What's behind China’s Filecoin hype?

China has a sweeping plan to develop and build high-tech infrastructure during the next several years, from data centers and 5G networks to ultra-high voltage power lines and electric vehicle charging stations. 

And the Filecoin network?

On March 25, People’s Daily, the mouthpiece of the Chinese Communist Party (CCP), published a report entitled “Decentralized storage opens a hundred billion [dollar] market to deepen digital transformation.” The report made the case that distributed storage and decentralized cloud computing, serving as the critical infrastructure for “Web 3.0,” is poised to transform society and the economy in “profound” ways.

The People’s Daily article didn’t explicitly mention Filecoin. But it did cite the Interplanetary File System (IPFS), which was developed by Protocol Labs, the same team behind Filecoin. The newspaper said that IPFS is an example of a faster, safer and more stable network structure that could transform the use of digital data.

In China, the names Filecoin and IPFS tend to be used interchangeably. The idea behind Filecoin is to create an incentive mechanism on top of IPFS. Miners use the Filecoin protocol to store and retrieve data for users and receive FIL in return. 

The price of FIL, the native cryptocurrency of the decentralized storage network Filecoin, increased by over 30% the next day to above $100 and more than doubled to an all-time-highs over $230 in the following week before plunging to around $170 as of press time.

Chinese crypto traders have played a significant role during the price rally. The Filecoin custodian wallet of Huobi, one of the largest crypto exchanges catering Chinese traders, ranks third on the FIL rich list with 17.7 million FIL after the Filecoin Foundation and Protocol Labs.  

But Filecoin has yet to prove itself as a mainstream storage platform. So where is this narrative coming from? Perhaps that has more to do with Filecoin mining — and the price of FIL — than it does decentralized data storage itself.

“New Infrastructure”

Last year, China’s government unveiled what it calls the “New Infrastructure” plan, in which the nation has committed to spending $1.4 trillion on infrastructure development during the 14th five-year economic cycle, which spans from 2021 to 2025.  Digitalization of services is a critical part of this new plan, and one area of focus is “big data centers.”

Decentralized networks like IPFS and Filecoin are indeed a potentially new way to store big data. But in fact the technology behind Filecoin and IPFS has yet to prove itself at a large scale, and there are a number of competing decentralized storage protocols. So why would People’s Daily call it out by name? 

In fact, Filecoin mining has become a big business in China. The network and the cryptocurrency have also received massive amounts of hype here. 

Numerous articles, usually written anonymously, have been published in China during the past year, all of which have pushed the same general narrative: Filecoin is bound to see exponential growth and is thus a once-in-a-lifetime investment opportunity — in part because the New Infrastructure strategy will boost its real value to the digital economy.  

The second part of that story has been reinforced lately by certain government leaders who have also painted Filecoin and IPFS in a positive light. 

 In late March, RRMine, a long-established bitcoin mining firm based in China’s Sichuan province that is also a top Filecoin miner, hosted an IPFS industry summit in the city of Ya’an. At the event, RRMine touted its more than 370 pebibyte (PiB) of effective storage power on the Filecoin network, which accounted for almost 10% of the network’s total. 

The keynote speaker that RRMine invited was Gou Yiquan, the deputy mayor of Ya’an, located in the western part of Sichuan. The province already accounts for half of the total bitcoin network’s hash power. 

The local government of Ya’an appears to see Filecoin mining as a way to boost the utility of the city’s Western Sichuan Big Datacenter, a $700 million investment project that opened in 2019. The government is trying to attract more outside businesses to the area.  

Gou placed his remarks in the context of China’s 14th five-year economic plan. “The computing power is becoming the core driver of the digital economy and also a core advantage for national competition,” he said. 

He went on: “Ya’an is a main Chinese hydropower hub and turning green energy into computing services and digital assets is a main area for Ya’an’s future digital economy development.”

Gou’s remarks were similar to those made by a group of municipal government officials in Fuzhou, Jiangxi province, during a mid-February research tour of a local Filecoin mining datacenter owned by IPFSUnion, a provider of Filecoin mining equipment.

The officials included Xiao Yi, the Fuzhou municipal general secretary of the CCP, who is the effective chief of the city, as well as Fuzhou mayor Zhang Hongxing. 

Such research tours are a typical way for Chinese officials to learn from enterprises and to boost those firms' confidence and morale. But in Fuzhou’s case, it was also the local government’s investment that was at stake. 

According to IPFSUnion’s business record, the Fuzhou government’s state-owned capital is a minor shareholder in the firm. IPFSUnion boasted in a blog post in February that through this public-private partnership, the city of Fuzhou is poised to have “the country’s largest IPFS decentralized storage industrial park.”

The plan, according to IPFSUnion, is to have both production factories that can manufacture 100,000 Filecoin miners annually and datacenters that can house 100,000 to 150,000 servers. “We hope IPFSUnion can stick to its core technology and grab shares in the competitive global market and contribute to China’s ‘New Infrastructure’ national strategy,” Xiao said during the tour. 

Random data 

While that vision may be grand, the reality is that Filecoin and IPFS are storing hardly any useful data. And it’s not clear that there is much demand for decentralized data storage. 

As of writing, Filecoin’s effective storage power — predominantly contributed by Chinese Filecoin miners — is 4 exbibyte (EiB). That’s like 4.8 million one terabyte hard drives, each capable of storing roughly 250,000 photos taken with a 12-megapixel camera, or 500 hours of high definition video.

Does that mean that 4 EiB worth of real digital memories produced by human beings is now stored by Filecoin miners? The short answer is no.

Xie Jinbin, a Filecoin ecosystem investment manager at Huobi, wrote a research piece in late February that explained how the network's underlying economic dynamics have meant that most of the data stored on Filecoin — at least for now — consists of just random files and very little useful piece information.

To mine FIL, first miners must demonstrate to the network that they have “effective storage power” by “sealing” (or encoding) data and pledging FIL. Simply having one tebibyte (TiB) of empty storage space produces nothing. Only a miner seals one TiB of data can they demonstrate that amount of effective storage power — which, as of press time, is enough to mine 0.0890 FIL in 24 hours. 

The faster Filecoin miners can seal data — a pace that depends to some degree on their CPUs and software optimization — the quicker they can grow their effective storage power. But the question is: where is that data going to come from?

That’s up to the miners. They can either seal real data that they deem useful or they can seal randomly generated data. According to Xie, the random data approach can add effective storage power as much as 10 times faster.

Xie added that the other reason Filecoin miners are opting for random data is that Filecoin mining power has been oversold in China. In the months leading up to the mainnet launch of Filecoin, Chinese Filecoin miners had not only been accumulating equipment but also selling Filecoin cloud mining contracts to retail investors with a promise of rewards per TiB of storage power sold. 

According to Xie, the randomly-generated data does have some utility in that they provide the network with the security and stability that it needs in its early stages. Xie explained that this is part of Filecoin’s “chicken and egg problem.” The network needs enough stored data to ensure its security. “But without enough demand for real data, how to ensure the network is secure?”  

Xie added: “Filecoin is based on the demand for decentralized storage but is that demand really sound to begin with?” 

It's not for the lack of trying by Filecoin. The project's Slingshot Competition awards a total of 50,000 FIL to miners storing 10 pebibytes (PiB) of data that meet certain “qualifying” standards. So far, the data stored via the Slingshot competition is at 6.5 PiB, accounting for 0.14% of the total network’s total storage capacity

Protocol Labs has also rolled out the Filecoin Plus program, which introduces yet another economic incentive: the program will reward miners with 10 times more FIL if they choose to seal real data.