Author: J1N, Techub News
Preface
The Kuzco that the author shares today is a network dedicated to serving the LLM large language model computing power mining network. This year, it was selected for the a16z Crypto Startup Accelerator (CSX) Fall Accelerator Program launched in New York. Projects selected for this program will receive at least $500,000 in investment from a16z within 8 weeks, and will receive guidance and support from the a16z operations team. Currently, there are about 2 weeks left before the end of this program.
The Kuzco project is still in a very early stage, and belongs to the GPU computing power network field like io.net, but it is taking a different development direction. However, due to its strong financial support and background, io.net has completed the airdrop to users earlier than Kuzco, and has successfully launched on mainstream cryptocurrency exchanges such as Binance, Coinbase, and Bybit, gaining greater market exposure.
The author has personally participated in the mining of io.net. To be honest, the whole process is like "earning money while the graphics card is idle". You only need to hang the graphics card, and it will automatically generate income for you, and you don't even spend much on electricity. Each mining machine consumes only 1.2 kilowatt-hours per day, and the total electricity cost for three months is only $10.8. Adding the depreciation of second-hand graphics cards and the cost of internet tools, the cost of 100 to 200 dollars can be easily handled. Finally, a profit of $4,000 was obtained, and no matter how you look at it, this project is like a "castle in the air" that spirals upwards with the left foot on the right foot.
The subsequent decline in token prices has verified this. In addition, the author has also compiled an article titled "Price Plunges 70%: How the AI Computing Power Rental Bubble Burst?", which further reflects the bubble in the AI computing power market. Whether it is the price of the rental market or the decline of the project tokens, they are the market's correction of the real value of these projects. This price adjustment is in line with market rules, indicating that the future may develop in a more rational direction.
Kuzco has significant differences in the mining mechanism from the previous popular project io.net. io.net is a "castle in the air", relying on hanging the graphics card to easily obtain dozens of times the income, and the actual value behind it is not obvious. Kuzco, on the other hand, is more down-to-earth and has the support of a16z. I spent a few weeks carefully testing mining rigs with different configurations, including single-card, multi-card, and different system setups. Through these practical operations, I will share these practical experience with you, so that you can understand a mining project that provides real value to users, rather than a false prosperity supported only by bubbles.
A network dedicated to serving the LLM large language model computing power mining
Kuzco is a decentralized GPU network in the Solana ecosystem, aiming to utilize the idle GPU resources of individual users to provide efficient and inexpensive computing power services for large language models (such as Llama3, Mistral, Phi3). Users can use these models through APIs compatible with OpenAI. Kuzco's distributed architecture can integrate idle computing power resources to enable the operation of large language models, while incentivizing miners who provide computing power through a reward mechanism.
Project Operation Status
As of October 21, the Kuzco project has 2,000 online GPU miners, with a peak of 6,000 at one time. The most commonly used GPU models include 3090 and 3060. Miners are currently receiving KZO Point rewards, but these points cannot be cashed out yet, and the project has not yet announced its token economic model. The author speculates that with the progress of the a16z accelerator program, the project may have more new developments and updates in the future.
How to Deploy
Official Hardware Requirements
The Kuzco project can run on Mac, Windows, and Linux operating systems, and supports a variety of different hardware configurations. The minimum system requirements are 16GB of RAM, 30GB of available disk space, and at least 10MB/s of network bandwidth. Kuzco supports NVIDIA (N-card) and AMD (A-card) graphics cards with at least 8GB of video memory. The minimum requirement for ordinary users' N-cards is GTX750, while most professional computing power graphics cards are supported.
However, the author does not recommend using A-cards for computing power mining, because AMD graphics cards have poorer compatibility, especially in their performance in AI tasks compared to N-cards. If you must use A-cards, you need to check the official compatibility support list. In the field of AI, the author must say: AMD, NO!
The author's five-card platform, with a Huazi Z490 motherboard
The author has tested the mining efficiency of several hardware, but due to network fluctuations, the efficiency performance may fluctuate.
GTX1070: 20 tok/s
RTX2060: 30 tok/s
RTX2070S: 40 tok/s
RTX3080: 80 tok/s
RTX4060Ti: 50 tok/s
RTX4070S: 70 tok/s
M2: 20 tok/s
M3: 30 tok/s
Unit: Avager Tokens / Second (Average Tokens calculated per second)
Real-time monitoring of mining machine operation
Deployment Method
Kuzco provides a client application, and users can download and use it to start mining, but this method is sometimes unstable and may disconnect, and it will not automatically restart. The author suggests a more stable approach is to use a Linux system or the Windows WSL (Windows Subsystem for Linux) environment, and start mining through the command line (CLI) or Docker containers. If you need to use multiple graphics cards, you can specify a GPU in Linux through Docker containers to perform multi-card mining. For example, to specify GPU0: "docker run --rm --runtime=nvidia --gpus '"device=0"' -d kuzcoxyz/worker:latest --worker <workerid> --code <code>", you can start multiple GPUs.
