RECENT
ARHITECTURE
TRENDS
PRESENTED TO:
DR. SAIMA FARHAN
PRESENTED BY:
MARYAM JAMIL
UM E HABIBA
MARYAM RAZZAQ
01
01 08 EDGE COMPUTING
INTRODUCTION
c
09 CLOUD COMPUTING
02 MULTI CORE PROCESSOR
ontent MEMORY CENTRIC
03 MANY CORE PROCESSOR 10
s ARHITECTURE
HETEROGENEOUS DOMAIN SPECIFIC
04 11
COMPUTING ARHITECTURE
APPROXIMATE 12 CONCLUSION
05
COMPUTING
NEUROMORPHIC
06
COMPUTING
07 QUANTUM COMPUTING
01
INTRODUCTIO
N
• Computer architecture is the fundamental
design and organization of a computer system.
• It dictates how components like the processor
(CPU), memory, and input/output (I/O) devices
interact.
• Think of it as the "blueprint" for building a
computer.
01
WHY ARE TRENDS
IMPORTANT ?
Increased Performace: Enabling us to tackle
complex tasks more quickly.
Save Energy: Better for the environment.
New tech: Enable amazing things like AI and VR.
MULTI CORE PROCESSOR
Refer to processor design that incorporate
multiple individual processor cores on a
0 single chip.
Each core function as an independent
2 processor capable of executing it's own set
of instructions concurrently
01
WORKING OF MULTI CORE
PROCESSOR
The processor chip houses two or more identical
processor cores that operate independently but
share the same memory hierarchy.
Different cores can execute separate threads or
tasks simultaneously, enabling true parallel
processing and improved overall system
performance
02 ARHITECTURE OF MULTI CORE
PROCESSOR
02
EXAMPLE
S
02
AMD Ryzen 9:
A high-performance
Intel Core i7:
01 processor with up to
A high-performance
16 cores.
processor with up to 10 0 ARM Cortex-A53:
cores.
3 A low-power
processor with up to
8 cores.
02
DISADVANTAG
ADVANTAGES
ES
1. Parallel processing: Allows 1. Amdahl's law: Some tasks are
multiple cores to execute inherently serial and cannot
different threads or tasks benefit from multiple cores
simultaneously, improving overall limiting in potential speed up for
system throughput and certain applications.
performance. 2. Shared Resources: Cores in
multi-core architectures share
2. Efficient Multitasking: With the same memory hierarchy and
multiple cores, a multi-core
processor can handle multiple interconnect, which can lead to
applications or tasks efficiently contention for shared resources
without sacrificing performance and potential bottlenecks.
MANY CORE PROCESSOR
0 A many-core processor is a type of processor
that contains a large number of processing
3 cores, typically ranging from tens to hundreds
or even thousands of cores.
03
Example: 1
Intel Xeon Phi
• Up to 72 cores
• Used for scientific computing,
simulations, and big data processing.
01 02
Example: 2
NVIDIA Tegra Series
• Multiple cores for mobile and embedded
devices
• Efficient in handling graphics, gaming, and
complex tasks on mobile platforms.
03
01 02
ADVANTAGES DISADVANTA
GES
1. Faster: More cores = quicker work, 1. Tricky to program: Needs
especially for big tasks. special software to use all
2. Efficient: Less power per core can save the cores.
energy overall. 2. Not always faster: Some
3. Multitasking: Runs many programs jobs can't be split up easily.
smoothly at once. 3. Hot: More cores = more
heat to manage.
HETEROGENEOUS
COMPUTING
Heterogeneous computing refers to a system that
0
employs multiple types of processing units, each
specialized for specific tasks. This approach
4 contrasts with traditional homogeneous systems,
which rely solely on CPUs.
04 ARHITECTURE OF HETEROGENEOUS
COMPUTING
04
DISADVANTA
GES
• Complex System Design
B
• Software and
Programming Model
Complexity
ADVANTAG A • Resource Management
ES and Scheduling
• Enhanced Performance
• Improved Energy Efficiency
• Flexibility and Scalability
APPROXIMATE COMPUTING
Approximate computing in computer architecture means
0 designing computers to purposely allow small errors in
calculations to make them run faster and use less power. Instead
of making sure every calculation is perfect, it focuses on getting
5 results quickly and efficiently, even if they aren't 100% accurate.
