“Intel Inside”? In Russia, the Chips are Down
Western sanctions have shut off Russia's access to the leading global chip designers and manufacturers. The longer the sanctions last, the more the Russian microelectronics industry will fall behind.
One of the most eye-opening results of the Ukrainian war has been the revelation of the role of foreign electronics in Russian weaponry. Several Western think tanks, notably the UK’s Royal United Services Institute (RUSI), working together with the Ukrainian military, have examined thousands of captured or downed Russian weapons. The findings are startling: every single one contains microelectronic components from leading Western and Asian firms, many of them supplied since the imposition of sanctions following Russia’s occupation of Crimea in 2014. While it is too early to tell whether the same pattern has continued since the February invasion, the study provides dramatic evidence of the long-standing Russian dependence on foreign microelectronics.
A LONG HISTORY OF BACKWARDNESS
The Russian microelectronics industry is minuscule compared to those of Europe and Asia, and its capabilities have long lagged far behind those of the world leaders. The reasons go back to the Soviet period. The Soviet command economy, until its collapse in 1991, was largely closed to the outside world. When the first transistors and integrated microcircuits sparked the computer revolution in the West from the 1960s through the 1980s, the Soviet Union remained on the sidelines. While the Soviets had world-class skills in physics, the bureaucrats who ran its military-industrial system restricted them to copying Western designs and products, which were acquired—despite Western export controls--by a worldwide network of spies (and the occasional turncoat). Thus, dependence on foreign technology was etched into the Soviet microelectronics industry from Day One.
A BRIEF HEYDAY, 1989-2014
But for a crucial quarter-century, following the break-up of the Soviet Union, Russia benefited from the era of good feelings that followed the end of the Cold War. The United States relaxed its export controls, applying them only to technologies with direct military uses. This reflected the optimism of the age: trade and innovation would generate growth and wealth; and these would trump geopolitics. This was particularly the case with microelectronics.
Meanwhile, however, another fundamental development, outside Russia, was transforming the prevailing business models and trade patterns of the microelectronics industry. This was the so-called “fabless revolution.” As historian Chris Miller recounts in his brilliant new book, Chip War, until the early 1990s most producers of semiconductors and processors combined both design and fabrication. But then a new player in Taiwan, TSMC, pioneered a model under which it would no longer design its own chips; instead, it would manufacture other companies’ designs under contract. This model transformed chip designers into TSMC’s loyal customers and, thanks to it, TSMC has built a commanding position in chip-making. TSMC now holds a virtual world monopoly, producing over 90% of the advanced chips in the world.
For Russia’s chip companies, the “fabless revolution” represented a golden opportunity. By signing manufacturing contracts with TSMC, the Russian companies could bypass their own weaknesses in manufacturing, and hope to produce competitive leading-edge microprocessors and servers, which in turn would open the way for modern data centers—in effect, a Russian “cloud.” It would continue to depend as before on foreign architecture and chip-making machinery, but the “fabless” model would enable the Russian chip designers to move forward quickly from design to mass manufacture. During this period, the Russian companies established working relations with leading chip manufacturers, especially TSMC.
DEPENDENCE DEEPENS
The new openness of the Russian economy, and the eased access to manufacturers, only deepened the dependence of the Russian microelectronics sector. This happened for several reasons. First, the newly-privatized Russian economy chronically underinvested in its own electronics industry. According to the Russian government, China invests $280 billion a year in computer-related electronics, the United States around $50 billion, but Russia only $1 billion. As a result, the entire Russian technological chain, from the manufacture of silicon chips through microprocessors to servers and data centers--not to mention finished computers and smartphones—remained weak.
The second cause was Russian customers’ rejection of Russian products. For about twenty-five years, Russia became part of the global supply chain in microelectronics and computers, and consumer products based on them, such as smartphones. Given a choice, both Russian companies and consumers preferred the foreign items, which were suddenly widely available. Civilians were not alone; the same, as the recent revelations from the battlefield show, was true of the military.
Lastly, the government as a whole paid little attention. It was only from the mid-2010s, as the geopolitical atmosphere turned cold, that the leadership became concerned about dependence in microelectronics. Government ministries began churning out ambitious plans to displace foreign products, chiefly by requiring Russian technology in state contracts. But these programs had little success: They were unpopular with both the Russian electronics companies and their customers, and even state agencies largely side-stepped the government’s orders.
