Skip to Content

Why China’s Homemade Microchips Will Struggle to Displace Western Giants

Loongson is making steady progress toward a viable microprocessor, but its chosen architecture could keep it from supplanting its rivals.
December 19, 2012

If China’s ultimate aim in the sphere of technology is to become completely self-sufficient, it is well on the way to achieving this ambitious goal.

Central processor: Loongson’s Godson 3A chip.

Huawei and ZTE, both based in Shenzhen, supply much of the globe’s telecommunications infrastructure. Lenovo, another Chinese company, is the world’s number-one PC maker and creeping up the smartphone charts. And Chinese Web giants like Alibaba, Baidu, and Tencent lead the domestic market in their respective fields.

One area, however, is still missing from the grand self-sufficiency plan: a world-class rival to microprocessor companies like Intel, AMD, and ARM. That’s why the buzz is increasing around Loongson Technologies and its Godson chips, the latest of which is set to make a splash at the International Solid-State Circuits Conference in San Francisco next February.

Loongson is a partnership between Beijing-based chip designer BLX IC Design Corp. and the Institute of Computing Technology at the Chinese Academy of Sciences. With government funding, the project began in 2001, using the MIPS instruction set from Stanford-based MIPS Technologies—an architecture that’s different from Intel and AMD’s x86 or ARM’s designs. Swiss electronics manufacturer ST Microelectronics is used for fabrication. MIPS-based processors have a global market share of little more than zero in these categories, although this jumps up to around 9 percent when categories like set-top boxes and digital TVs are included, according to market research firm IDC.

Loongson’s first effort, produced in 2001, was a 32-bit CPU running at 266 megahertz; this was followed in succeeding years by several versions of a 64-bit chip called Godson 2, based on the same architecture but running at faster speeds. Progress culminated in the one-gigahertz, single-core Godson 2-H, which was designed for netbooks, mobile devices, low-powered PCs, and embedded systems.

So far, Loongson chips have found their way into very few products in the traditional PC or server category, with the exception of the odd model launched by Chinese firm Lemote, itself a spin-off of the Institute of Computing Technology, little known French vendor EMTEC, and one or two others.

Loongson has had marginally more success in supercomputing. Godson designs are already used in Sunway BlueLight MPP—China’s first home-grown supercomputer—and 3,000 Godson 3-B chips planned for the delayed Dawning 6000 supercomputer. Jennifer Bernier-Santarini, a spokeswoman for MIPS Technologies, argues that Godson chips “have an increasing opportunity in servers and high-performance computing.” The MIPS architecture is “very streamlined and elegant compared to the competition. With fewer instructions and less complexity, it can lead to better performance efficiency,” she says.

But Loongson still lags far behind U.S. chip giant Intel in the consumer electronics market. The 1.35 gigahertz, eight-core Godson-3B1500 chip, which will be on show in San Francisco next year, has features 32 nanometers in size and provides 172.8 gigaflops of performance while drawing 40 watts of power—impressive power saving stats for a CPU. However, the 32-nanometer architecture is a generation behind Intel’s newest Ivy Bridge processors, which are not only the smallest around at 22 nanometers, but use 3D Tri-Gate technology to reduce power consumption further. Intel’s 14-nanometer technology is reportedly on track for production by the end of 2013. 

Software could also be a stumbling block. Because the Godson chips are based on a different architecture from most of the PCs sold worldwide, the main operating system for them is Linux. That said, the new 3B1500 chip benefits from x86 emulation, which means it can run Windows.

But IDC program director Shane Rau says “emulation has had very little success in the marketplace” due to the processing inefficiency that results. “At worst it’s a marketing gimmick and at best a bridge into the applications of another [platform],” he says.

Other experts contacted by MIT Technology Review were pessimistic about Loongson’s chances of success outside of government-supported projects. Frost & Sullivan analyst Jun Chen argues that commercial success for Loongson will remain elusive because the stability of its products hasn’t been demonstrated yet in enough implementations. He says Huawei, whose Hi-Silicon business is ramping up production of ARM-based chips, may offer China a greater chance of processor self-sufficiency precisely because there are more commercial opportunities for it.

Tom Hackenberg, principal analyst at IHS Technologies, suggests that Loongson’s original choice of MIPS could backfire. “They have little competitive value in the commercial arena,” he says, adding that the new Godson 3B-1500 will do little to change this. Hackenberg adds that Godson simply isn’t innovative enough to disrupt the market and provide “a compelling reason enticing a hardware designer to risk isolating themselves in an uncommon ecosystem.”

IDC’s Rau agrees that a “compatible base of software” is essential to the success of the project, but is more optimistic that China’s hefty state backing could make the difference. Large government contracts for cloud and high-performance implementations could give the Godson architecture a push, he says, spurring efforts to build out the ecosystem over time.

“At the start of the effort you need a central plank of control to develop compatible software; it’s essential down the road. If tools are put in place—compilers, libraries, etc.—by these high-performance monolithic efforts, then these tools could be applied to diversification efforts down the line,” says Rau.

The key to success going forward, however, will be in Godson’s ability to scale up and down, to find its way into high-volume, low-power devices (smartphones) at one end and low-volume, high-performance applications at the other. This is something Intel, ARM, and AMD have already worked out.

Processor self-sufficiency may be China’s ultimate goal, but it is also likely to invest in other architectures to spread the risk. Even so, given time, Loongson could yet become a real competitor to Intel, AMD, and ARM.

IDC’s Rau notes: “Godson has been around about 12 years. Other implementations of processors have surely been born, lived, and died in shorter periods, which is the benefit of the government backing.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.