software is either targeting x86 or ARM.
> "Loongson however will have that problem, as software is either targeting x86 or ARM."
This is one (of many) advantages of open source software. If you have access to the source, you can build it on a ridiculously wide range of hardware. I've been downright shocked over the years at some of the strange things people have run actually useful (and entertaining) open source software on. Doom runs pretty much everywhere just for one example. ;)
Even as someone who hasn't compiled a kernel in a loooong time (10-12 years?) this is a weird position to me.
AFAIK Loongson's ISA is some sort of superset of MIPS64. Not exactly ubiquitous but MIPS64 has been around for a while so I would expect bootstrapping to be fairly simple.
Of course, bootstrapping is not the same as, say, using the parts of the superset that help with x86 emulation to speed things up, but even QEMU has supported LoongArch since 7.1 (over a year ago.)
That's not an insurmountable problem with a billion people as captive audience. At first glance the ISA is nothing special and that's a good thing. It's a run of the mill 64 bit RISC ISA with little endian byte order and no unusual alignment or store order constraints to watch out for. If something compiles for x86-64, ARM64, RV64GC, etc. it should compile for that LoongArch if there is a compiler (backend) for it. If the MMU stayed as similar to MIPS64 as their earlier chips it shouldn't be too much work to write and maintain a Linux (or *BSD port to it).
The custom ISA becomes a real problem if you need Windows and other closed source software on it, but China has declared it political goal to rid itself its dependency on Microsoft.
I don't think anyone is suggesting that China isn't capable of building an efficient and powerful CPU or not, but will they, at any point, catch up to contemporary CPUs from Intel, AMD, Apple, etc.?
I am suggesting that they can't. They have a serious innovation problem. A lot of their designs started out stolen or cloned from the West. They also don't have anything like ASML available to them. And to get this out SMIC ripped off TSMC intellectual property.
I'm not saying they won't deliver useful silicon but they won't be delivering comparable silicon. And it'll be under trade embargo no doubt due to the IP issues so they will have a very limited market to scale into.
This happened before (GDR) and the products were never competitive.
I think the best thing they have done though is kick out compatible ATmega parts and the STM32 parts. They're really killing it there.
> And it'll be under trade embargo no doubt due to the IP issues so they will have a very limited market to scale into.
This was created for their own domestic market. China's goal is achieve independence of foreign technology. With 1.3 billion people they have a very large market.
> A lot of their designs started out stolen or cloned from the West
This is also called learning by doing. This is how skill and expertise is most efficiently developed in the very beginning.
This trope of assuming that copying is inherently valueless or devoid of creativity or innovation is simply wrong, and comes from an implicit assumption of superiority of the self.
They may not innovate as much as you expect now. But they are building a foundation for future innovation.
Also: Huawei is still leader in 5G technology, with the best products. You should look at their patents, they own the majority of 5G patents.
> I think the best thing they have done though is kick out compatible ATmega parts and the STM32 parts. They're really killing it there.
GigaDevice and WCH have completely new MCUs now based on RISC-V.
They keep the STM32 peripheral register interface for easy driver re-use, but they are well beyond mere copies these days.
If you told anyone in 2013 that TSMC would be a few nodes ahead of Intel you'd be laughed out of the room. So who knows, maybe China can do the same. I'm flabbergasted how the west lost its lead in manufacturing in geral. Nothing surprises me anymore there.
China have a population of 1 billion people, access to the pool of advanced manufacturing and hardware experience of East Asia and demonstrated expertise at high tech and manufacturing.
"They can't buy machines from this Swiss company!" is not really much of a barrier when we look at the resources they have available. Over 12 months, that is a problem for them. Maybe 5 years. In a decade, ASML relying very heavily on the CCP shooting themselves and their people in the foot to still be a differentiator.
When talking about the sheer scale of 10% of the globe's population, you need a lot more than one technical dot point to say they can't do something.
Don’t disagree with your general point though. Also there is always the option of industrial espionage as we have seen.
Leaving the politics aside, it will be interesting to see what technical solutions emerge if the ‘easy’ option of spending a few hundred million dollars on a couple of EUV machines isn’t available.
Heh, oops. I thought precision and went with Swiss. How thoughtless.
