Intel Core i7 Processor

Ekansh Priyam Agarwal

Bachelor of Technology in Computer Science and Bachelor of Technology in Computer Science and
Engineering Engineering

Shiv Nadar University Shiv Nadar University

Greater Noida, India Greater Noida, India

[[email protected]] [[email protected]]

Abstract— [1][4]Intel’s Core i7 processors are based I. INTRODUCTION

on the Penryn manufacturing process the company
introduced last year, but otherwise there are few Intel core i7 processor is one of the first core i-
similarities with the Core 2 Duo, Core 2 Quad, and series processors built based on the new Intel’s
Core 2 Extreme lines. A Core i7 processor fits in microarchitecture technology called Nehalem. Intel’s
familiar LGA775 motherboard chip sockets, the Intel Nehalem microarchitecture has key features that outrank
norm for years now. Core i7 processors employ a new from previous processor technologies.
technology, the Intel Quick Path Interconnect (QPI),
for increased bandwidth and reduced latency. Hyper- In the Nehalem architecture, the memory controllers and
Threading has also been reintroduced to the line, so North Bridge are directly put into the processor. In the
each core process two threads simultaneously making Nehalem processor technology, core i7 is the first CPU
eight-core processing a reality. And in case that not introduced in 2008.
enough multiprocessing for the technology can
support as eight physical cores meaning that 16-core Intel core i3, i5, and i7 naming scheme for their CPUs
processing The road to the Core i7 actually started for quite a while now, but what these labels mean tends
with the demise of Intel’s Netburst architecture. to slowly change over time as new features are
Intel's old strategy for producing microprocessors introduced or older ones get replaced. On top of this, the
was to simply increase the core clock speeds, naming scheme between desktop and mobile CPUs is
instruction sets and cache sizes a few ticks every year. often different as well. The biggest thing one needs to
Every time this happened the power draw and heat know in regards to the i3, i5 and i7 naming scheme is
levels would increase as well, until eventually Intel hit that it is primarily a way for Intel to separate their CPUs
a brick wall with the Pentium 4. They are 64 bit into three performance tiers:
processors in computer architecture, 64-bit integers,
memory addresses or other data units are those that a. Intel Core i7: High-end
are at most 64 bits wide. Also 64-bit CPU and ALU b. Intel Core i5: Mainstream
architectures are those that are based on registers,
c. Intel Core i3: Entry-level
address buses or data buses of that size. The need for
core i7 processors requires a comparison with their
Until recent times, core i7 family processors were the
immediate predecessors. The comparison can be
best particularly for extreme 3D gaming, intensive
summarized as follows. The Core i7 is a completely
graphics tasks, multimedia production in standard
new architecture which is much faster and more
computer level. Now, this processor becomes a previous
efficient than the Core 2 Duo.
generation CPU, for the reason that Intel introduced new
and improved processors called 2nd generation Intel®
Keywords— Intel core i7, Penryn manufacturing,
CoreTM processors family. But this doesn’t mean that
Motherboard chip, Quick Path Interconnect (QPI),
the Core i7 processor is out of order. It still does the job
Hyper-Threading, Intel’s Netburst architecture, ALU
it is designed for.
architecture, Core 2 Duo.
Core i7 processor has several versions both in the high III. PURPOSE OF INTEL CORE I7 PROCESSOR
end and budget group. Depending on ones work type
and budget, one can pick the right one. The Core i7 processors series targets the gaming
industry and for the applications that demand efficient
performance.