This approach is useful for tasks where small mistakes don't
matter much, like video processing or some types of AI.
05 ARHITECTURE OF APPROXIMATE COMPUTING
05
ADVANTAG DISADVANTAGE
ES S
• Faster Computations • Less accurate
• uses Lower Power • Unpredictable Results
• Good for some tasks: (like • Not Suitable for All
video streaming or some Applications
types of AI)
NEUROMORPHIC
COMPUTING
Neuromorphic computing is a type of computer design that
tries to make computers work like the human brain. It uses
0 special hardware that mimics how brain cells (neurons) and
connections (synapses) process information. This makes the
6 computer better at tasks like learning, recognizing patterns,
and making decisions, just like our brain does.
06 ARHITECTURE NEUROMORPHIC OF
COMPUTING
06
ADVANTAG
ES
• Efficient for AI tasks
• Low power usage
• Faster learning
DISADVANTA
GES
• Hard to build
• Not fully understood
• Not suitable for everything
QUANTUM COMPUTING
Quantum computing in computer architecture is a new type of
computing that uses the principles of quantum mechanics (the
0 science of very tiny particles like atoms) to process information
in a completely different way than regular computers.
7 How it works:
Instead of using bits (which are either 0 or 1), quantum
computers use qubits that can be both 0 and 1 at the same
time. This allows them to solve certain problems much faster
than traditional computers.
07 ARHITECTURE OF QUANTUM COMPUTING
07
DISADVANTAG
ADVANTAGES
ES
• Super fast for certain
tasks • Hard to build
• Handles huge amounts • Expensive
of data • Not useful for
• Could revolutionize everything.
technology
EDGE COMPUTING
0 Edge computing is like having mini-computers closer to
where data is collected (like sensors or cameras) instead of
8 sending all that data to a far-off big computer (the cloud).
This makes things faster, more reliable, and better for
privacy.
08 ARHITECTURE OF EDGE COMPUTING
08
ADVANTAGES DISADVANTAGES
• Faster Processing • Limited Power
• Lower Bandwidth Use • Harder Management
• Better Reliability • Scalability Issues
CLOUD COMPUTING
Cloud computing in computer architecture
0 refers to the practice of using remote
servers on the internet (the "cloud") to
9 store, manage, and process data, instead
of relying on local servers or personal
computers.
09
DISADVANTAGE
ADVANTAGES
S
• Scalability • Security Concerns
• Cost-Effective • Internet Dependency
Add your words here,
• Accessibility • toOngoing
according your need toCosts
• Automatic Updates draw the text box size.
Please read the instructions
and more work at the end of
the manual template.
MEMORY CENTRIC
ARHITECTURE
Memory-centric architecture in computer
1 architecture refers to a design where memory
(RAM) plays a central role in the system's
performance, and the architecture is optimized to
0 maximize the efficiency of memory usage. In this
design, the processing elements are designed to
work closely with memory, reducing the need to
move data frequently between memory and
processors.
10
DISADVANTAGE
ADVANTAGES
S
• Faster data access • Memory Bottleneck
• Efficient for Big Data • Costly
• Reduced Data Delays • Complexity
DOMAIN SPECIFIC
ARHITECTURE
1 Domain-specific architecture (DSA) in computer
architecture refers to designing specialized hardware or
systems tailored for a specific application or task, rather
1 than using a general-purpose processor. For example,
creating custom chips or systems optimized for tasks like
image processing, artificial intelligence, or cryptography.
11
DISADVANTAGE
ADVANTAGES
S
• Better Performance • Limited Flexibility
• Energy Efficiency • Development Time
• Lower Cost • Higher Initial Cost
12
CONCLUSION
Computer architecture trends focus on improving
performance, energy efficiency, and scalability. Key
developments include multi-core processors, AI and
GPU specialization, and integration with cloud and
edge computing. Emerging technologies like
quantum computing and neuromorphic systems
promise to push the boundaries of computation,
enabling faster, more intelligent, and efficient
systems.
THANK YOU!