And I'd assume the answer is spending a few billion dollars developing a couple of EUV machines over a long time. It isn't like a billion dollars is much money and the Chinese economy is reaching the point where people will be looking to muscle in to precision manufacturing. It isn't like cheap labour is an option any more. They've already overwhelmed the Europeans in a bunch of phone and computer related areas so there is evidence that the Europeans probably can't maintain a competitive edge if the Chinese want to get good at something.
Honestly, given what happened with the iPhone and Nokia I don't know how anyone is placing faith in the Chinese not being able to catch up. European tech companies have a history of looking unbeatable until they're just knocked completely out of the race. If anything the threat is that Europe won't be able to keep up with Asia.
You may be right, but we will need to wait and see.
China is not the first to take a crack at this. Attempts to replicate and catch up with American and European chipmaking and computing efforts go back to the dawn of the discipline, and had involved some of the largest economies of their times, yet here we are.
Other attempts to crack this failed mostly for economic reasons, not technical. Of course, not understating how hard it is technically, but the bigger problem was that competitors didn't have the funds nor the risk appetite to invest 20+ years of R&D into something that may or may not work.
We now know it works. Before, why should China bother when they can just buy from western suppliers and TSMC instead? Now we gave China no other choice but to invest 10-20 years into this until they have something that works. This is why you see all the chip manufacturing equipment makers in China have something like 150-200% YoY growth in the past 3 years. Their R&D centers are pursuing multiple avenues to make sub-5nm chips possible. Not only multiple avenues of making EUV work, but also non-EUV avenues. It just so happens that starting this year we know that nanoimprint lithography works, and is cheaper than EUV. Huawei's newest CPU performs on par with CPUs on smaller nodes. That's because Huawei invested heavily in CPU design and heat dissipation technology. They haven't even gone into advanced packaging yet.
Will they stay behind western technology forever because we have an R&D lead? People who claim this are ignoring a big factor: they will no longer be able to profit from Chinese buyers. These profits are necessary for future R&D investment, which is necessary to maintain the lead and R&D speed.
> Other attempts to crack this failed mostly for economic reasons, not technical.
That's debatable.
Samsung, a behemoth in Korea and one of the largest holdings in the world, has tried to compete with TSMC at the same node size, and failed so far.
> Will they stay behind western technology forever because we have an R&D lead? People who claim this are ignoring a big factor: they will no longer be able to profit from Chinese buyers.
Doesn't this work the other way around as well? Chinese companies in the same space will not be able to profit from EU and North American customers, historically the largest spenders.
Samsung "failed"? As far as I know they are pretty close behind TSMC. Sure, not at the exact same node, but being this close I don't call failure at all. Even staying in the race and keeping up the pace is a huge achievement.
> Doesn't this work the other way around as well?
It does. But the difference is:
- They're coming from behind and enjoy latecomer's advantage. They can take advantage of western R&D results. They may not be able to obtain all the details, but just knowing that a particular R&D path is viable is already an advantage.
- They have a captive market thanks to sanctions.
- They're coming from a smaller size. Chinese semiconductor companies had such small market share that capturing even a small part of only the Chinese market already translates to tremendous growth. They can only go up while the western supply chain will go down.
Finally, there's the global south which is also growing. This is a huge market that they can sell to. In the future, western suppliers will have to compete with Chinese suppliers in this market, even if Chinese suppliers are locked out of western markets.
That wording is a bit harsh. They haven't failed, but they haven't been able to compete at scale, which is what matters.
> They're coming from behind and enjoy latecomer's advantage. They can take advantage of western R&D results.
Isn't that the main issue, that they can't?
> They have a captive market thanks to sanctions.
That isn't true. Current sanctions target exports of certain consumer and professional products, none of which are chips as far as I understand; and imports of certain products branded strategic such as AI related hardware.
> They're coming from a smaller size. Chinese semiconductor companies had such small market share that capturing even a small part of only the Chinese market already translates to tremendous growth.
I don't think this means anything in this context. Capturing a larger chunk of a single market doesn't even mean that they are better than most at what they do, just that their go to market strategy is more effective than their competitors. SMIC could be releasing chips with comically large transistors, as long as their scale and prices are preferable to those of their competitors.