Generally, Core i7 processor is recommended for:-

• Multitasking, for running multiple programs at the

same time

• Multithreading applications

• Extreme 3D gaming

• Creating professional movies and editing graphical

tasks

• More than enough for basic tasks such as word

processing, internet browsing and e-mail

Fig1 - Intel Core i7 processor

IV. Tabled details of INTEL Core i7 Processors[4]

[15]
II. CORE I7 GENERAL SPECIFICATIONS
Processor Intel Core i7
• All support 64-bit execution Series Nomenclature i7-6xx, i7-7xx,
i7-8xx, i7-9xx,
• Integrate 2-8 Cores (latest Core i7 processor i7-2xxx, i7-37xx,
incorporates 8 cores) i7-38xx, i7-47xx,
i7-48xx, i7-58xx,
• Support upto 16 threads
i7-59xx, i7-67xx,
i7-68xx, i7-69xx,
• Speed ranges from 2.66GHz to 5.10GHz
i7-7700K
• Support DDR3 and DDR4 main memory Code Name Bloomfield,
Nehalem,
• Support Hyper-threading technology
Clarksfield,
Clarksfield XM,
• Support up to 16MB cache
Lynnfield, Sandy
Bridge, Sandy
• Enhanced Intel SpeedStep Technology
Bridge-E, Ivy Bridge,
• Intel Virtualization Technology Ivy Bridge-E,
Haswell, Haswell
• Intel Turbo Boost Max Technology Refresh, Devil‘s
Canyon, Broadwell,
• Streaming SIMD Instructions (MMX) Skylake, Kaby Lake,
Coffee Lake
• Overclocking capability

• Intel thermal velocity boost Production date 2008-Present

Clock Rate 1.6Ghz-4.4GHz

• Intel speed shift technology

• Thermal Monitoring Technologies

 Socket LGA 1156, Bus Speed 2.5GT/s - 6.4 GT/s
LGA 1155, LGA
1366, LGA 2011, L1 Cache 64 KiB per core
LGA 1150, LGA
L2 Cache 6x256 KiB
1151
L3 Cache 12 MiB - 20 MiB
Fabrication 14 nm, 22 nm, 32 Overclock capable Yes
nm, 45 nm

TDP 45W- 130W

Number of cores 4,8

Bus Speed 4.8 GT/s, 6.4 GT/s

L1 Cache 64 KiB per core

L2 Cache 4×256 KiB

Fig2 - Intel Core i7 internal die photograph
L3 Cache 6 MiB - 10 MiB

V. Features of intel core i7 Processors

Processor Intel Core i7 (Extreme Edition
1. Higher-Performance Multiprocessor Systems with
QPI [1]
Series Nomenclature i7-970, i7-980, i7-980x,
i7-990x, i7-39xx, i7-49xx,
Not only was the memory controller moved to the
i7-5820K, i7-59xx, i7-6800K,
CPU for Nehalem processors, Intel also introduced a
i7-6850K, i7-6900K,
distributed shared memory architecture using Intel
i7-6950X, (i5-7640X),
Quick Path Interconnect (QPI). QPI is the new point-to-
i7-7740X, i7-7820X
point interconnect for connecting a CPU to either a
chipset or another CPU. It provides up to 25.6 GB/s of
total bidirectional data throughput per link. Intel’s
Code Name Gulftown, Sandy Bridge-E, decision to move the memory controller in the CPU and
Ivy Bridge-E, Haswell-E, introduce the new QPI data bus had an impact for single-
Broadwell-E, Skylake, Kaby processor systems. However, this impact is much more
Lake significant for multiprocessor systems.

Production date 2011- Present

Clock Rate 3.0 GHz - 5.0 GHz

Socket LGA 1366,

LGA 2011, LGA 2011-v3,
LGA 2066

Fabrication 14 nm, 22 nm, 32 nm

TDP 130 W - 150 W

Fig 3 -Typical system based on the Intel® CoreTM i7
Number of cores 4, 6, 8 or 10 processor
(with hyperthreading)
The higher-level architectural differences between the
previous generation and the new Nehalem
microarchitecture for multiprocessor systems:-

