Original | Odaily Planet Daily ( @OdailyChina )
Author: Golem ( @web3_golem )
Every time a Web3 narrative explodes, it not only brings wealth effects to users, but also opportunities to entrepreneurs. Meme is the most watched narrative in this cycle, and the wealth effect it brings far exceeds that of traditional altcoins. As a result, a large number of players migrated from the secondary market and centralized exchanges to the chain, and P Junior, Chain Front, and Dev became new user labels.
For entrepreneurs, in this round of Meme craze, the user demand they face is fast and safe Meme trading tools, so this has undoubtedly become the main battlefield for competition, and excellent Meme trading tools such as GMGN, OKX Wallet, Bullx and Axiom have been born. In the Chinese-speaking region, the most popular Meme tool is undoubtedly GMGN, which had a monthly income of over 10 million during its peak period, and even exceeded Uniswap in a single day .
However, as GMGNs business grew, the huge storage capacity and computing requirements caused by the sharp increase in the number of users and transaction volume had a great impact on the AWS Aurora database that GMGN originally used, which in turn hindered business growth. You know how many Meme trading tools have been killed on the beach because they could not bear the huge user traffic and data service pressure.
Ultimately, GMGN chose to migrate data from AWS Aurora to the open source distributed SQL database TiDB, which effectively improved the scalability and computing power of the system. During the peak period of the Meme market, it was still able to quickly process on-chain data and respond to user growth. While increasing business agility by 30%, it also reduced operating costs by 50% compared to before.
Today, GMGN supports multi-chain Meme transactions, including Solana, BSC, Ethereum, Base, Tron, and Blast. According to Dune data, GMGNs fee income on the Solana chain alone has exceeded US$71 million, with 743,000 users. GMGNs successful experience can also provide reference for entrepreneurs in other tracks of Web3.
1. GMGN’s Advanced Path
2024 is the year of Meme explosion. Pump.fun unified the issuance standards of Meme coins, allowing Meme coins to be continuously produced in the form of factory assembly lines. According to Dune data, Pump.fun has issued more than 10 million Meme coins, and its cumulative fee income to date has exceeded US$680 million. The explosion of Pump.fun has also attracted many entrepreneurial teams to develop downstream Meme trading tools.
But counterintuitively, GMGN was actually born earlier than Pump.fun. Back in 2023, the Solana Meme coin market had not yet exploded, and only Ethereum occasionally produced some Meme coins. However, GMGN founder Haze had already sensed the huge business opportunities that Meme coins might bring. It is low-cost, highly efficient, and can quickly issue assets based on hot spots and sentiment. This is not just a sector of the blockchain, but also the ultimate embodiment of asset issuance methods.
As a result, a tool focusing on on-chain transactions and with data visualization capabilities, GMGN, was born in June 2023. GMGN initially cut into the Ethereum chain, but the effect was not obvious. In order to expand the market, Haze has since then considered expanding to chains such as Solana, Blast, and Base.
Finally, the opportunity came in 2024. With the birth of Pump.fun and the popularity of BOME in March 2024, the Solana Meme coin market surpassed Ethereum and ushered in exponential growth. GMGN also launched Solana Meme coin transactions and on-chain analysis in April 2024, successfully seizing this opportunity. In November 2024, GMGNs data ushered in the first wave of major outbreaks, with a daily peak transaction volume of more than 500,000 times, and then ushered in the second wave of data outbreaks in January 2025, with a daily transaction volume of 230 million US dollars.
Although most of GMGN’s users are born in the 2000s, Haze was born in the 1980s. Before founding GMGN, Haze was a quantitative grid trader. This experience not only gave him a deeper understanding of the blockchain industry, but also gave him a deeper understanding of “speed” due to his trader’s qualities.
GMGN is actually designed to solve a pain point, for speed, Haze said in a keynote speech titled GMGN is born for speed at the 2025 Hong Kong Web3 Carnival. He specifically divided speed into four types, namely fast discovery, fast decision-making, fast operation and fast execution.
What really made GMGN popular was its pioneering rat warehouse function, which solved the most concerned issue of users in the process of dog-buying, exposed the conspiracy group, and really helped users make quick decisions. Since then, GMGN has continuously enriched the dimensions of data, adding new labels such as diamond hands, paper hands, VC, and smart money, and at the same time, it has tied the on-chain address to the social account of KOLs, so that the community can have an objective judgment on the dog-buying ability of various KOLs.
Nowadays, many Meme trading tools are replicating GMGN’s functions and completing better iterations. Faced with the latecomers, GMGN, which has been developing for two years, does not seem overly anxious, but continues to serve its loyal community users, while actively participating in the construction of the public chain in its own way. We are public chain builders, and we made 60 million US dollars on Solana, but we didn’t sell a single one, and we pledged all of it, Haze said proudly on the stage.
2. TiDB builds the Web3 data base for GMGN
As the Meme coin market became increasingly popular, GMGN also faced pressure. GMGNs server was overwhelmed 6 times in 4 days, Haze said in an interview with Wu in November 2024, talking about the problems caused by user and transaction growth.
The database is the power behind the operation of App/dApp, so ordinary users do not have direct access to it. In the early days of GMGN, it mainly used AWS Aurora database to store a large amount of real-time transaction records and smart contract data, and then integrated and pre-processed the data through applications.