> They can only go up while the western supply chain will go down.
This is wishful thinking, at best. But again, growth means nothing here. SMIC could "grow" a 1000% for all I care, and it wouldn't move the needle.
> Finally, there's the global south which is also growing. This is a huge market that they can sell to.
That depends highly on China and their allies. China may capture markets in Africa, but it is actively alienating most of Asia Pacific, minus perhaps Russia.
Samsung are claiming a 3nm process node within 24 months of TSMC, and 90% of TSMC's transistor density.
That is commercially significant and entitles TSMC to bragging rights and huge profits. But if China fails like Samsung, they're at no strategic disadvantage at all and the US technology embargo will have effectively failed to do anything.
> But if China fails like Samsung, they're at no strategic disadvantage at all and the US technology embargo will have effectively failed to do anything.
China is in no position of "failing like Samsung", right now. At most, they may fail like Samsung did five years ago.
Lets not ignore the considerable effort they've spent in hacking companies and stealing proprietary information. (My last engagement was with a company that had a ton of information of interest to State sponsored actors.)
They have spent significiant resources in hoovering as much information as possible from companies that have data of interest.
The issue is whether China will get behind the RISC-V stuff and contribute to a semi-unified ecosystem or whether the individual companies will fuck it up by balkanizing everything in an attempt to one-up their competitors. I'm not optimistic.
The West mostly doesn't really care about RISC-V. ARM licensing costs don't seem to be completely out of line unless you are talking super-cheap devices--which mostly come from China.
(A good counterexample is Nordic Semiconductor. They are having to compete against cheap Chinese BLE alternatives so they needed to start looking at RISC-V).
I was answering the parent remark which was about RISC-V.
And my point is that if RISC-V achieves a significant amount of open designs and tooling then anyone who joins is roughly speaking guaranteed parity (in design - something like ISA and microarchitecture, not in foundry). Which is much better than the current situation.
> Sure, but Apple's phone chips started out with performance comparable to low-end laptop CPUs.
Hummingbird was a bit of a quantum leap for ARM in my opinion. That said, I think Apple was able to do better with what they had thanks to their level of control over things (most notably, the way iOS suspends apps) makes it far easier to have a smooth user experience with a little less muscle.
I haven't looked at the TDP on this, one would probably want to normalize the measurements.
Is there much info about the LoongBT x86 and ARM emulation instructions? I can see there's stuff in qemu/hw/loongarch and target/loongarch, but would love a more general overview.
113 comments
[ 2.4 ms ] story [ 223 ms ] threadBut then just go with RISC-V, I mean there are legions of ravenous fanboys out there _and_ it's already got support.
So the only thing that makes sense is that you have legacy MIPS clients and machines where either you want a drop-in replacement, or compatibility with them... but also want to piggy back on RISC-V and will end up with a CPU with a split personality.
"The release shows that domestic CPUs have reached a new height in self-sufficiency and performance. It also demonstrates that China is capable of producing first-class products based on a proprietary CPU structure,"
As this is also kind of one of the responses to the West trying to technology-boycot them (what can be considered now hugely failed and soon, if not already, backfiring..) could also be seen as a feature, and not a problem?! ;)
As far as I understand it, the new Loongson CPU and the Asus mainboard are available commercially today. I don't think there are any RISC-V CPUs with compare single-core performance.
And afair LoongArch started before RISC-V was as popular as it is today.
Their long-term strategy is to support other ISAs through binary translation, thereby skirting around IP issues associated with them and making the hardware ISA matter less anyway.
As for the real costs the upper bound seems to be no more than ~30% on average for the software layers to do it so probably less than that if the hardware is built from the ground up to run x86 code only. My guess is the real difference is <10%. Still appreciable, but not something that really drives the gaps.
Open source could really facilitate a golden age in CPU design where backwards compatibility really doesn't matter.
At the bottom of the open software pyramid are drivers, and they are often closed-source. Qualcomm is a major contributor to a lot of Android devices having short support periods. Even though Android is open source, Qualcomm's drivers are not, and Qualcomm would rather have consumers buy devices with the latest snapdragon than keep using an upgraded device from 5 years ago.
