This document discusses computer architecture and trends in computer performance. It covers topics like computer architecture definitions, design goals that influence architecture choices, factors that affect performance like instructions per cycle and clock speed, different types of memory and their speeds, trends in processor architecture over time including increasing transistor counts and multicore processors, and how latency and bandwidth impact network and system performance. It also provides some interesting historical facts about competition between Intel and AMD in the CPU market.

1. PREPARED BY MAHAMMADSALEH ABBAS , AVAZ
QARAYEV, KANAN CHALABI
SUBJECT: COMPUTER ORGANIZATION
AND ARCHITECTURE
SPECIALITY: IT
TRENDS IN COMPUTER
ACRCHITECTURE

2. 2
COMPUTER ARCHITECTURE
In computer engineering, computer architecture is a set of rules and methods that
describe the functionality, organization, and implementation of computer systems. Some
definitions of architecture define it as describing the capabilities and programming
model of a computer but not a particular implementation. In other definitions computer
architecture involves instruction set architecture design, microarchitecture design, logic
design, and implementation.

3. DESIGN GOALS
The exact form of a computer system depends on the constraints and goals.
Computer architectures usually trade off standards, power versus performance,
cost, memory capacity, latency (latency is the amount of time that it takes for
information from one node to travel to the source) and throughput. Sometimes
other considerations, such as features, size, weight, reliability, and expandability are
also factors. The most common scheme does an in-depth power analysis and figures
out how to keep power consumption low while maintaining adequate performance.

4. PERFORMANCE
 Modern computer performance is often described in instructions per cycle (IPC), which
measures the efficiency of the architecture at any clock frequency; a faster IPC rate
means the computer is faster. Older computers had IPC counts as low as 0.1 while
modern processors easily reach near 1. Superscalar processors may reach three to five
IPC by executing several instructions per clock cycle.
 Counting machine-language instructions would be misleading because they can do
varying amounts of work in different ISAs. The "instruction" in the standard
measurements is not a count of the ISA's machine-language instructions, but a unit of
measurement, usually based on the speed of the VAX computer architecture.
 Many people used to measure a computer's speed by the clock rate (usually in MHz or
GHz). This refers to the cycles per second of the main clock of the CPU. However, this
metric is somewhat misleading, as a machine with a higher clock rate may not
necessarily have greater performance. As a result, manufacturers have moved away
from clock speed as a measure of performance.
 Other factors influence speed, such as the mix of functional units, bus speeds, available
memory, and the type and order of instructions in the programs.

5. ► There are two main types of speed: latency and throughput. Latency is the time between the start
of a process and its completion. Throughput is the amount of work done per unit time. Interrupt
latency is the guaranteed maximum response time of the system to an electronic event (like when
the disk drive finishes moving some data).
► Performance is affected by a very wide range of design choices — for example, pipelining a
processor usually makes latency worse, but makes throughput better. Computers that control
machinery usually need low interrupt latencies. These computers operate in a real-
time environment and fail if an operation is not completed in a specified amount of time. For
example, computer-controlled anti-lock brakes must begin braking within a predictable and limited
time period after the brake pedal is sensed or else failure of the brake will occur.
► Benchmarking takes all these factors into account by measuring the time a computer takes to run
through a series of test programs. Although benchmarking shows strengths, it shouldn't be how
you choose a computer. Often the measured machines split on different measures. For example,
one system might handle scientific applications quickly, while another might render video games
more smoothly. Furthermore, designers may target and add special features to their products,
through hardware or software, that permit a specific benchmark to execute quickly but don't offer
similar advantages to general tasks.

6. Past Trends
Processors have undergone a tremendous evolution throughout their history. A key milestone in this
evolution was the introduction of the microprocessor, term that refers to a processor that is implemented
in a single chip. The first microprocessor was introduced by Intel under the name of Intel 4004 in 1971. It
contained about 2,300 transistors, was clocked at 740 KHz and delivered 92,000 instructions per second
while dissipating around 0.5 watts. Since then, practically every year we have witnessed the launch of a
new microprocessor, delivering significant performance improvements over previous ones. Some studies
have estimated this growth to be exponential, in the order of about 50% per year, which results in a
cumulative growth of over three orders of magnitude in a time span of two decades . These improvements
have been fueled by advances in the manufacturing process and innovations in processor architecture.
According to several studies, both aspects contributed in a similar amount to the global gains.