Fig 6- Intel Core i7 internal core diagram

Fig 4- Previous Generation Microarchitecture

2. Improved Cache Latency with Smart L3 Cache[1]

Cache is a block of high-speed memory for

temporary data storage located on the same silicon die as
the CPU. If a single processing core, in a multicore
CPU, requires specific data while executing an
instruction set, it first searches for the data in its local
caches (L1 and L2). If the data is not available, also
known as a cache-miss, it then accesses the larger L3
Fig 5 - Nehalem Microarchitecture cache. Exclusive L3 cache, if that attempt is
unsuccessful, then the core performs cache snooping
The Nehalem microarchitecture integrated the memory searches the local caches of other cores – to check
controller on the same die as the Core i7 processor and whether they have data that it needs. Attempt also results
introduced the high-speed QPI data bus. In a Nehalem- in a cache-miss it then accesses the slower system RAM
based multiprocessor system each CPU has access to for that information. The latency of reading and writing
local memory but they also can access memory that is from the cache is much lower than that from the system
local to other CPUs via QPI transactions. For example, RAM, therefore a smarter and larger cache greatly helps
one Core i7 processor can access the memory region in improving processor performance.
local to another processor through QPI either with one
The Core i7 family of processors features an
direct hop or through multiple hops. With these new
inclusive shared L3 cache that can be up to 12 MB in
features, the Core i7 processors lend themselves well to
size. The NI PXIe-8133 embedded controller features
the creation of higher-performance processing systems.
four cores, where each core has 32 kilobytes for
For maximum performance gains in a multiprocessor
instructions and 32 kilobytes for data of L1 cache, 256
system, application software should be multithreaded
kilobytes per core of L2 cache, along with 8 megabytes
and aware of this new architecture. Also, execution
of shared L3 cache. The L3 cache is shared across all
threads should explicitly attempt to allocate memory for
cores and its inclusive nature helps increase performance
their operation within the memory space local to the
and reduces latency by reducing cache snooping traffic
CPU on which they are executing. By combining a
to the processor cores. An inclusive shared L3 cache
multiprocessor computer with PXI-MXI-Express to a
guarantees that if there is a cache-miss, then the data is
PXI system, processor intensive applications can take
outside the processor and not available in the local
advantage of the multiple CPUs. Examples of these
caches of other cores, which eliminates unnecessary
types of applications range from design simulation to
cache snooping. The L3 is designed to use the inclusive
hardware-in-the-loop (HIL).
nature to minimize snoop traffic between processor
cores. The latency of L3 access may vary as a function 2) To find how system load, specifically the number of
of the frequency ratio between the processor and the threads running concurrently, affects when and how
uncore sub-system. often Turbo Boost gets engaged, and finally,
This feature provides improvement for the overall
performance of the processor and is beneficial for a 3) To determine how scheduling decisions that distribute
variety of applications including test, measurement and load in a processor affect the potential performance
control. Each physical processor may contain several improvements offered by Turbo Boost.
processor cores and a shared collection of subsystems
that are referred to as "uncore". Specifically in Intel Core To this end, the SPEC CPU2006 benchmark suite
i7 processor the uncore provides a unified third-level [13]benchmark applications with diverse qualities (inte-
cache shared by all cores in the physical processor,Intel ger versus floating point applications, memory-intensive
QuickPath Interconnect links and associated logic. versus computationally-intensive applications)was
selected. Benchmarks were run individually and in
3. Turbo boost feature in Intel Core i7 Processor[14] groups while monitoring system performance with and
without the Turbo Boost feature. The results of the study
The latest multi-core processor from Intel found were that Turbo Boost can provide on average up
codename Nehalem [9] has a unique feature called to a 6% reduction in execution time. Turbo Boost
Turbo Boost Technology [10]. With Turbo Boost, the Technology had the most impact on performance when
processor opportunistically increases the frequency of the scheduling was not optimal; however, in all cases,
the cores based on the core temperature, the number of Turbo Boost enhanced performance. Turbo Boost also
active cores, the estimated current consumption, and the resulted in a significant increase in energy consumption
estimated power consumption. Normally, the Core i7 because the processor requires a higher voltage to
processor can operate at frequencies between 1.5 GHz operate at Turbo Boost frequencies. However, current
and 3.2 GHz (the maximum non-Turbo Boost frequency processors also support low power sleep states where
or the base frequency) in frequency steps of 133.33 they consume very little power[12]. Disks, memory and
MHz. With Turbo Boost enabled, the processor can other platform components can also be big contributors
increase the frequency of cores two further levels to 3.3 to platform power consumption.This research is quite
GHz and then 3.4 GHz. We refer to the first frequency useful to both CPU designers as it demonstrates the
above the base frequency as the lower Turbo Boost benefits and costs of Turbo Boost technology, and to
frequency (3.3 GHz) and to the maximum frequency as software designers as it provides insight into the
the higher Turbo Boost frequency (3.4 GHz). If multiple benefits of this technology for applications.
physical cores are active, only the lower Turbo Boost
frequency is available. Prior work has shown that such a processor
configuration offers higher performance per watt in most
Turbo Boost is made possible by a processor situations when compared with symmetric multi-core
feature named power gating. Traditionally, an idle processors [6], and a great deal of other work has
processor core consumes zero active power while still analyzed the performance, versatility, and energy-
dissipating static power due to leakage current. Power efficiency of AMP systems either theoretically or
gating aims to cut the leakage current as well, thereby through simulation [2], [8], [6], [7], [11].
further reducing the power consumption of the idle core.
The extra power headroom available can be diverted to Prior work from Intel [2] has shown that such a
the active cores to increase their voltage and frequency processor can be leveraged to mitigate Amdahl’s law for
without violating the power, voltage, and thermal parallel applications with sequential phases. Amdahl’s
envelope. law states that the speedup of a parallel application is
limited by its sequential component. A typical parallel
Turbo Boost Technology essentially makes the Nehalem application might divide a computational task into many
a dynamically asymmetric multi-core processor (AMP); threads of execution executing in parallel, and then
cores use the same instruction set but their frequency can aggregate the results using only a single thread. This
vary independently and dynamically at runtime. division of work results in an execution pat- tern where
parallel phases of execution are interspersed with
A detailed evaluation of the Turbo Boost feature was sequential “bottleneck” phases. A dynamically
performed with the following goals: asymmetric processor can accelerate such bottleneck
phases while staying within its energy budget.
1) To understand how Turbo Boost behaves depending
on the properties of the application such as its degree of When a program enters a sequential phase, the processor
CPU or memory intensity, would automatically turn off idle cores and boost the
frequency on the active core. When the program returns Hyperthreading: Hyperthreading is a technology that
to the parallel phase, all the cores would be activated, divides a physical core into two threats or logical
but the frequency of each core would be reduced. The cores.These cores are important for the power of a
benefits of such an architecture are demonstrated by device but virtual cores are never powerful as physical
Annavaram et al. [2]. They observe performance cores.The i5 processor lacks this technology and i7 has
improvements of as much as 50% relative to symmetric this.
systems using a comparable energy budget. Nehalem,
with its Turbo Boost feature has the potential to mitigate Turbo Boost : This is the technology that allows the
Amdahl’s law for parallel applications with sequential processor to raise its frequency for a certain amount of
phases, therefore we evaluated this capability using time. Both the i5 and i7 processors have this technology,
several parallel applications from the PARSEC [5] and but the difference is that i5 can raise its frequency to 3.5
BLAST [3] benchmark suites. GHz while the i7 can raise its frequency to 4.1GHz.

