Non-Uniform Memory Access

Posted on by lechuck park

From DALL-E with some prompting
The image depicts the NUMA (Non-Uniform Memory Access) architecture in computer systems. Key elements include:

  1. Operating System: Manages and controls processes running on the CPU.
  2. CPU: Central Processing Units where computing tasks are executed.
  3. NUMA Nodes: Guide CPUs to use the nearest memory, with each NUMA node having memory areas closer to specific CPUs.
  4. Memory Access Paths: “Short Path” indicates a fast and low-energy memory access that is closer, while “Long Path” represents a slower and more energy-consuming memory access that is farther away.

The structure illustrates that memory access times in a NUMA system are not uniform across all memory, suggesting that memory access optimization can enhance overall system performance.


Requires for DL

Posted on by lechuck park

From DALL-E with some prompting
The image outlines the importance of data in the era of deep learning (DL). It starts with “Data,” representing various sources and types, which feeds into “Deep Learning,” depicted by a neural network diagram. The process leads to “Result,” symbolized by charts and graphs indicating the output or findings. The central message, “Data determines the results,” stresses that the quality of data significantly impacts the outcome of deep learning processes. Below, “Data Verification” suggests the need for ensuring data accuracy, which ties into the cycle of “UPDATE” and “Analysis,” highlighting an iterative process to refine and improve deep learning applications. The phrase “What to deal with DL” hints at the challenges and considerations in managing and utilizing deep learning effectively.

Overall IP Network

Posted on by lechuck park

From DALL-E with some prompting
The image is a diagram explaining the overall structure and data flow of an IP network.

Overall IP network: The entire structure of an IP network
Ethernet In the LAN: Ethernet used within the Local Area Network (LAN)
Identification in the internet: Identifying devices on the internet
OSPF short path with IP addresses: Open Shortest Path First (OSPF) routing protocol finds the shortest path using IP addresses
Addressing/Routing to the peer: Assigning addresses and routing to peer devices
BGP to get/share IP (other & me): Border Gateway Protocol (BGP) is used for obtaining and sharing IP addresses between others and oneself
Service Connection: Establishing a service connection
IP address ↔ Domain address: The relationship between IP addresses and domain addresses
DNS Easy to keep an internet address by Domain name: Domain Name System (DNS) makes it easy to maintain an internet address by using domain names
On TCP/UDP: Operating on TCP (Transmission Control Protocol) and UDP (User Datagram Protocol)
The diagram illustrates how data moves within a network. For instance, when a user accesses web services using the HTTP protocol, the DNS translates domain names into IP addresses, and then a service connection is established using the IP address over TCP/UDP protocols. Routing protocols such as OSPF and BGP are used to find the optimal path for data transmission through internal networks and the wider internet, respectively.


Unexplainable

Posted on by lechuck park

From DALL-E with some prompting
The image intends to explain two critical perspectives of AI/ML. First, it illustrates that while traditionally digitalized data was defined by rules, AI/ML enables us to judge human ‘feelings’ as data based on a more extensive dataset. Second, AI/ML allows for the prediction of the future using data; however, some parts of these significant advancements remain unexplainable and difficult for humans to comprehend fully. This interpretation suggests that while AI aims to quantify and use non-visible elements like emotions for predictions through data standardization and optimized processing, there are aspects that cannot be fully articulated or understood.

Data Standardization

Posted on by lechuck park

From DALL-E with some prompting
The image emphasizes the importance of data quality in the digital transformation of large-scale operations. By securing “Data Quality” through data standardization, optimized operations based on verified data enable reliable decision-making, monitoring, and optimization. AI-enhanced analysis and optimization accelerate business transformation, drive data-led innovation, and achieve sustainable operation and customer satisfaction.

  1. Data Standardization: Emphasizes the importance of “Data Quality,” indicating that high-quality, standardized data is foundational.
  2. Operation based on verified data/system: Shows the use of verified data to ensure reliable decision-making, monitoring, and optimization, leading to sustainable operations, business intelligence, and customer satisfaction.
  3. Accelerating (AI) digital business transformation: Describes how optimized and customized processing, along with an AI data analysis platform, can accelerate digital transformation. This leads to work automation, user customization, resource optimization, data-driven innovation, AI predictions and analytics, and expanding standardization.

The overall message suggests that standardizing data quality is crucial for building AI systems that can drive digital transformation and improve business operations and customer satisfaction.

On-device AI

Posted on by lechuck park

From DALL-E with some prompting
The image is a diagram explaining the concept of “On-Device AI,” which describes the process of operating artificial intelligence within a device. The stages are as follows:

Data: It begins with the collection of data from large databases and the internet, represented by a “Big data” icon and various icons representing different internet services.

Machine Learning: The collected data is used to train models through the machine learning process, depicted by a neural network icon.

Model: The trained model is represented by an AI model icon, including learned features or vectors, indicated by the term “Learned.”

Optimized Data: The trained model is transformed into optimized data for use in on-device AI.

On Device AI: The on-device AI operates using an inference engine and a dedicated inference chip, supporting AI functionalities on the end-user’s device, such as a smartphone, as illustrated by a device icon.

The image represents the flow from data collection, through model training and optimization, to the execution of AI within a device. This process allows AI to function independently within a personal device rather than on cloud servers, providing benefits such as reduced response times and enhanced privacy protection

The time is

Posted on by lechuck park

From DALL-E with some prompting
The image conveys that innovation is more than just seeking new things; it is achieved through numerous changes performed within a given absolute time. The upper section shows three change attempts within a limited timeframe, suggesting that evolution occurs through these trials and failures. The lower section emphasizes the capability to undertake nine changes in the same time period, highlighting faster and more substantial evolution. It underlines the idea that digital transformation enables more rapid and extensive changes than those made by humans, which is a crucial element in driving innovation and evolution.