7. Figure 3 shows a high-level block diagram of a typical
contemporary microprocessor. The main components
are a number of general purpose cores, a graphics
processing unit, a shared last level cache, a memory
and I/O interface, and an on-chip fabric to
interconnect all these components. Below, we briefly
describe the architecture of these modules.
7
Current Microprocessors

8. Multicore Processors
The vast majority of current microprocessors have multiple general
purpose based on architecture described in the previous section. Figure
5 depicts the main components of a multicore processor. There are a
number of cores, each one with private L1 caches (separate caches for
instructions and data) and a private second level cache (some
processors do not have a private L2 cache), a shared last level cache
and an interconnection network that allows all the cores to
communicate through the memory hierarchy. The lower levels of the
memory hierarchy are normally a main memory and a disk storage, and
they are located off-chip. A multicore processor can run multiple
threads simultaneously in different cores with no resource contention
among them except for the shared resources, which are basically the
memory hierarchy and the interconnection network. The architecture of
these two components is key for the performance of multicore
processors and are described in more detail below. The architecture of
each one of the individual cores is basically the same as in a single-core
processor, and has been described in the previous section.

9. 9
I'll talk about this on the
next page
In the past, CPU manufacturers thought that increasing MHz
would increase the CPU performance, but later realized that the
higher MHz, higher heat and energy consumption. And that's
why the increase in MHz stagnated after 2010, and then
declined in the following years.
increase
decrease
After the increasing number of cores in the
CPU, single-core performance has decreased
over time. In the past, manufacturers focused
more on single core performance but now its
changed .A rough example for you: imagine
that your team in a fight and your team with
more people beat only one man.
İf The more people in the team, your team will
be stronger , right? So Number of cores looks
like that.
Increasing the number of cores instead
of increasing the MHz has been a good
solution to get a better performance and
therefore the number of cores seems to
increase from year to year.
5 things affect cpu performance

10. TRANSISTOR PERFORMANCE
More transistors can be used to increase the processor throughput. Theoretically, doubling the number
of transistors in a chip provides it with the capability of performing twice the number of functions in the
same time, and increasing its storage by a factor of two. In practice, however, performance gains are
significantly lower. Fred Pollack made the observation long time ago that processor performance was
approximately proportional to the square root of its area, which is normally referred to as Pollack’s rule
of thumb . The increased transistor density allows architects to include more compute and storage units
and/or more complex units in the microprocessors, which used to provide an increase in performance of
about 40% per process generation . The 30% reduction in delay can provide an additional improvement,
as high as 40% but normally is lower due to the impact of wire delays .
Finally, additional performance improvements come from microarchitecture innovation. These innovations include deeper pipelines,
more effective cache memory organizations, new instruction set architecture (ISA) features, larger instruction windows and multicore
architectures just to name some of the most relevant.

11. 11
• Basically has been taken to mean that
the standard computer`s performance
improves , with a doubling time of 18
months
And ı want to say another thing.
Nowaday Moore’s Law
Breaking Down !
The end of Moore’s Law as we know
it was always inevitable. Because
There is a physical limit to what can
fit on a silicon chip after start
working with nanometers.

12. Latency and Bandwidth
The emergence and the fast growth of the web performance optimization industry within the past few years is a sign of
the growing importance and demand for faster user experiences by the consumer. And this is not simply a
psychological need for speed in our ever devoloping and connected world: EX
Faster sites lead to better user engagement.
Faster sites lead to better user retention.
Faster sites lead to higher conversions.
Simply put, speed is a feature. And to deliver it, we need to understand the many factors and fundamental limitations.
In this part, we will focus on the two critical components that dictate the performance of all network traffic: latency
and bandwidth
12
Latency
The time from the source sending a packet to the destination receiving it.
For example: 5G is becoming widespread nowadays and as far as we
know 5G has a very low latency (around 1ms) and it has many
benefits for us. Now doctors will be able to perform operations
remotely, cars can be managed via the internet and ( it can be
decrease in the number of accidents ) etc.
Bandwidth
Maximum throughput of a logical or physical communication path.
the maximum rate of data transfer across a given path