But yes, for consumer electronics it's a whole other story.
From what I've seen it actually beats the M3 series in single-core performance.
That Nuvia acquisition has really paid off for Qualcomm.
Citations? Because from what I have found [1], Qualcomm only claims it’s 7% faster than a non-Pro, non-Max M2 in single-core. It’s 50% faster than the M2 in multi-core but that’s using 50% more cores
And all of this on a system with a TDP almost 4 times that of the M2 system, meaning it’s consuming a lot more power to get essentially the same performance.
All of this on a chip that’s coming out in 2024 at the earliest, when Apple will be introducing the M4 series — two generations down the road from what the Elite X is performing at.
—
[1] https://www.tomshardware.com/news/snapdragon-x-elite-outperf...
https://cdn.mos.cms.futurecdn.net/zwxHUu7zdH4y5wgbzpwnD9-970..., from https://www.windowscentral.com/hardware/laptops/results-are-...
In 2021, it claimed to be on track to shipping M-series-level chips in 2023. [1]
In 2022, it claimed to be on track to shipping M-series-level chips in 2023. [2]
Now, in 2023, it claims to be on track to shipping M-series-level chips in 2024. At least that date is still in the future, but I think it should be looked at skeptically in light of their past unfulfilled promises.
—-
[1] https://www.theverge.com/2021/11/16/22785181/qualcomm-next-g...
[2] https://www.pcworld.com/article/1382740/qualcomm-dubs-nuvia-...
Edit: https://en.wikipedia.org/wiki/Loongson#Compiler_support
For Longsoon in general, it was apparently enough support for RMS to use one as far back as 2010: https://usesthis.com/interviews/richard.stallman/
https://doc.rust-lang.org/nightly/rustc/platform-support/loo...
https://www.tomshardware.com/news/loongson-launches-3a6000-c...
https://wccftech.com/china-loongson-3a6000-cpu-reaches-core-...
https://en.wikipedia.org/wiki/Semiconductor_Manufacturing_In...
Zoinks!
Other different thing is that thankfully for us, most of those political delegates kills part of the initiative of those companies due corruption.
Also I'm underestimating that there are countless Chinese citizens studding overseas, and the government have nasty resources for to make them return back if their knowledge becomes relevant.
The matter is the sanctions. As I see it, comes more from other facts,
The first one is the Chinese government has been pursuing a coordinated long term dumping plan towards independent companies allocated in the rest of world. Independent companies are hopeless when a whole developed country that is able to manipulate freely the prices target them and run against them; the companies enters in economical problems and are acquired by the Chinese government for keeping the influence expansion. Chips manufacturers are between their top targets.
By other way, such government want to invade Taiwan, one of the main focal points were the chips are manufactured for the rest of the world, and this would be done with armament guided through chips manufactured ironically by those same factories.
What worries to many countries also, and affects the world in general, if Taiwan were invaded, the first consequence of losing the ability/independence in chips manufacturing through some years of prices dumping, is that -among other things- the armament would require of Chinese's government approval. Right now is being subsidized the creation of chip factories from Taiwanese companies across the world, whom -for not lose protection- are not exporting their top tech into those factories.
They are also worried with telecommunications, because Chinese government is involved in companies commercializing products where they are able to introduce backdoors within the designs, and is promoting their use through prices dumping.
More or less I understand it like this, but Im sure it does exist even more factors that keep affecting all of us.
This is one (of many) advantages of open source software. If you have access to the source, you can build it on a ridiculously wide range of hardware. I've been downright shocked over the years at some of the strange things people have run actually useful (and entertaining) open source software on. Doom runs pretty much everywhere just for one example. ;)
also, the unix dependcy graph is formidable and needs to be supported to just bootstrap.
AFAIK Loongson's ISA is some sort of superset of MIPS64. Not exactly ubiquitous but MIPS64 has been around for a while so I would expect bootstrapping to be fairly simple.
Of course, bootstrapping is not the same as, say, using the parts of the superset that help with x86 emulation to speed things up, but even QEMU has supported LoongArch since 7.1 (over a year ago.)
The custom ISA becomes a real problem if you need Windows and other closed source software on it, but China has declared it political goal to rid itself its dependency on Microsoft.
