Pentium M

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Pentium M
General information
LaunchedMarch 12, 2003
DiscontinuedMay 2, 2008[1][better source needed]
Common manufacturer(s)
  • Intel
Max. CPU clock rate900 MHz to 2.26 GHz
FSB speeds400 MT/s to 533 MT/s
Architecture and classification
Min. feature size0.13µm to 90nm
MicroarchitectureP6 variant
Instruction setx86, MMX, SSE, SSE2
Physical specifications
  • 1
Products, models, variants
Core name(s)
  • Banias
  • Dothan
PredecessorIntel Pentium III
SuccessorIntel Core (Duo),

The Pentium M is a family of mobile 32-bit single-core x86 microprocessors (with the modified Intel P6 microarchitecture) introduced in March 2003 and forming a part of the Intel Carmel notebook platform under the then new Centrino brand.[2] The Pentium M processors had a maximum thermal design power (TDP) of 5–27 W depending on the model, and were intended for use in laptops (thus the "M" suffix standing for mobile). They evolved from the core of the last Pentium III–branded CPU by adding the front-side bus (FSB) interface of Pentium 4, an improved instruction decoding and issuing front end, improved branch prediction, SSE2 support, and a much larger cache. The first Pentium M–branded CPU, code-named Banias, was followed by Dothan.[3] The Pentium M-branded processors were succeeded by the Core-branded dual-core mobile Yonah CPU with a modified microarchitecture.


The Pentium M represented a new and radical departure for Intel, as it was not a low-power version of the desktop-oriented Pentium 4, but instead a heavily modified version of the Pentium III Tualatin design (itself based on the Pentium II core design, which in turn had been a heavily improved evolution of the Pentium Pro). It is optimized for power efficiency, a vital characteristic for extending notebook computer battery life. Running with very low average power consumption and much lower heat output than desktop processors, the Pentium M runs at a lower clock speed than the laptop version of the Pentium 4 (The Pentium 4-Mobile, or P4-M), but with similar performance - a 1.6 GHz Pentium M can typically attain or even surpass the performance of a 2.4 GHz Pentium 4-M.[4] The Pentium M 740 has been tested to perform up to approximately 7,400 MIPS and 3.9 GFLOPS (using SSE2).[5]

The Pentium M coupled the execution core of the Pentium III with a Pentium 4 compatible bus interface, an improved instruction decoding/issuing front end, improved branch prediction, SSE2 support, and a much larger cache. The usually power-hungry secondary cache uses an access method which only switches on the portion being accessed. The main intention behind the large cache was to keep a decent-sized portion of it still available to the processor even when most of the L2 cache was switched off, but its size led to a welcome improvement in performance.

Other power saving methods include dynamically variable clock frequency and core voltage, allowing the Pentium M to throttle clock speed when the system is idle in order to conserve energy, using the SpeedStep 3 technology (which has more sleep stages than previous versions of SpeedStep). With this technology, a 1.6 GHz Pentium M can effectively throttle to clock speeds of 600 MHz, 800 MHz, 1000 MHz, 1200 MHz, 1400 MHz and 1600 MHz; these intermediate clock states allow the CPU to better throttle clock speed to suit conditions. The power requirements of the Pentium M varies from 5 watts when idle to 27 watts at full load. This is useful to notebook manufacturers as it allows them to include the Pentium M into smaller notebooks.

Although Intel marketed the Pentium M exclusively as a mobile product, motherboard manufacturers such as AOpen, DFI and MSI shipped Pentium M compatible boards designed non-mobile enthusiasts, HTPC, workstation and server applications. An adapter, the CT-479, was developed by ASUS to allow the use of Pentium M processors in selected ASUS motherboards designed for Socket 478 Pentium 4 processors. Shuttle Inc. offered packaged Pentium M desktops, marketed for low energy consumption and minimal cooling system noise. Pentium M processors are also of interest to embedded systems' manufacturers because the low power consumption of the Pentium M allows the design of fanless and miniaturized embedded PCs. The Pentium M also responds very well to undervolting, which can be done with the program Notebook Hardware Control or RMClock.

Intel Pentium M processor family
2003-2005 Logo 2006-2008 Logo Laptop
Code-name Process Date released
Original Pentium M brand logo New Pentium M brand logo Banias
(130 nm)
(90 nm)
Mar 2003
Jun 2004
List of Intel Pentium M microprocessors


Pentium M 1.4 with Banias core
Backside of a Pentium M 1.4

As the M line was originally designed in Israel idc/israel-development-center under manager shmuel-mooly-eden[6] the first Pentium M was identified by the codename Banias, named after an ancient site in the Golan Heights. The Intel Haifa team had previously been working on the memory controller for Timna, which was based on earlier P6 memory controller designs giving them detailed knowledge of P6 architecture which they used when Intel gave them a crash project to create a backup mobile CPU. [7] Given the product code 80535, it initially had no model number suffix, but was later identified as the Pentium M 705. It was manufactured on a 130 nm process, was released at frequencies from 900 MHz to 1.7 GHz using a 400 MT/s FSB, and had 1 megabyte (MB) of Level 2 cache. The core average TDP (Thermal Design Power) is 24.5 watts.

