Math Co-processor

and
Multi-core Processor
Course Teacher:
Md. Obaidur Rahman, Ph.D.
Professor
Department of Computer Science and Engineering (CSE)
Dhaka University of Engineering & Technology (DUET), Gazipur.

Course ID: CSE - 4503

Course Title: Microprocessors and Assembly Language
Department of Computer Science and Engineering (CSE),
Islamic University of Technology (IUT), Gazipur.
Lecture References:
 Book:
 Microprocessors and Interfacing: Programming and Hardware,
Author: Douglas V. Hall
 An Introduction to Parallel Programming, Author: Peter
Pacheco

 Lecture Materials:
 M. A. Sattar, BUET, Dhaka, Bangladesh.

2 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Math Co-processor
 Intel family of math coprocessor are generally labelled
as 80x87, which support 80x86 processors in
computing.
 Instruction sets and programming of all devices are
almost identical
 Intel family of math-coprocessors is able to
 add  find square root
 subtract  calculate partial tangent
 multiply  calculate partial arctangent
 divide  logarithms
3 CSE-4503: Microprocessors and Assembly Language
Islamic University of Technology (IUT)
Intel Processors and Co-processors

Processor Corresponding Co-processor

8086/8088 8087
80186/80188 80187
80286 80287
80386 80387SX, 80387DX
80486SX 80487SX

4 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
CISC and RISC Processors
 CISC (Complex Instruction Set Computers):
 CISC was developed to make compiler development
simpler.
 It shifts most of the burden of generating machine
instructions to the processor.
 For example, instead of having to make a compiler write
long machine instructions to calculate a square-root, a
CISC processor would have a built-in ability to do this.

5 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
CISC and RISC Processors
 RISC (Reduced Instruction Set Computers):
 RISC is a type of microprocessor architecture that
utilizes a small, highly-optimized set of instructions,
rather than a more specialized set of instructions often
found in other types of architectures.
 History
The first RISC projects came from IBM, Stanford, and
UC-Berkeley in the late 70s and early 80s. The IBM 801,
Stanford MIPS, and Berkeley RISC 1 and 2 were all
designed with a similar philosophy which has become
known as RISC.

6 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
CISC and RISC Processors
The main characteristics of CISC microprocessors are:
 Extensive instructions.
 Complex and efficient machine instructions.
 Extensive addressing capabilities for memory operations.
 Relatively few registers in use.
In comparison, RISC processors are more or less the
opposite of the above:
 Reduced instruction set.
 Less complex, simple instructions.
 Pipelining is used to allow for simultaneous execution of
parts, or stages, of instructions
 Many symmetric registers are used.

7 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Age of Multi-core Processors
 From 1986 – 2002, microprocessors were speeding like a
rocket, increasing in performance an average of 50% per
year.
 Since then, it’s dropped to about 20% increase per year.
 An Intelligent Solution:
 Instead of designing and building faster
microprocessors, put multiple processors on a single
integrated circuit.

8 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Is Multi-core Processors is the Solution?
 Up to now, performance increases have been attributable
to increasing density of transistors in multiple
microprocessors.

 But there are inherent problems:

 Smaller transistors = faster processors
 Faster processors = increased power consumption
 Increased power consumption = increased heat
 Increased heat = unreliable processors

9 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Multi-core Processor and Parallel
Programming
 Adding more processors doesn’t help much if
programmers aren’t aware of them…
 … or don’t know how to use them.

 Serial programs don’t benefit from this approach (in most

cases).
 Move away from single-core systems to multi-core
processors and Introduce parallelism!!!

 “Core” = central processing unit (CPU)

10 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Parallel Programming Concept for
Multi-core Processors
 Compute n values and add them together.
 Serial solution:

 n iteration requires to add.

11 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Parallel Programming Concept for
Multi-core Processors
 8 cores, n = 24, then the calls to
 1,4,3, 9,2,8, 5,1,1, 5,3,7, 2,5,0, 4,1,8, 6,5,1, 2,3,9
 The Master Core runs the serial program

Core 0 1 2 3 4 5 6 7
my_sum 8 19 7 15 7 13 12 14

Core 0 1 2 3 4 5 6 7
my_sum 95 19 7 15 7 13 12 14
12 CSE-4503: Microprocessors and Assembly Language
Islamic University of Technology (IUT)
Parallel Programming Concept for
Multi-core Processors
But wait! There’s a much better way to compute the global sum.

