The Irtysh microprocessor: a Russian chip from China

Source: cnews.ru
Irtysh equals Loongson
Tramplin Electronics received 1,3 billion rubles to create a Russian server processor. It sounds like a breakthrough. But let's put it more simply. What are we actually getting? A chip built on the Chinese LoongArch architecture, with LA664 cores developed by the Chinese company Loongson. The specifications are practically mirror images of their Chinese counterparts. The 16-core Irtysh C616 is Loongson, the 32-core C632 is also Loongson, and the 64-core flagship C664 is again Loongson. The only Russian components are a few security modules, a cryptographic accelerator, and the microchip housing. Essentially, Tramplin Electronics is acting as a licensee and integrator, customizing the Chinese base to suit its needs, rather than creating the architecture from scratch.

Source: telesputnik.ru
Microprocessors have always been a problem in modern Russia. Leaving aside the Soviet period, the first chips appeared immediately after the country's collapse in the early 90s. More precisely, development began on the basis of Soviet technology. Then came the Elbrus, a microchip with a unique architecture that in some respects even outperformed its Western counterparts. Then came new generations, each one becoming more powerful. There were also other projects, such as the Baikal, which used the popular ARM architecture.
There was essentially one problem: all of this was manufactured abroad, primarily in Taiwan. And after 2022, when the West imposed sanctions, this option was closed. The country found itself in a situation where critical systems—banking, energy, government—were running on Intel and AMD processors, and components for them were no longer being supplied. At least, not through official channels.
The development of a domestic microprocessor has become a matter of national security—and nothing less. In parallel with the chip's development, the complex process of creating a Russian photolithograph—or, in simpler terms, a machine for printing microprocessors—is underway. While we're still waiting for the photolithograph, our own microchip is almost ready. The extent to which it's truly Russian is discussed above.
The most interesting thing is that at the end of 2022, Beijing declared Loongson processors strategically important and banned their export. A year later, Russian engineers somehow obtained a license to develop the architecture and launched their own line. How this happened is a closely guarded secret. The terms of the license are not disclosed. But the fact remains: we have replaced dependence on the West with dependence on the East. Is this good or bad? So far, so good. China is a partner. But a partner with its own interests. And in stories There have been examples of technological alliances where friendships ended abruptly and painfully. The Tramplin team are former engineers from the Moscow SPARC Technology Center, the same ones who developed Elbrus.
The investor is Omsk businessman Svyatoslav Kapustin, founder of Tramplin Ventures. He decided to invest in Russian chip production, hoping they would provide a viable alternative to unaffordable Western solutions. And his calculations appear to be correct: laws require import substitution, the lists of chips approved by the Ministry of Industry and Trade and the Ministry of Digital Development are growing, and any Russian chip is guaranteed to find customers in banks, telecoms, energy, and government agencies.
The Irtysh's specs are impressive: 16 to 64 cores, up to 128 threads, up to 2 TB of RAM, and built-in encryption. The chip can handle complex calculations—roughly on par with processors from Intel and AMD. Manufacturers claim their chip performs as well as popular competitors, even though each core is slower individually. But in servers, this isn't important—they prioritize multiple cores working simultaneously, not just one core being as fast as possible.
Tests have shown that a server powered by this chip easily handles large user flows and responds quickly to requests. This is perfectly adequate for banks and large IT systems. If anyone's wondering, the Irtysh isn't a gaming processor—the screen image will be very choppy. But that's not what the chip was designed for.
Technological breakthrough or trap?
The schedule for introducing the Irtysh processor to the masses is tight: an application to the Ministry of Industry and Trade's registry is scheduled for the second quarter of 2026, engineering samples are expected in the second quarter of 2027, market launch is expected in 2027, and the international market launch is expected in 2028. 1,3 billion rubles should be enough for preparatory work and the production of several thousand processors. Converted to dollars, that's about 15 million. For a processor project, that's not much. By comparison, Intel spends about 20 billion dollars annually on development. Therefore, it's safe to say that Russia is getting a very good deal on the Irtysh processor.

The main problem with the Irtysh is its production. The chips will almost certainly be made in China at a SMIC plant; in Russia, they are only assembled into packages, and that's a small part of the work. The most advanced Russian process technology is 65 nanometers, a level that's been around for twenty years. The Irtysh, on the other hand, was created in China using a 12-nanometer process. Russia simply doesn't have the photolithography equipment for the corresponding class. While the crystals are being made in China, Tramplin Electronics only oversees the final assembly. Full localization will require tens of billions of rubles and the development of a domestic process technology to at least 28 nm. This will take years.
Irtysh also has a unique feature: it's not directly compatible with Windows. Enthusiasts tested the chip on The Witcher (which runs on Windows), but this required some clever tricks. Ultimately, the processor will perform worse than advertised on the familiar Russian Windows operating system. The lack of Windows support is Irtysh's main obstacle to mass adoption. For decades, corporate Russia has relied on Windows: accounting databases, document management, specialized software for banks, factories, and hospitals. Migrating all this to the unfamiliar Linux platform is not just difficult, but impossible for many companies.
Without Windows, the processor loses access to a vast universe of programs. Millions of applications—from the familiar 1C and Office to highly specialized engineering suites—were created specifically for this OS. Adapting them to another platform would require colossal expenditures, which no one would bear for the sake of a negligible market. As a result, Irtysh users will find themselves in a closed world with a limited software selection, where half of their vital tasks simply won't have a solution.
But there's another aspect to independence from Windows. Microsoft isn't just an operating system manufacturer; it's a tool of influence. With sanctions changing monthly and major IT companies already blocking licenses for Russian users, relying on third-party software is extremely dangerous. If a processor runs Windows, it can be rendered useless at any time: through an update, through license revocation, or through remote blocking.
This won't happen with Linux and domestic operating systems—the code is open, and decisions are made here, not at headquarters on the other side of the world. For banks, government agencies, and defense companies, this isn't a matter of convenience, but of survival: we control what runs on our servers, and no one can shut it down. Plus, this is a powerful incentive for the Russian IT industry. As long as Windows is available, developers have no incentive to create alternatives—why invest if the market is saturated? But when Windows is unavailable, demand arises. This means that programmers will emerge who will learn to write for other platforms. Companies will emerge that will build their businesses on this.
The Irtysh developers, albeit on a Chinese platform, deserve congratulations on their initiative. One can only hope for a lasting geopolitical balance between Beijing and Moscow and the rapid technological development of their own production base, not just the packaging line for foreign-made crystals.
Information