VI. ADVANTAGES OF INTEL CORE i7 B. Intel Core i7 vs i9

PROCESSORS
Core Concept: The Core i7 processors feature six to
1. It offers a big cache size which is about 8 MB. eight cores, whereas the i9 CPUs all ship with a
This helps to increase instruction execution minimum of eight.
speed.
Cache: A larger cache size on a multi-threaded CPU
2. Processor speed is very fast due to a helps deal with multiple tasks more efficiently. The i7
combination of turbo boost and hyperthreading processors feature an L3 cache size of 12MB, whereas
technologies. the i9’s come commonly equipped with 16MB. The X-
series processors can be even larger, with the i9-9980XE
3. It consists of a 4 pin connector to control the featuring an L3 smart cache size of 24.75MB.
speed of the fan. It provides cooler running
technology for less heat and less noise. Hyperthreading: The feature that creates two processing
threads for every physical core. For 9th Gen Core
4. It can be run at overclocking speed. It is the processors, Intel has restricted Hyperthreading to i9
process by which computers run at a higher only. For previous generation Core processors, i7 was
clock rate than it is designed for. the class with Hyperthreading.

VII. DISADVANTAGES OF INTEL CORE i7

PROCESSORS
IX. CONCLUSION
1. It is expensive compared to core-i3 and core-i5
The Intel Core i7 processor is Intel’s first CPU designed
processors.
based on Nehalem micro architecture.[9] This processor
is ideal for computer 3D games, multitasking and multi-
2. Power consumption of core-i7 is not better threading applications. The Core i7 family of processors
compared to core-2 duo processor type. based on the Intel Nehalem microarchitecture offers
many new and improved features that benefit a wide
3. It requires newer motherboards. variety of applications including test, measurement and
control. Engineers and scientists can expect to see
4. It is sensitive to higher voltages. processing performance gains as well as increases in
memory and data throughput when comparing this micro
architecture to previous micro architectures. Now with
VIII. DIFFERENCES BETWEEN PROCESSORS the introduction of the 10th generation of intel core i7, i5
and i3 processors, Intel has placed itself apart from its
A. Intel Core i5 vs i7 competitor like AMD in terms of performance. The main
thing to remember is that i7 and i9 CPUs are at the high
Cache: The more the cache in a processor the faster it end of the product line for heavy computing stuff, i5
functions as it reduces the interaction of the CPU with CPUs are in the middle for an intermediate user, and i3
RAM and the SSD or HDD, The i5 processor has 4Mb- CPUs are entry level for a beginner or one with not so
6Mb Cache and the i7 processor as a rule has 8Mb of much intensive work to perform on his computer.
cache.
 X. ACKNOWLEDGEMENT MICRO-36. Proceedings. 36th Annual IEEE/ACM International
Symposium on, pages 81–92, 2003.