13 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Parallel Programming Performance for
Multi-core Processors
 Analysis
 In the first example, the master core performs 7 receives and 7
additions.
 In the second example, the master core performs 3 receives
and 3 additions.
 The improvement is more than a factor of 2!

 Imagine, If we have 1000 cores:

 The first example would require the master to perform 999
receives and 999 additions.
 The second example would only require 10 receives and 10
additions.
 That’s an improvement of almost a factor of 100!

14 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Processor Cores
 A core can be thought of as in individual processor.
 A dual-core processor, therefore has two internal
processors, a quad-core model has four.
 More cores are useful for multi-tasking; for example, you
can run two applications at the same time, each one
having access to its own dedicated processor.
 More cores are also useful for multi-threaded
applications, such as video editing.
 Core i3 processors have two cores, Core i5 CPUs have
four and Core i7 models also have four. Some Core i7
Extreme processors have six or eight cores.

15 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Thread and Hyper-Thread
 Thread: a thread of execution is the smallest sequence of programmed
instructions that can be managed independently by a scheduler, which is
typically a part of the operating system. Multiple threads can exist within
one process, executing concurrently and sharing resources such as memory,
while different processes do not share these resources.

 Hyper-Thread: Hyper-Threading is Intel's technology for creating two

logical cores in each physical core. In other words, to your operating
system it appears as though your CPU has double the number of cores
than it really does.

 In terms of performance, Hyper-Threading speeds up multi-tasking and

multi-threaded applications. It's not as fast or as efficient as extra 'real'
cores, but it's an improvement over a single Core. Core i3 and i7
processors have this technology, Core i5 processors do not.
16 CSE-4503: Microprocessors and Assembly Language
Islamic University of Technology (IUT)
Turbo-Boost
 Turbo Boost is Intel's technology for automatically
overclocking a processor, boosting its clock speed higher
than the default setting.
 The CPU monitors its temperature and, when it's running
cool enough, will apply the overclock. Core i5 and i7
CPUs have this technology, Core i3 models do not.

17 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Dual Core and Core 2 Duo
 Architecture

 Dual Core: The dual core is the older of the two having an
architecture that sports two cores on one processor, but
its older technology puts it in a position of disadvantage.

 Core 2 Duo: The core 2 duo has two cores on the same
processor. However while the architecture sounds
vaguely the same, it is more advanced than the dual core
processors.

18 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Dual Core and Core 2 Duo
 Performance
 Dual Core: This is without a doubt one of the best Intel
processors, however it lacks the bite in terms of
performance. Acceptable enough clock speeds of about
2.33 GHz for the best ones.

 Core 2 Duo: This newer processor beats the dual core in

all bench-marking tests and therefore performance-wise,
is definitely the better pick. Slams the opposition clocking
speeds of 3.33 GHz clock as seen of the higher end ones.

19 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Core i3
 Entry level processor
 2-4 Cores
 4 Threads
 Hyper-Threading (efficient use of processor resources)
 3-4 MB Cache
 32 nm Silicon (less heat and energy)
 Core i3 processors do support 64-bit versions of
Windows
 Suitability:
 If you use your computer for basic tasks such as word
processing, email, surfing the web, etc., a Core i3 processor is
more than enough to handle all of that with ease

20 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Core i5
 Mid range processor
 2-4 Cores
 4 Threads
 Turbo Mode (turn off core if not used)
 3-8 MB Cache
 32-45 nm Silicon (less heat and energy)
 Suitability:
 Core i5's offer enough performance to do stuff like video
editing and gaming, and more than enough performance to do
basic stuff like-word processing, internet surfing, and email.

21 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Core i7
 High end processor
 4-8 Cores
 8 Threads
 Turbo Mode (turn off core if not used)
 Hyper-Threading (efficient use of processor resources)
 4-8 MB Cache
 32-45 nm Silicon (less heat and energy)

 Suitability:
 Core i7's offer enough performance to do stuff like video
editing and gaming.

22 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Salient Features: Core i9
(10th and upper Generation Processor)
 8-16 cores
 Supports Hyper-Threading
 Upto 16 MB of Smart Cache
 1400 Billions Transistor
 Turbo-mode compatible

 Suitability:
 Core i9's offer enough performance to do stuff like video
editing and gaming.

23 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)
Thank You !!

24 CSE-4503: Microprocessors and Assembly Language

Islamic University of Technology (IUT)