When mining with multiple cards, please note the following hardware device hardware level:
Power Supply: The power supply is crucial and cannot be skimped on. It is recommended to purchase according to the standard of "1 RMB = 1 Watt", and use gold-rated power supplies as much as possible. When mining, choose a 1500W to 2000W power supply according to the number of graphics cards, or use multiple power supplies, but multiple power supplies need additional startup lines connected to the motherboard to be able to power on normally. Otherwise, it cannot start normally.
Power consumption information
Cables: Under high-power conditions, power cables are easily damaged, so the author recommends using higher-quality power cables. In addition, different brands of modular power supplies use different cable interfaces, which are not interchangeable. If you use cables of different brands, it may cause device damage or burnout. Therefore, the cables must be consistent with the brand of the power supply to ensure compatibility.
Aging and damaged power cables
Motherboard: Each channel of the motherboard (whether x1, x8, or x16) can only support one graphics card. For example, you can support as many graphics cards as there are channels. The B85 motherboard popular during the Ethereum mining era is a good choice.
The B85 platform the author participated in Ethereum mining, which has been scrapped
CPU: Because multiple tasks need to be processed at the same time, the more CPU threads the better. When starting mining through Docker, it will initially consume a large amount of CPU resources. If using multiple graphics cards, they must be started in sequence, otherwise the system may crash or freeze.
Young people should use i9, there is no need to use e5 for the sake of face
Graphics card: A Docker process takes up about 6 GB of video memory (the official document says 8 GB, but in fact 6 GB is enough to run), if the graphics card has 12GB of video memory, you can run two Docker processes on one graphics card. During mining, the graphics card's workload will occupy 50% to 90%, and the graphics card's temperature should be kept below 85 degrees, which is reasonable and safe.
PCIE adapter cable: It is recommended to use a x1 to x16 PCIE adapter cable, with x1 inserted into the motherboard and x16 inserted into the graphics card. If using a 40 series graphics card, you need to use a x16 extension cable.
Network connection quality: The quality of the network connection will have a great impact on the mining efficiency. The author's test found that the computing power tasks received from the Singapore network node are more than the Hong Kong node, which means that choosing a better network node can improve the mining efficiency.
If a fault occurs during the mining process, you should first check the running status of the motherboard and software. If the problem is at the software level, you can judge it by checking the terminal error information, it may be a memory error, which requires a reboot; or it may be because the official file is updated, and the local mining machine has not updated the code in time, the solution is to change the node or update the code.
Running error
At the hardware level, if the mining machine cannot be powered on, first check the fault light on the motherboard. Taking the ASUS motherboard as an example, the most common one is the white VGA light, which indicates that there is a problem with the graphics card power. In this case, you can try to re-insert the PCIE and graphics card power cables. However, sometimes the white light may be on but the machine is still running normally.
Summary
Although the author calls io.net a "castle in the air", meaning that its market value is severely overestimated, it has indeed successfully raised $40 million with a valuation of $100 million. However, after io.net went online, there were many imitations, and the products and backers of these imitations could not withstand strict verification, indicating that the success of io.net is not something that every project can easily replicate.
Based on this observation, the author has been looking for a mining project with more practical value. Finally, the author found Kuzco. Kuzco first received incubation and support from a16z, which increased its credibility and potential. Secondly, Kuzco's mining mechanism actually provides computing power services through GPU.
More importantly, from a macro perspective, the field in which Kuzco is located, large language models (LLM), is the most widely used AI product by the general public, and countless people around the world use these models every day, and these models require huge computing power support. The author believes that providing computing power for such a huge demand is not only very meaningful, but also has practical commercial value, so Kuzco is a project worth paying attention to.
Also, the cost of participating in the GPU computing power network is relatively low, especially the stable price of second-hand 40 series graphics cards in the market, with low depreciation costs, so it is worth buying. However, the author reminds you not to choose to rent graphics cards, because the rental fee is much higher than buying second-hand graphics cards. At the same time, Kuzco's airdrop incentives are not clear, and if you rashly rent a large number of graphics cards, there will be a higher risk.
And the mining machine itself has a very high practical value, not only can it be used to mine Kuzco, but it is also more stable and reliable than directly investing in high-risk Altcoins. Due to the good scalability of the mining machine, in addition to mining Kuzco, it can also be used to mine other GPU projects, or used as a node verifier in the blockchain, running some scripts and services to generate more income. This means that even if you no longer mine Kuzco, the mining machine can still continue to generate value.
Finally, many people ask how much profit can be made from mining per day, but there is no definite answer to this question. Except for the project party, no one can accurately know the final profit. The return on mining has great uncertainty, it may have very high returns, or it may be lower than expected, so it is impossible to determine in advance how much money can be earned.