The Banias family processors internally support Physical Address Extension (PAE) but do not show the PAE support flag in their CPUID information; this causes some operating systems (primarily Linux distributions) to refuse to boot on such processors since PAE support is required in their kernels.[8] Using the 'forcepae' linux boot option will allow linux to boot using PAE in these cases.


Pentium M 730 core Dothan
Pentium M 730 core Dothan backside

On September 17, 2003, Intel unveiled plans for releasing its then next-generation of Pentium M processors, codenamed "Dothan" by them. It was named after another ancient town in Israel, and it launched formally on May 10, 2004. Dothan Pentium M processors (product code 80536, CPUID 0x6DX) are among the first Intel processors to be identified using a "processor number" rather than a clockspeed rating; the initial Dothan versions with the 400Mhz Front-Side-Bus (FSB) are known as Pentium M 710 (1.4 GHz), 715 (1.5 GHz), 725 (1.6 GHz), 735 (1.7 GHz), 745 (1.8 GHz), 755 (2.0 GHz), and 765 (2.1 GHz).[9][10] These initial Dothan models all have a TDP of 21 W and a 2 MB L2 cache.

These 700 series Dothan Pentium M processors retain the same basic design as the original Banias Pentium M, but are manufactured on a 90 nm process, with twice the secondary cache. Die size, at 87 mm2, remains in the same neighborhood as the original Pentium M, even though the 1000 series contains approximately 140 million transistors, most of which make up the 2 MB cache. TDP is also down to 21 watts with the 400 MT/s FSB (from 24.5 watts in Banias), though power use at lower clockspeeds has increased highly . However, tests conducted by third party hardware review sites show that Banias and Dothan equipped notebooks have roughly equivalent battery life.[citation needed] Additionally third party hardware review sites have benchmarked the Dothan at approx 10-20% better performance than the Banias in most situations.

Revisions of the Dothan core were released in the first quarter of 2005 with the Sonoma chipsets and supported a 533 MT/s FSB and XD (Intel's name for the NX bit); and the PAE support flag in the CPUID was enabled, unlike earlier Pentium Ms that showed PAE unavailable. These revised Dothan processors include the 730 (1.6 GHz), 740 (1.73 GHz), 750 (1.86 GHz), 760 (2.0 GHz), 770 (2.13 GHz) and 780 (2.26 GHz) and have a TDP of 27 W and a 2 MB L2 cache.

In July 2005, Intel released the 780 (2.26 GHz) and the low-voltage 778 (1.60 GHz).

The processor line had models running at clock speeds from 1.0 GHz to 2.26 GHz as of July 2005. The models with lower frequencies were either low voltage or ultra-low voltage CPUs designed for improved battery life and reduced heat output. The 718 (1.3 GHz), 738 (1.4 GHz), and 758 (1.5 GHz) models are low-voltage (1.116 V) with a TDP of 10 W, while the 723 (1.0 GHz), 733 (1.1 GHz), and 753 (1.2 GHz) models are ultra-low voltage (0.940 V) with a TDP of 5 W.

Intel A100 Series[edit]

An ultra low-power microprocessor based on the Dothan built on a 90 nm process with 512 KB L2 cache and 400 MT/s front side bus (FSB).

Core Solo and Core Duo[edit]

The next generation of processors, codenamed Yonah, were based on the Enhanced Pentium M architecture, and released under the Intel Core brand, as Core Duo and Core Solo.

See also[edit]


