ZTC OER Stipend
Permanent URI for this communityhttps://hdl.handle.net/1969.6/96014
Browse
Browsing ZTC OER Stipend by Author "Hadimlioglu, Ismail Alihan"
Now showing 1 - 13 of 13
- Results Per Page
- Sort Options
Item Amdahl's Law(2024) Hadimlioglu, Ismail AlihanIntroduction: What is Parallel Processing? Executing tasks simultaneously using multiple processing units. Benefits of Parallel Processing, Increased throughput (more tasks completed per unit time), Reduced execution time for specific tasksItem Branch prediction(2024) Hadimlioglu, Ismail AlihanThe Branching Dilemma: Conditional branches are instructions that alter the program flow based on a condition. Processors typically fetch instructions sequentially. Encountering a branch creates a dilemma: Fetch the next instruction in sequence (assuming not taken). Fetch the target instruction of the branch (assuming taken).Item Fundamentals of computer design: Advanced computer architecture(2024) Hadimlioglu, Ismail AlihanIntroduction: In this topic, we will explore: What is Computer Design, Historical Context, Key considerations that go into creating efficient & powerful computer systemsItem Instruction set architecture (ISA) and RISC-V(2024) Hadimlioglu, Ismail AlihanWhat is an ISA? The bridge between software (instructions) and hardware (CPU). Defines the set of instructions a processor can understand and execute. Dictates how data is processed and programs are run.Item Instruction-Level parallelism and pipelining(2024) Hadimlioglu, Ismail AlihanIntroduction: What is Computer Performance? Execution Time, Instructions Per Second (IPS), Clock Speed (GHz), Clock Cycles Per Instruction (CPI)Item Introduction to advanced computer architecture(2024) Hadimlioglu, Ismail AlihanIntroduction: What is Computer Architecture? The science and art of designing, structuring, and optimizing the core components of a computer system. Defines how instructions are processed, data is accessed, and system components communicate. Acts as the bridge between the software (instructions) and hardware (physical components). Provides a foundation for understanding the trade-offs between performance, power consumption, and cost in advanced systems. Enables an in-depth analysis of advanced techniques like pipelining, superscalar execution, and memory hierarchies. Forms the basis for optimizing software for specific architectures and exploiting their capabilities. Helps in appreciating the challenges and opportunities presented by emerging technologies like multicore processors and GPUs.Item Memory hierarchy design(2024) Hadimlioglu, Ismail AlihanIntroduction: What is Memory Hierarchy? A layered system for storing and accessing data in a computer, Organized based on speed, capacity, and costItem Memory optimizations(2024) Hadimlioglu, Ismail AlihanIntroduction: Importance of Memory, Bottleneck in performance, Large capacity gap between registers and main memory, Challenges of Memory Access, Latency (access time), Bandwidth (data transfer rate)Item Pipeline scheduling and loop unrolling(2024) Hadimlioglu, Ismail AlihanIntroduction: What is Instruction Level Parallelism (ILP)? The concept of pipelining in a CPU, Data dependencies and their impact on performanceItem Pipelining hazards and performance(2024) Hadimlioglu, Ismail AlihanPipelining Hazards: Structural Hazards: Insufficient Hardware Resources, Data Hazards: Reading data before it's written, Control Hazards: Branch instructions and stallsItem Power, energy, and cost in computer architecture(2024) Hadimlioglu, Ismail AlihanIntroduction: We will explore the critical aspects of power, energy, and cost in computer architecture. We will discuss the design considerations, components, and mathematical models used to optimize these factors in modern computing systems.Item Reorder buffer(2024) Hadimlioglu, Ismail AlihanInstruction pipelines: Processors execute instructions in a sequence of steps fetch, decode, execute, write back. In-Order Execution: Instructions progress through the pipeline one at a time, waiting for the previous instruction to finish before starting. Out-of-Order Execution: Instructions can be issued, executed, and completed out-of-order as long as there are no dependencies.Item Tomasulo algorithm(2024) Hadimlioglu, Ismail AlihanIntroduction: Instruction Level Parallelism (ILP): ILP refers to the potential to execute multiple instructions concurrently within a processor. Modern processors exploit ILP to achieve higher performance. The Tomasulo Algorithm is a key technique that facilitates dynamic instruction scheduling for maximizing ILP.