FEBRUARY 24 2022: THE CHIPS ARE DOWN
The problems intensified in 2022. After the unprovoked attack on Ukraine in February 2022, TSMC abruptly severed all connections with Russian companies. The effect of sanctions has been to drive Russian semiconductor companies into the “grey market” or to outright contraband, especially for goods destined for the military or the security services. As early as the 2000s, as East-West relations cooled and the US revived its export-controls, the Russians increasingly turned to front companies and covert intermediaries which bought chips from US producers and passed them on to Russian middlemen. As the profusion of Western components found in Russian weapons in Ukraine has shown, such channels can continue to supply the Russian military.
But that is hardly a satisfactory solution. First, this illicit flow contains many defective chips. The leading foreign “fabs,”, above all TSMC, made their reputation by working hard to increasing the “yield,” that is, the percentage of non-defective chips that come off their assembly lines. But barred from direct access to manufacturers such as TSMC, Russian designer/manufacturers have discovered that up to 40% of the chips they buy on the open market are defective. Sorting the good ones from the bad takes time and raises costs.
Yet the real problem with the grey market and illicit purchases is that it they are necessarily limited to yesterday’s technology. But for the Russian economy to be competitive, it requires direct access to the companies that are at the leading edge today, driving the continuing headlong revolution in IT. The sanctions have abruptly interrupted that access; in this long-term perspective, the sanctions promise to have a profound and steadily worsening effect. To see why, let us take a look at the fast-moving frontier of transistor miniaturization.
THE NEW LANDSCAPE: UTRA-SMALL TRANSISTORS
To handle the torrents of data generated by modern IT, companies and governments need ever more advanced computers with millions of transistors packed ever more closely together. The separation between two transistors is measured in “nanometers,” or billionths of a meter. State-of-the-art chips today are on the verge of reaching separations of only 2 or 3 nanometers, compared to several hundred a mere decade ago. These will be the basis of the next generation of smartphones, servers, and other applications. But there are only two companies in the world that are capable of achieving this breakthrough—TSMC and Samsung—and they in turn depend on the small handful of partners that supply them with architecture (mainly Arm), photolithography (mainly ASML), and production machinery (such as Applied Materials).
The Russian chip companies are so far limited to much greater widths. “Mikron,” Russia’s most advanced microelectronics company (co-owned by the giant military-industrial conglomerate Rostekh), can produce small quantities of chips in the 90-to-180 nanometer range. But its main product is a processor with a 250-nanometer node, for use in bank cards and passports. Russia’s newest producer, Akvarius, is similarly handicapped.
22 nanometers turns out to be a crucial Rubicon. Crossing it requires a completely new design concept—3-dimensional chips (called “FinFET”), coupled with “extreme ultraviolet” photolithography equipment, plus a new manufacturing “fab” to match. TSMC, thanks to massive investment starting a decade ago, successfully crossed the 22-nanometer divide, with Samsung right behind, to reach today’s 2- and 3-nanometer chips. This is the cutting-edge technology that will enable the fifth generation of applications, known as “5G.” But without licenses to these new processes, and with no access to the “fabs” that will produce them, the Russians are stuck.
SUMMING UP: THE THREE-TIERED IMPACT OF WESTERN SANCTIONS
To sum up, the impact of Western sanctions on the Russian microelectronics industry can be divided into three tiers: first, the near-term effects on the current generation of Russian weapons; second, the mid-range effects on design and production in both military and civilian sectors; and third, the longer-range impact of sanctions on Russia’s entire IT chain, mainly civilian, ranging from chips to servers and data centers, and ultimately on the Russian “cloud.”
In the first category, Western sanctions are already too late. The current generation of Russian weapons, as the findings from the Ukrainian battlefield show, depend on older Western chips and other components that are widely available on “grey” markets, or are covered by licensing agreements that were concluded before the Crimean sanctions in 2014. In both cases, however, Russian weapons producers will be increasingly limited to obsolete designs using these simpler products. This is meager comfort, to be sure, since such weapons, though not leading-edge, can be devastating on the battlefield or as part of a “scorched-earth” strategy. (The Iranian-produced drones being used in Ukraine, though hardly advanced in their control systems, demonstrate the point.)
