Please enable JavaScript.

Coggle requires JavaScript to display documents.

SoC Interconnect (4. Network on chip (NoC) (Data traversal (a message is…

- - - - Difficulty in timing closure
      - All types of transaction are placed on the common bus. This reduces the amount of bandwidth available for any one specific type of activity, such as coherence or inter-processor messaging
    - - switching overheads (high impedance, tri-state, etc)
    - - Access right
        
        Each transaction is presented to all entities connected to the bus regardless of whether that entity is the transaction's destination or not. This means every transaction is also snooped by every processor on the shared bus whether it needs to be or not.
- - - - burst
      - stream
    - - high impedance, previous state
- - - - Each component can be scaled independently
      - Easy to assign and manage multiple clock and voltage domains
      - greater concurrency in application
      - Easy to build derivative chips
        
        To respond quickly to engineering change orders during development
      - Regular structure
        
        easy to understand control and data flow
        
        easy to verify and debug
        
        easy to layout
        
        easy timing closure
        
        increase maximum operating frequency
  - - - each packet has header information that allows the receiver to re-construct the original message
    - - flits do not contain additional headers
  - - - minimizing power required for routing
      - minimizing logic and routing tables to achieve a lower area footprint
      - increasing performance by reducing delay and maximizing traffic utilization of the network
      - improving robustness to better adapt to changing traffic needs
    - - Static
        
        fixed path are used to transfer data between a particular source and destination
        
        does not take into account current state of the network
        
        advantages
        
        easy to implement, since very little additional router logic is required
        
        in-order packet delivery if single path is used
      - Dynamic
        
        routing decision are made according to the current state of the network
        
        considering factors such as availability and load on links
        
        path between source and destination may change over time
        
        as traffic conditions and requirements of the application changes
        
        more resources needed to monitor state of the network and dynamically change routing paths
        
        able to better distribute traffic in a network
  - - - connects to the core
      - handles payload transfer to and from the core
      - implemented as sockets
        
        Open core protocol (OCP)
    - - handles network protocol
      - responsible for packet assembly and disassembly, synchronization etc
- - - - idle
      - non-sequential
      - busy
      - sequential
    - - slaves are accessed by multiple masters, have to keep track of who accessing the slaves and block others trying to gain access
        
        semaphore
    - - byte, half word, word, double word, etc
    - - slave inserts wait states if needs more time to provide or sample the data
    - - data/user/privilege access right
      - bufferable/non-bufferable
      - cacheable/non-cacheable
    - - pass the data in sequentially using 32 bits
    - - slaves need to have a buffer, once the buffer is full than execute
  - - - LSB stored first
    - - MSB stored first