However, as the business expands, facing the huge growth of Web3 data and user activities, GMGN faces the following database challenges:
Poor scalability
With the rapid increase in Web3 data and the number of users, GMGNs database at the time was difficult to expand effectively. According to Dune data, the number of users and transactions on the GMGN platform ushered in the first explosive growth in November 2024, with the daily transaction number reaching a maximum of 495,000 and the number of active addresses reaching nearly 300,000; the second explosive growth came in January 2025, with the daily transaction number reaching a maximum of 782,000 and the number of active addresses increasing to 430,000.
However, as shown in the figure above, the on-chain market is also cyclical. AWS Aurora mainly relies on vertical expansion for expansion, which cannot adapt to the rapid changes in on-chain user traffic and data processing needs. It is easy to cause the database to be paralyzed during peak trading periods and idle during low trading periods, which also increases operating costs.
Poor data query performance
Web3 data is not only large in volume but also needs to be updated in real time. A complete Solana node or Ethereum node needs to store 1-2 TB of data. In particular, GMGN involves multiple chains, which makes data query more difficult. Due to the huge load, general databases such as AWS Aurora perform poorly in supporting complex queries, and a system that can quickly and accurately handle large amounts of data queries is urgently needed.
Poor agility
The complexity of the database architecture leads to high development and maintenance costs. GMGN itself is forced to form a data analysis team with big data processing and multi-blockchain expertise to maintain multiple database instances. This not only increases the teams operating costs, but also seriously damages GMGNs agility as a startup company.
At the same time, the key to Web3 data services is the need to ensure that data is always accessible and not unavailable due to any single point of failure, which forces GMGN to maintain multiple database instances, but this also undermines the agility and adaptability of GMGNs business.
TiDB helps GMGN improve scalability and reduce costs
Faced with the above challenges, GMGN spent two weeks integrating all Aurora MySQL databases into a cloud solution called TiDB, which helped improve scalability and reduce operating costs.
TiDB is an open source distributed SQL database developed by PingCAP in 2015. Simply put, in the Web3 field, TiDB can efficiently support a large amount of on-chain data storage, OLTP queries and real-time analysis to support the fastest acquisition of on-chain data insights, and provide efficient, reliable and easy-to-manage database solutions for data-based platforms.
Simplify the original architecture and reduce operating costs by 50%
First, TiDB helped GMGN simplify the database architecture. In the original architecture, different chain data (such as ChainA, ChainB or ChainC) were stored in separate MySQL databases and then processed by the data platform. In TiDBs data architecture, all data from different chains are directly stored in TiDB, as shown below.
The benefit of doing so is that GMGN no longer needs to maintain multiple independent databases, which reduces the complexity of operation and maintenance and reduces GMGNs operating costs by 50%. At the same time, because it has a unified data storage layer, GMGN only needs to import newly generated blockchain data into TiDB in real time, which makes it easier for GMGN to achieve data consistency and transaction management.
Provide horizontal and elastic expansion capabilities
As a distributed database, TiDB has good horizontal scalability (adding nodes to handle more load) helping GMGN to scale seamlessly and handle rapid data and user growth without affecting performance.
In addition, TiDBs elastic scalability is also very consistent with the rapid changes in user traffic and data processing requirements on the chain. In TiDB, computing resources and storage resources can be expanded independently, and computing and storage can be expanded or reduced online as needed, so as to flexibly respond to traffic peaks or data surges.
HTAP function significantly improves GMGN data query capabilities
Faced with the problems of on-chain real-time query capabilities and complex query support, TiDBs HTAP function has significant advantages. This function supports real-time online transaction processing (OLTP) and online analytical processing (OLAP), meeting the needs of real-time data writing and analysis of blockchain data.
In actual operation, GMGN only needs to use some simple SQL queries to easily meet the requirements of big data processing, eliminating the worries about data synchronization, cleansing and other processing issues. The data query response time for normal business queries has been improved to an average of single-digit millisecond latency compared to before the migration.
Improve business agility by 30%
TiDBs improved architecture means that the GMGN team no longer needs to maintain multiple separate databases. TiDB automatically maintains multiple data copies and has automatic fault detection and recovery functions, providing GMGN with enterprise-level high availability. At the same time, TiDB is a fully managed SaaS service, which does not require GMGN to intervene in database maintenance and operation issues. With the professional support of PingCAP, GMGN can quickly solve problems and ensure fast and stable business operations.
This provides GMGN with extremely reliable system stability, allowing it to focus on products and improving user experience, and increasing business development agility by 30%.
3. TiDBs Web3 data service blueprint
Of course, TiDB is not customized for GMGN. In fact, many well-known Web2 and Web3 companies have used TiDB services, including Meituan, Xiaohongshu, Xiaomi and China Merchants Bank, and Web3 includes a global leading DeFi data platform, Chaintool and a leading Asian decentralized wallet. TiDB has received more than 37,000 stars on GitHub, and is becoming an important data service provider in the Web3 field.
TiDBs Web3 Service Cases
Looking at the entire Web3 industry, whether it is DEX, on-chain tools or games, they all face problems such as storage pressure, computing challenges and insufficient complex data processing capabilities. Open source distributed SQL databases such as TiDB are solving these pain points for Web3 project parties.
Project owners will not have to worry about whether the growth of on-chain node data will reduce user experience; whether the surge in user access frequency and data will cause server downtime; whether the failure of a single node will cause business stagnation, etc. TiDB is like a flexible and controllable foundation, allowing Web3 entrepreneurs to focus more on iterating functions and experiences.
References
Full text of Hazes speech at the Hong Kong Carnival: GMGN is born for speed