  1. ^ "Product Change Notification #106928–02". Intel. December 27, 2006. Retrieved October 14, 2019.
  2. ^ Stokes, Jon (25 February 2004). "A Look at Centrino's Core: The Pentium M". Ars Technica. Retrieved 11 November 2019.
  3. ^ Shimpi, Anand Lal. "Intel's 90nm Pentium M 755: Dothan Investigated". Retrieved 2019-12-05.
  4. ^ Intel's Centrino Duo Notebook Technology
  5. ^ "Intel Pentium M 740 PCSTATS Review - Benchmarks: Office Productivity, SiSoft Sandra 2005". Archived from the original on 2013-10-29. Retrieved 2007-09-14.
  6. ^ ",1864.html from google ( tomshardware pentium m green card israel ) result 1 :
    shmuel-mooly-eden: A look into the origins of Core 2 Duo.
    writer : The Editors of Tom's Hardware.
    publish-date : 20-sep-2006.
    (usa-country->california-state->santa-clara-city) - If you were to look into story of Intel's latest micro-processors core-2-duo, you'd quickly learn that core-2-duo aren't based on most likely predecessor, Pentium 4. Track-down family-line and you end up at Banias, the 1st Pentium M. Join us for an interview with shmuel-mooly-eden, man in charge of a project that convinced Intel to put giga-hertz campaign into reverse. [source:'As head of Intel's Israeli development center, Shmuel "Mooly" Eden, a brash and garrulous engineer, led the billion-dollar Pentium M project and later designed the Centrino chip.' in from google ( mooly eden ) result 3]. 1 of great advantages of being a technology-journalist is the fact that we can receive deep insight/idea into technology that is penetrating our every-day life. An essential part of the job is doing direct-contact with source and developers of new technologies and, while we ( technology-journalist ) spend much time on phone-call, we also receive opportunities to physically-meet developers and inventors of new products. From time to time, when we meet unique personalities, we are pursuing interview appointments to know more about some project and person behind that project. shmuel-mooly-eden, vice president and general manager of Intel's Mobile Platforms Group, has unique personality. and with that kind of back-ground, a thorough/complete interview on TG Daily is overdue. shmuel-mooly-eden joined intel year 1982 becoming member of start-up staff in israel-country->jerusalem-city->Fab-8 and eventually led "Banias" project, which resulted in launch of the 1st Pentium M in mar-2003. [foto:shmuel-mooly-eden]. Pentium M project is credited with prompting Intel to rethink its strategy which rapidly-increase clock-speed and rapidly-increase electricity-consumption of cpu. Pentium M, designed in Israel-country, carries many of genes of Core-2-Duo, which enabled Intel to regain leadership position in CPU market. We met shmuel-mooly-eden in Intel's Santa Clara HQ to learn more about Banias-history, and learn about a unique personality as well.
    ---TG Daily: A few weeks ago, Intel introduced more electricity-efficient Core micro-architecture, which has been considered as 1 of intel-company's most significant product strategy turn-arounds in quite some time. If we try to find origins of this new thinking at Intel then we end up at Banias, the 1st Pentium M. You were in charge of that product and are often described as 'man who changed Intel.' How does it feel to be credited with invention of such an important product?
    ---shmuel-mooly-eden: I don't believe that there is a single person who can change Intel. Well, perhaps CEO who is doing those big things. The article you are referring to had a nice head-line, but I would not go to such an extreme.
    ---TG Daily: So, if you did not invent Banias-processor then who invent Banias-processor ?
    ---shmuel-mooly-eden: I am not a modest guy and I am not trying to shy away from credit, but I don't think anyone at Intel can say "this is my baby." banias-cpu is an effort of hundreds of people. I was lucky enough to have been captain of ship back-then. And, if ship is successful and wins battle then captain receive a lot of credit. But there are many other people who do the work and deserve credit. For example, if you are manager then you should think wise enough to ask the right questions; you should think smart enough to pick-up the right answers and you should navigate every-thing. I receive the opportunity to do that. But there are also very talented architects - and those include architects who invented banias-architecture. As manager, you need to know how to how to solicit architects, how to motivate architects, how to encourage architects and how to move architects forward. If someone needs to be credited with the invention of the die of Pentium M, then I would have to say credit goes to team in Israel: design team, because of (Israel Development Center)/idc's brilliant idea for the best processor in the world. And management because we took big-risk and said "we will change paradigm." At that time, consumers were shopping for high-cpu-frequency and we said: "Guys, this is end-of-line. You need to do a right-hand-turn. Let's look at performance and other values beyond high-cpu-frequency."
    ---Banias and the risk for Intel
    ---TG Daily: What you effectively did, was putting giga-hertz-cpu-frequency in reverse. From a world that seemingly was driven, at least inside Intel, only by giga-hertz-cpu-frequency, you told people to throw-out foundation of intel-marketing-strategy. How much of a risk was it ?