We would like to thank our professor Dr. Divya Lohani [13] Aashish Phansalkar, Ajay Joshi, and Lizy K. John. Analysis of
for giving us this opportunity to work on this topic. We redun- dancy and application balance in the SPEC CPU2006
would also like to thank our teaching assistant, Miss benchmark suite. In ISCA ’07: Proceedings of the 34th annual
International Symposium on Computer Architecture, pages 412–423,
Krati for helping us throughout this assignment. New York, NY, USA, 2007. ACM.

Nevertheless, we would like to thank all those whose [14] J. Charles, P. Jassi, N. S. Ananth, A. Sadat and A. Fedorova,
"Evaluation of the Intel® Core™ i7 Turbo Boost feature," 2009 IEEE
research papers have helped us in this assignment. We
International Symposium on Workload Characterization (IISWC),
tried our best to mention each of them in our References Austin, TX, 2009, pp. 188-197.
section.
[15] J. M. Cebrián, L. Natvig and J. C. Meyer, "Improving Energy
Efficiency through Parallelization and Vectorization on Intel Core i5
REFERENCES and i7 Processors," 2012 SC Companion: High Performance
Computing, Networking Storage and Analysis, Salt Lake City, UT,
[1] Series Datasheet , “Intel® CoreTM i7-900 Desktop Processor 2012, pp. 675-684.
Extreme Edition Series and Intel® CoreTM i7-900 Desktop
Processor”, Volume 2

[2] Murali Annavaram, Ed Grochowski, and John Shen. Mitigating

Am- dahl’s Law through EPI Throttling. In ISCA ’05: Proceedings of
the 32nd annual international symposium on Computer Architecture,
pages 298–309, Washington, DC, USA, 2005. IEEE Computer Society.

[3]S.F.Altschul,W.Gish,W.Miller,E.W.Myers,andD.J.Lipman.Basic
Local Alignment Search Tool. J Mol Biol, 215(3):403–410, October
1990.

[4] White Paper Introduction to Intel® Architecture, the basics

[5] Christian Bienia, Sanjeev Kumar, Jaswinder Pal Singh, and Kai Li.
The PARSEC Benchmark Suite: Characterization and Architectural
Implications. In Proceedings of the 17th International Conference on
Parallel Architectures and Compilation Techniques, October 2008

[6] A. M. El-Husseini and M. Morrise, "Clocking design automation in

Intel's Core i7 and future designs," 2011 IEEE/ACM International
Conference on Computer-Aided Design (ICCAD), San Jose, CA, 2011,
pp. 276-278.

[7] G. Shamanna, N. Kurd, J. Douglas and M. Morrise, "Scalable,

sub-1W, sub-10ps clock skew, global clock distribution architecture for
Intel® Core™ i7/i5/i3 microprocessors," 2010 Symposium on VLSI
Circuits, Honolulu, HI, 2010, pp. 83-84.

[8] Mark D. Hill and Michael R. Marty. Amdahl’s Law in the

MulticoreEra. Computer, 41(7):33–38, 2008.

[9] Intel® Corporation. First the tick, now the tock: Next
generationIntel® microarchitecture (Nehalem). Whitepaper, Intel
Corporation,
April 2008.

[10] Intel® Corporation. Intel® Turbo Boost Technology in Intel®

CoreTM Microarchitecture (Nehalem) Based Processors. Whitepaper,
Intel® Corporation, November 2008.

[11] M. Aater Suleman, Onur Mutlu, Moinuddin K. Qureshi, and Yale

N.Patt. Accelerating critical section execution with asymmetric multi-
core architectures. In ASPLOS ’09: Proceeding of the 14th
International Conference on Architectural Support for Programming
Languages and Operating Systems, pages 253–264, New York, NY,
USA, 2009. ACM

[12] R. Kumar, KI Farkas, NP Jouppi, P. Ranganathan, and DM

Tullsen. Single-ISA heterogeneous multi-core architectures: The
potential for processor power reduction. In Microarchitecture, 2003.