DATA PATHS AND CONTROL UNITS
The Data Path, also known as the Arithmetic Logic Unit (ALU), is responsible
for executing
arithmetic and logical operations on data. It performs calculations and data
transformations,
such as:
1.Arithmetic operations (addition, subtraction, multiplication, division)
2. Logical operations (AND, OR, NOT)
3. Bitwise operations (shifting, masking)
Data Path
This refers to the route data takes through hardware/software components to
perform graphical computations (e.g., transforming vertices, applying
textures).
Core Components of a data path
Functional Units:
These are the components that perform specific operations on the data, such
as ALUs for arithmetic and logic operations, multipliers, or other specialized
units.
Registers:
These are small, high-speed storage locations that hold data temporarily
during processing.
Buses:
These are communication channels that allow data to be transferred
between different parts of the data path.
DATA PATH SERVES DUAL ROLES
1. Bottleneck Source:
o Slow or overloaded data paths (e.g., memory bandwidth,
CPU→GPU transfers) that create performance limits.
o Example: High-res textures clogging memory bandwidth
→ lag.
2. Optimization Target:
o Improving data paths (via techniques like
instancing/mipmapping) removes bottlenecks.
o Instancing → Fewer CPU→GPU calls (optimizes command
path).
o Mipmapping → Better texture memory use (optimizes
texture data path).
Control Unit
Control units manage the flow of data and synchronize operations across the
GPU/CPU. The control unit (CU) is a component of a computer's central
processing unit (CPU) that directs the operation of the processor. A CU
typically uses a binary decoder to convert coded instructions into timing and
control signals that direct the operation of the other units (memory,
arithmetic logic unit and input and output devices, etc.).
Most computer resources are managed by CU. It directs the flow of data
between the CPU and the other devices. John von Neumann included the
control unit as part of the von Neumann architecture.[1] In modern computer
designs, the control unit is typically an internal part of the CPU with its
overall role and operation unchanged since its introduction.[2]
Key Functions:
1. Instruction Execution → Covers fetching, decoding, and executing
(Points 1–3, 5).
2. Data Processing → Uses decoded instructions + registers/memory
(Points 3–4).
3. Memory Management → Fetching/storing data/instructions (Points 1,
4).
4. Input/Output (I/O) → Implied by control signals/data movement
(Point 3, extended beyond CPU).
5. Parallelism: Control units schedule shader threads to keep thousands of cores busy.
6. Synchronization: Prevents race conditions (e.g., two shaders writing to the same pixel).
Key Control Units:
How Data Paths + Control Units Work Together
1. CPU issues a draw call → Command Processor schedules it.
2. Vertex data flows through shader cores → Memory Controller fetches
textures.
3. Fragments are shaded → ROP units resolve visibility (depth test).
4. Final pixels are dispatched to the Display Controller.
How Data Moves through Data Paths / Relationship
The Data Path and Control Unit work together to execute instructions:
1.The Control Unit fetches and decodes instructions.
2. The Control Unit sends control signals to the Data Path.
3. The Data Path performs arithmetic and logical operations.
4. The Control Unit manages data transfer and storage.
Importance:
The Data Path and Control Unit are crucial for
1. CPU performance
2. Instruction execution
3. Data processing
4. System efficiency
Key Considerations for Data Path Design
Efficiency:The data path should be designed to minimize the time it takes to
perform operations and transfer data.
Flexibility:The data path should be able to handle different types of data and
operations.
Resource Utilization:The data path should be designed to make efficient use
of hardware resources.