LoongBT, faster x86 and ARM binary translation, 213 instructions
I don't think anyone is suggesting that China isn't capable of building an efficient and powerful CPU or not, but will they, at any point, catch up to contemporary CPUs from Intel, AMD, Apple, etc.?
I'm not saying they won't deliver useful silicon but they won't be delivering comparable silicon. And it'll be under trade embargo no doubt due to the IP issues so they will have a very limited market to scale into.
This happened before (GDR) and the products were never competitive.
I think the best thing they have done though is kick out compatible ATmega parts and the STM32 parts. They're really killing it there.
IIRC, Loongson was a straight copy of MIPS with the names changed (in the first few iterations, at least).
This was created for their own domestic market. China's goal is achieve independence of foreign technology. With 1.3 billion people they have a very large market.
This is also called learning by doing. This is how skill and expertise is most efficiently developed in the very beginning.
This trope of assuming that copying is inherently valueless or devoid of creativity or innovation is simply wrong, and comes from an implicit assumption of superiority of the self.
They may not innovate as much as you expect now. But they are building a foundation for future innovation.
Also: Huawei is still leader in 5G technology, with the best products. You should look at their patents, they own the majority of 5G patents.
GigaDevice and WCH have completely new MCUs now based on RISC-V. They keep the STM32 peripheral register interface for easy driver re-use, but they are well beyond mere copies these days.
https://wch-ic.com/products/categories/47.html?pid=5
https://www.gigadevice.com/product/mcu
Closest may be SMIC 7nm on multi patterning, which they cannot produce at scale and is still 2-3 generations behind TSMC [0]
Canon nanoimprint will not be available to China either.
[0] https://www.csis.org/analysis/contextualizing-national-secur...
But no one had ever beaten Intel in 50 years, until someone did.
There’s no monopoly on knowledge and I wouldn’t bet against the world’s second economy with its government directly financing a strategic sector.
"They can't buy machines from this Swiss company!" is not really much of a barrier when we look at the resources they have available. Over 12 months, that is a problem for them. Maybe 5 years. In a decade, ASML relying very heavily on the CCP shooting themselves and their people in the foot to still be a differentiator.
When talking about the sheer scale of 10% of the globe's population, you need a lot more than one technical dot point to say they can't do something.
Don’t disagree with your general point though. Also there is always the option of industrial espionage as we have seen.
Leaving the politics aside, it will be interesting to see what technical solutions emerge if the ‘easy’ option of spending a few hundred million dollars on a couple of EUV machines isn’t available.
And I'd assume the answer is spending a few billion dollars developing a couple of EUV machines over a long time. It isn't like a billion dollars is much money and the Chinese economy is reaching the point where people will be looking to muscle in to precision manufacturing. It isn't like cheap labour is an option any more. They've already overwhelmed the Europeans in a bunch of phone and computer related areas so there is evidence that the Europeans probably can't maintain a competitive edge if the Chinese want to get good at something.
Honestly, given what happened with the iPhone and Nokia I don't know how anyone is placing faith in the Chinese not being able to catch up. European tech companies have a history of looking unbeatable until they're just knocked completely out of the race. If anything the threat is that Europe won't be able to keep up with Asia.
China is not the first to take a crack at this. Attempts to replicate and catch up with American and European chipmaking and computing efforts go back to the dawn of the discipline, and had involved some of the largest economies of their times, yet here we are.
We now know it works. Before, why should China bother when they can just buy from western suppliers and TSMC instead? Now we gave China no other choice but to invest 10-20 years into this until they have something that works. This is why you see all the chip manufacturing equipment makers in China have something like 150-200% YoY growth in the past 3 years. Their R&D centers are pursuing multiple avenues to make sub-5nm chips possible. Not only multiple avenues of making EUV work, but also non-EUV avenues. It just so happens that starting this year we know that nanoimprint lithography works, and is cheaper than EUV. Huawei's newest CPU performs on par with CPUs on smaller nodes. That's because Huawei invested heavily in CPU design and heat dissipation technology. They haven't even gone into advanced packaging yet.
