Kaisar Provider
Last updated
Last updated
Providers offer CPU/GPU physical devices for the Kaisar Network, also known as Containers. This is where the actual usage of the cloud takes place, including application execution and rendering. The purpose of Containers is to ensure a seamless experience in the cloud, delivering an optimal user experience.
Notes on Activities:
Availability: Containers must always be in a high state of availability, ready to be activated immediately upon consumer request.
Usability: Each Container must have the necessary applications or services installed and configured to allow users to access and start up as quickly as possible.
Processing capacity: Containers need to meet specific computing power and graphics requirements to handle the unique demands of applications or services.
Network efficiency: Containers must have stable bandwidth and network infrastructure to support high-speed data transmission and low-latency interactions.
Container Selection Process:
Performance-based: Containers are randomly selected based on their ability to provide the highest quality of service with the lowest possible latency and cost.
Experience optimization: Containers are evaluated for their ability to deliver the best possible user experience, considering factors such as processing speed and uptime.
Worker Node Provision
Steps
Provider provides computing resources as worker nodes that perform computational tasks.
Provider provisions a worker node.
Kaisar Network collects data from the node.
Worker node sends proof of provisioned work to checkers.
Checkers Validate PoPW and calculate points
Provider is rewarded based on points.
Worker Node Registration
Steps:
Provider registers a new worker node.
Registration request sent to Kaisar Network.
Console Portal responds with setup scripts.
Worker node runs setup scripts.
Worker node connects to a wallet.
Worker node requests a new peer node from blockchain.
Connection established.
Worker node executes a benchmark job.
Benchmark job result sent to Kaisar Network.
Worker node registered successfully.
Reward Calculation Mechanism for Providers
Factors
A (Renting Ratio): Ratio of hours rented to total time of 1 Epoch.
B (Bandwidth - BW): Node's network bandwidth score.
C (GPU Model - GM): GPU performance level score.
D (Uptime - UT): Ratio of uptime to total time of 1 Epoch.
Weights
WA = 0.25
WB = 0.25
WC = 0.3
WD = 0.2
Formula
Bandwidth Scoring System
GPU Model Scoring
Calculating Average Monthly Income for Providers
To calculate monthly revenue for each device, we use the following parameters:
Pool: Monthly bonus pool
Devices: Number of devices
Renting Price: Device rental price
Coverage: Coverage percentage
KAI Price: The price of KAI token
Revenue = (Pool + Coverage x Devices x Renting_Price/KAI_Price)/Devices
For example:
Pool: 1,620,000 KAI (50% Emission)
Devices: 100,000 H100 GPUs
Renting Price: $3,000/device
Coverage: 50%
KAI Price: $0.5
Revenue = (Pool + Coverage x Devices x Renting_Price/KAI_Price)/Devices
= (1,620,000 + 0.5x100,000x3,000x2)/100000 = 3,016 KAI ~ $1,508
Tier | Download/Upload | Score |
---|---|---|
Supplier | GPU Model | Score |
---|---|---|
Low Speed
100 Mb/s / 75 Mb/s
0.2
Medium Speed
400 Mb/s / 300 Mb/s
0.4
High Speed
800 Mb/s / 600 Mb/s
0.6
Ultra High Speed
1600 Mb/s / 1200 Mb/s
0.8
NVIDIA
GeForce RTX 1080 Ti
0.25
NVIDIA
GeForce RTX 2080 Ti
0.25
NVIDIA
GeForce RTX 3050
0.25
NVIDIA
GeForce RTX 3050 Ti
0.25
NVIDIA
GeForce RTX 3060
0.25
NVIDIA
GeForce RTX 3060 Ti
0.25
NVIDIA
GeForce RTX 3070
0.25
NVIDIA
GeForce RTX 3080
0.25
NVIDIA
GeForce RTX 3080 Ti
0.4
NVIDIA
GeForce RTX 3090
0.5
NVIDIA
GeForce RTX 3090 Ti
0.75
NVIDIA
GeForce RTX 4080
0.5
NVIDIA
GeForce RTX 4090
0.75
NVIDIA
RTX A100
2.5
NVIDIA
H100 PCIe
10