13. NETWORK
Let’s take a closer look at some common contributing components for a typical router on the Internet, which is responsible for
relaying a message between the client and the server:
Propagation delay
Amount of time required for a message to travel from the sender to receiver, which is a function of distance over speed with
which the signal propagates.
Transmission delay
Amount of time required to push all the packet’s bits into the link, which is a function of the packet’s length and data rate of the
link.
For example, to put a 10 megabyte (MB) file "on the wire" over a 1Mbps link, we will need 80 seconds. 10MB is equal to 80Mbps
because there are 8 bits for every byte! ex: You can download 1 Megabit of data per second with a speed of 1 Mbps. This means
only 128 Kilobytes (KB).
Processing delay
Amount of time required to process the packet header, check for bit-level errors, and determine the packet’s destination.
Queuing delay
Amount of time the packet is waiting in the queue until it can be processed.
The total latency between the client and the server is the sum of all the delays just listed.
In some countries today, internet speed has reached 1000 megabitpersecond.
In fact, Nasa's internet speed is 91 gigabits per second.
8 Gigabit = 1 Gigabyte
so
91 Gigabit = 11.375 Gigabytes .
So you can download gta5 in just 5 seconds

14. SSD vs. HDD Speed and Performance (SECONDARY MEMORY)
► Solid state drives (SSDs) are faster than conventional hard disk drives (HDDs) and they are also more
reliable and use less power. That means that when it comes to choosing between SSD or HDD storage,
SSDs would be preferably to HDDs in all cases if it weren't for one fact: SSDs are more expensive than
HDDs when measured by cost per Gigabyte of storage.
► To understand why there is a big difference between SSD v HDD speed, it's necessary to consider
the difference between SSD and HDD technology.
SSD have dramatically faster read and write
speeds when compared with hard disk
drives.
nowadays ssd's more popular than hdd
(TREND)

15. 15
But not in Azerbaijan (
you can see how bandwidth and latency
increase year by year

16. INTERESTING FACTS
I prepared this interesting subject with quotations from Mesut
Çevik's Youtube channel, I hope you like it.
If AMD did not exist, we would probably see only Intel in the
processor world and only Nvidia as the GPU manufacturer in the
graphics card world. And in a world where there is no competition,
users always pay more and buy less.
IBM chooses Intel's x86 architecture as its processor technology
while setting PC standards. As the operating system, it chooses
Microsoft's DOS operating system, which we all know very well
today.
Of course, DOS is far from the Windows we currently use, but this is
a milestone for both companies. But IBM makes Intel a condition to
avoid any trouble with processor supply, and they say that you will
share the x86 license with a second manufacturer. they choose it as
a manufacturer and license x86 to AMD and start producing the first
processors for IBM.
16

17. 17
Everything is going very well in the first years. İntel begins to do the first annoyances in 1982. It
refuses to fully share the 80286 architecture with AMD, and the source files and updates it
shares are of a nature that will not work for AMD. That's why AMD lags behind Intel in this
processor production for a long time. Now an AMD is always a must for Intel, but a weak AMD is
a must. He tries his best to make the company appear as a second backup supplier, each time
with different ways and different strategies. But the process is not limited to this. Because
Intel is constantly making agreements with companies in a monopoly movement against AMD.
But not looking at this, we see the world's first desktop processor that exceeds the 1 GHz clock
frequency from AMD. These challenges did not stop AMD, and the company continued to
improve its technology, introducing end users to dual-core desktop processors in 2005. But no
matter what AMD did, whatever technological advances it showed, how it got ahead of its
competitors, major manufacturers were not willing to sell AMD products. AMD also thought that
there might be different jobs behind it, and because it can support it with some evidence, it
filed a sizable monopoly suit against Intel in 2005.

18. 18
The first claim in AMD's indictment is that Intel is threatening big
manufacturers not to buy discounts. Another claim is that Intel
pays Sony millions of dollars just to do business with it. As a
result, AMD's market share in Sony products has dropped from 23
percent in 2002 to 8 percent in 2003 and zero percent today. So it
talks about 2005. You can no longer see any AMD processors in
Sony products and for these reasons, AMD is winning the case.
Things like this are progressing until today and AMD is now one of
the best graphics card and CPU manufacturers. If you like it, if
you want to know more about this subject, you can use the
YouTube channel of "Mesut Çevik".