But Western sanctions will have little effect in stopping this flow; there are simply too many sources of supply. The Western powers, with both primary and secondary sanctions, will be able to stamp out individual illicit traders and jobbers, but such “whack-a-mole” efforts are unlikely to have more than marginal effects on the overall traffic in this category, as new players will constantly appear.
In the second category, Western sanctions are already having a substantial impact. The current Russian microelectronics industry has some decent design capabilities, but only weak fabrication capacity. During the brief “heyday” described above, Russian chip designers began turning to foreign “fabs,” while underinvesting in their own production facilities. The post-invasion sanctions have abruptly interrupted this emerging chain, cutting the access of Russian designers to foreign foundries such as Samsung or TSMC. As a result, Russian designs such as “Elbrus” and “Baikal” were left in the lurch, having lost access to production abroad, while having only limited production capacity at home.
In this second category, then, Western sanctions stand a good chance of being effective for the next several years. Russian policy-makers have responded to the cutoff with an emergency program to increase fabrication capacity by strengthening the connections between Russian “fabless” designers and “semi-fab” producers. But increasing fabrication will take large-scale investment—and time. They will now have a captive domestic market, inasmuch as their customers, both private and state, will no longer be able buy foreign products. But considering the length of time it took Asian fabs to tool up, progress in Russian fabrication will be at least a decade away.
Finally, in the third category, the impact of Western sanctions is likely to be crippling. There are only a handful of US and Asian designers and producers that are capable of developing the necessary architecture, advanced lithographic equipment, and chip designs, for the next generation of semiconductor systems—and even fewer that are able to combine them into actual mass production. Below 90 nanometers, the Russians have neither design nor production capacity. No amount of crash government programs will enable them to overcome a gap that took decades of misguided government policy to create.
But for how long? Apple has recently announced that it will use TSMC’s 3-nanometer chip in its new iPhones and Macs, starting in 2023. Once that happens, the latest chips will be incorporated in millions of new phones and laptops all over the world, which the Russians will be able to buy through networks of “parallel imports.”
Game over? Not at all: By that time, the race will have moved on, and the Russian economy, including the military sector, will still be a decade or more behind…
___________________________________________________________________________
Thane Gustafson is the author and co-author of eight books on Russian affairs, including most recently Wheel of Fortune: The Battle for Oil and Power in Russia (2012), The Bridge: Natural Gas in a Redivided Europe (2020), and Klimat: Russia in the Age of Climate Change (2021), all with Harvard University Press.
Some good analysis here. Thank you. Coming out of the university in the late 1980's one of my first jobs was with a company that worked to import high technology products & processes from the USSR & EEuro. We had business exchanges with some high-end Soviet chip manufacturers. It soon became clear to us & our Soviet contacts how far off the pace the Soviets were in chip manufacturing. (Fiber optic tech was a different story).
As your report briefly touches on, there is a whole ecosystem behind the public face of chip manufacturing. This report in the FT is sobering analysis for those who want to reshore, or go it alone in chip manufacturing https://on.ft.com/3zCckNA.
Also, software is a crucial part of the efficiency & effectiveness of information technology. However, there is a limit to how much software can overcome hardware deficiency.
Western sanctions on Russian energy are a self-inflicted wound with disturbing global implications - and limited short/medium term impact on Russia. However, western technology sanctions will have a rapid & debilitating impact on Russian domestic development - generally unappreciated by the pre-internet elements of the Kremlin elite.
This is a good summary of the situation but suffers a bit from an overly pro Western wishful thinking perspective. Why is it necessary to describe the Russian intervention in Ukraine as 'unprovoked'? First this is an article about micro chips not whether or not Russia was justified in invading Ukraine. It is a complicated question. Then you say ' both private and state, will no longer be able buy foreign products.' Really? China? India? Turkey? Brazil? Russia will have problems buying from the collective West, that is all. It is a notable feature of this conflict that the US has totally failed to force the whole world into blocking sales to Russia.
As to how long it will all take for Russia to correct its undoubted shortcomings in this sector, you may be right that it will be 'ten years or more...' but I doubt it. Catching up is always easier than pioneering. Remember what happened with nuclear weapons. Russia will be working closely with China. Finally there is the possiblility of a yet unknown, unimagined, technical alternatives being invented to bypass the present near monopoly in Taiwan. Nature abhors a monopoly. The Soviets put the first man into space. Russia has developed hypersonic weapons. There is every reason to think that they will work round this problem one way or another and hopefully the war will be over long before then.