Author: lechuck park

AI Infrastructure Architect & Technical Visualizer "Complex Systems, Simplified. I translate massive AI infrastructure into visual intelligence." I love to learn computer tech and help people by the digital.

OSPF

Posted on by lechuck park

+ 224.0.0.6 (All Designated Routers) : Designated Router (DR) and Backup Designated Router (BDR). It is utilized to optimize communication between the DR and BDR, and regular OSPF routers do not receive messages from this address.

From DALL-E with Some prompting
The image is a visual representation of the operation of the OSPF (Open Shortest Path First) protocol. Here is the interpretation of each step depicted in the image:

get LS (Link State): OSPF routers collect cost values from all physically connected routers. This step involves determining the adjacency relationships between routers and the state of each link.

LSA (Link State Advertisement): Each router creates an LSA that contains its link-state information and disseminates it to other routers within the network. During this process, the multicast address 224.0.0.5 is used to broadcast the information to all OSPF routers.

LSDB (Link State Database): The information from the received LSAs is compiled into the LSDB of every OSPF router. This database should be identical across all routers within the Autonomous System (AS) and contains the complete topology information of the network.

Shortest Path Tree Calculation: Using the LSDB, each router calculates the shortest path tree from itself to all other destinations employing the Dijkstra algorithm. This calculation aids each router in determining the optimal routing paths.

Routing Table Update: The shortest path information calculated is then used to update the routing table of each router. This enables routers to forward packets using the optimal routes.

At the bottom, there’s a section titled Dynamic Updates, indicating that when there are changes in the network topology, new LSAs are generated and propagated through the network. This ensures that all routers’ LSDBs are updated and, as a result, the routing tables are also updated to reflect the new optimal routes.

In the top-right corner, it states “224.0.0.5 Broadcast IP for all OSPF router”, which indicates the multicast address used by all OSPF routers to receive LSA broadcasts.

This diagram provides a visual explanation of the core routing processes of OSPF, highlighting the mechanisms that enable efficient routing within the network and facilitate rapid convergence.


Start Regression (ML)

Posted on by lechuck park

From DALL-E with some prompting

Linear Regression:
Yields a continuous output.
Relates independent variable X with dependent variable Y through a linear relationship.
Uses Mean Squared Error (MSE) as a performance metric.
Can be extended to Multi-linear Regression for multiple independent variables.

Linear & Logistic Regression

  • The process begins with data input, as indicated by “from Data.”
  • Machine learning algorithms then process this data.
  • The outcome of this process branches into two types of regression, as indicated by “get Functions.”

Logistics Regression:
Used for classification tasks, distinguishing between two or more categories.
Outputs a probability percentage (between 0 or 1) indicating the likelihood of belonging to a particular class.
Performance is evaluated using Log Loss or Binary Cross-Entropy metrics.
Can be generalized to Softmax/Multinomial Logistic Regression for multi-class classification problems.

The image also graphically differentiates the two types of regression. Linear Regression is represented with a scatter plot and a trend line indicating the predictive linear equation. Logistic Regression is shown with a sigmoid function curve that distinguishes between two classes, highlighting the model’s ability to classify data points based on the probability threshold.

Processing UNIT

Posted on by lechuck park

From DALL-E With some prompting

Processing Unit

  • CPU (Central Processing Unit): Central / General
    • Cache/Control Unit (CU)/Arithmetic Logic Unit (ALU)/Pipeline
  • GPU (Graphics Processing Unit): Graphic
    • Massive Parallel Architecture
    • Stream Processor & Texture Units and Render Output Units
  • NPU (Neural Processing Unit): Neural (Matrix Computation)
    • Specialized Computation Units
    • High-Speed Data Transfer Paths
    • Parallel Processing Structure
  • DPU (Data Processing Unit): Data
    • Networking Capabilities & Security Features
    • Storage Processing Capabilities
    • Virtualization Support
  • TPU (Tensor Processing Unit): Tensor
    • Tensor Cores
    • Large On-Chip Memory
    • Parallel Data Paths

Additional Information:

  • NPU and TPU are differentiated by their low power, specialized AI purpose.
  • TPU is developed by Google for large AI models in big data centers and features large on-chip memory.

The diagram emphasizes the specialized nature of NPU and TPU for AI tasks, highlighting their low power consumption and specialized computation capabilities, particularly for neural and tensor computations. It also contrasts these with the more general-purpose capabilities of CPUs and the graphic processing orientation of GPUs. DPU is presented as specialized for handling data-centric tasks involving networking, security, and storage in virtualized environments.

N * same calc

Posted on by lechuck park

From DALL-E with some prompting
The image illustrates the evolution from traditional computational methods to AI, emphasizing how initially, human computations were limited by the inability to quantify all conditions, leading to varied outcomes. Through discussion, the most rational results were derived. In contrast, AI operates on big data (which is inherently limited since it is pre-defined data) and implements intelligence through an enormous number of identical calculations, showcasing the capabilities of AI to handle vast and complex datasets and perform sophisticated analyses leading to innovative solutions.

Event & Alarm

Posted on by lechuck park

From DALL-E with some prompting

The image illustrates the progressive stages of detecting alarm events through data analysis. Here’s a summary:

  1. Internal State: It shows a machine with an ‘ON/OFF’ state, indicating whether the equipment is currently operating.
  2. Numeric & Threshold: A numeric value is monitored against a set threshold, which can trigger an alert if exceeded.
  3. Delta (Changes) & Threshold: A representation of an alert triggered by significant changes or deviations in the equipment’s performance, as compared to a predefined threshold.
  4. Time Series & Analysis: This suggests that analyzing time-series data can identify trends and forecast potential issues.
  5. Machine Learning: Depicts the use of machine learning to interpret data and build predictive models.
  6. More Predictive: The final stage shows the use of machine learning insights to anticipate future events, leading to a more sophisticated alarm system.

Overall, the image conveys the evolution of alarm systems from basic monitoring to advanced prediction using machine learning.


Anomaly Connection Detection #0

Posted on by lechuck park

from DALL-E with some prompting
The image seems to illustrate the concept of anomaly detection in network security. A user is shown with a green line leading to a server, indicating normal interaction, while a red line leading from a network criminal suggests malicious activity. The network architecture is in place to mirror and tap into the data traffic, allowing for the steering of packets for closer inspection. An alert (!!) signifies the detection of an anomaly. Below, details of what is monitored are given: raw or sampled packets, TCP/IP 5-tuples, geographic IP locations, bandwidth, and new detection areas including DNS and HTTP header information. This represents a multifaceted approach to identifying and responding to potential security threats within a network.