Will they stay behind western technology forever because we have an R&D lead? People who claim this are ignoring a big factor: they will no longer be able to profit from Chinese buyers. These profits are necessary for future R&D investment, which is necessary to maintain the lead and R&D speed.
That's debatable.
Samsung, a behemoth in Korea and one of the largest holdings in the world, has tried to compete with TSMC at the same node size, and failed so far.
> Will they stay behind western technology forever because we have an R&D lead? People who claim this are ignoring a big factor: they will no longer be able to profit from Chinese buyers.
Doesn't this work the other way around as well? Chinese companies in the same space will not be able to profit from EU and North American customers, historically the largest spenders.
> Doesn't this work the other way around as well?
It does. But the difference is:
- They're coming from behind and enjoy latecomer's advantage. They can take advantage of western R&D results. They may not be able to obtain all the details, but just knowing that a particular R&D path is viable is already an advantage.
- They have a captive market thanks to sanctions.
- They're coming from a smaller size. Chinese semiconductor companies had such small market share that capturing even a small part of only the Chinese market already translates to tremendous growth. They can only go up while the western supply chain will go down.
Finally, there's the global south which is also growing. This is a huge market that they can sell to. In the future, western suppliers will have to compete with Chinese suppliers in this market, even if Chinese suppliers are locked out of western markets.
That wording is a bit harsh. They haven't failed, but they haven't been able to compete at scale, which is what matters.
> They're coming from behind and enjoy latecomer's advantage. They can take advantage of western R&D results.
Isn't that the main issue, that they can't?
> They have a captive market thanks to sanctions.
That isn't true. Current sanctions target exports of certain consumer and professional products, none of which are chips as far as I understand; and imports of certain products branded strategic such as AI related hardware.
> They're coming from a smaller size. Chinese semiconductor companies had such small market share that capturing even a small part of only the Chinese market already translates to tremendous growth.
I don't think this means anything in this context. Capturing a larger chunk of a single market doesn't even mean that they are better than most at what they do, just that their go to market strategy is more effective than their competitors. SMIC could be releasing chips with comically large transistors, as long as their scale and prices are preferable to those of their competitors.
> They can only go up while the western supply chain will go down.
This is wishful thinking, at best. But again, growth means nothing here. SMIC could "grow" a 1000% for all I care, and it wouldn't move the needle.
> Finally, there's the global south which is also growing. This is a huge market that they can sell to.
That depends highly on China and their allies. China may capture markets in Africa, but it is actively alienating most of Asia Pacific, minus perhaps Russia.
That is commercially significant and entitles TSMC to bragging rights and huge profits. But if China fails like Samsung, they're at no strategic disadvantage at all and the US technology embargo will have effectively failed to do anything.
https://en.wikipedia.org/wiki/3_nm_process
China is in no position of "failing like Samsung", right now. At most, they may fail like Samsung did five years ago.
It would also mean that the sanctions worked.
They have spent significiant resources in hoovering as much information as possible from companies that have data of interest.
Cite: https://en.wikipedia.org/wiki/Allegations_of_intellectual_pr...
And now RISC-V is likely to pretty much wipe them out.
The issue is whether China will get behind the RISC-V stuff and contribute to a semi-unified ecosystem or whether the individual companies will fuck it up by balkanizing everything in an attempt to one-up their competitors. I'm not optimistic.
The West mostly doesn't really care about RISC-V. ARM licensing costs don't seem to be completely out of line unless you are talking super-cheap devices--which mostly come from China.
(A good counterexample is Nordic Semiconductor. They are having to compete against cheap Chinese BLE alternatives so they needed to start looking at RISC-V).
And my point is that if RISC-V achieves a significant amount of open designs and tooling then anyone who joins is roughly speaking guaranteed parity (in design - something like ISA and microarchitecture, not in foundry). Which is much better than the current situation.
"Still basically an Intel Core i3-10100 competitor"
Looking at the TDP, these seem more promising than other Chinese silicon designs which have been terrible flops.
Hummingbird was a bit of a quantum leap for ARM in my opinion. That said, I think Apple was able to do better with what they had thanks to their level of control over things (most notably, the way iOS suspends apps) makes it far easier to have a smooth user experience with a little less muscle.
I haven't looked at the TDP on this, one would probably want to normalize the measurements.