    ---shmuel-mooly-eden: It was a huge risk for IDC (Israel Development Center). Banias market came just after Timna had been canceled [Timna was code-name for an integrated processor designed for entry-level market and originally scheduled for 2nd half of year 2000 - ed]. We had worked on Timna for 2 years and need to make sure that we didn't receive another project-cancellation-order from intel. In such a case, intel may lose confidence in idc. And worse than this, people may lose confidence. But the biggest risk in this industry is not-taking-risk, because then you are doomed. If you want to play-safe then you are out of the game.
    ---TG Daily: I would imagine that your idea to decrease clock-speed/cpu-frequency in half wasn't very well received at your 1st presentation to C-level executives. I remember a briefing from pat-gelsinger, in which pat-gelsinger out-lined a vision that CPU-frequency may reach 20 ghz/giga-hertz year 2009 and that kind of cpu could be compared to small-nuclear-electricity-generator. And there you are with a cpu that goes to other-direction. What was the reaction of engineers and marketing to your 1st presentation of Banias ?
    ---shmuel-mooly-eden: pat-gelsinger was actually 1 of the 1st people at Intel who figured-out/realise that giga-hertz-frequency's out-look would not be good thing. pat-gelsinger said that we must return, because we were running into a electricity-consumption-wall. You can blame many people for the increase in giga-hertz-cpu-frequency and electricity-consumption, but definitely not pat-gelsinger. At very early stage, pat-gelsinger was advocating/supporting exactly same thing what we were doing. We did that almost in parallel. But, of course, and with-out saying names, there were other people who felt that a lower clock-speed/cpu-frequency was too risky and they did not support low-clock-speed/low-cpu-frequency-idea. In very beginning, when we talked to people about our project, the 1st question we receive was "what's the cpu-frequency ?" We replied "1.6 ghz" and we receive answer "not interesting." that was not easy to push low-clock-speed/low-cpu-frequency-idea through. I do not want people to be misled: high-cpu-frequency was the right thing to do at that time. high-cpu-frequency gave you more performance and we had enough head-room. [foto:Die shot of Pentium M processor. This picture shows 2nd-generation Pentium M (90 nm "Dothan" core.)]
    ---TG Daily: Who do you think was most essential in getting a green-light for Banias ?
    ---shmuel-mooly-eden: There are several stages you need to look at. 1st, you have development-stage, where people aren't aware of that cpu. development is creating cpu and we are seeing where we are getting to. But after that, you need approval to move-forward and you do not get anywhere without CEO on your side. You can't continue, because forward-movement is a risk for Intel. For Banias, we receive support from all executives. ceo paul-otellini also embraced Banias.
    ---TG Daily: Do you think Banias will be remembered as CPU that changed Intel ?
    ---shmuel-mooly-eden: I would say Banias will be remembered as product that brought Intel on the right-track. Moving to low-electricity was inevitable. I am proud that we have been in the position of developing banias. But over-all, I believe Banias will be remembered as 1 of products that changed Intel. If you look at intel-company then 4004-cpu changed Intel, 8086-cpu also and 386-cpu also. I would be flattered, if Banias was remembered as 1 of mile-stones that put Intel in a leadership-position.
    ---TG Daily: Intel introduced its 1st electricity-saving technology - speed-step - back in jan-2000, in fact just 1 day before transmeta unveiled crusoe-cpu. There are people who claim that Intel may never have come-up with electricity-saving-technology with-out transmeta discovering electricity-saving-trend. Now Intel is making a turn-around largely because of pressure intel received from amd. Is Intel a company that now and then needs a step on its toes to see trends ?
    ---shmuel-mooly-eden: I was not in usa at the time-frame when speed-step was introduced. But I would not be surprised, if this was started way, way before. You need to work with many parties, including microsoft, to do something like that. Take, for example, yonah-dual-core. We started with yonah-dual-core-project many years ago; way before anybody ever spoke about yonah-dual-core-project. So, do we need some-body to step on our toes ? Hope-fully not. Is competition accelerating innovation ? Yes. Competition is good. I am not afraid of competition. The only thing I need to make sure is that competition is behind us.
    ---Getting personal: 25 years of Intel
    ---TG Daily: Let me give you a couple keywords, what's your 1st thought when you hear them ? Clock-speed
    ---shmuel-mooly-eden: A great opportunity that turned-out to be a burden.
    ---TG Daily: Timna
    ---shmuel-mooly-eden: The most brilliant engineering project, the biggest marketing failure. A great education for myself, in respect to the business.
    ---TG Daily: amd
    ---shmuel-mooly-eden: what? [pause] we are going to give them tough competition.
    ---TG Daily: 32-bit
    ---shmuel-mooly-eden: Part of evolution.
    ---TG Daily: Windows Vista
    ---shmuel-mooly-eden: No comment. It's the next OS, we need to look at it and see.
    ---TG Daily: Free time
    ---shmuel-mooly-eden: I love it.
    ---TG Daily: You have worked with Intel 25 years. Some-how, you don't fit the image of a 'dry' high-level manager of an IT corporation with 100,000 employees. Your casual "guy-next-door" appearance and your energy could be perceived as a much better fit for environment of a startup or a company like Google. How do you fit into Intel ?
    ---shmuel-mooly-eden: I believe that you always have different types of people and different styles. My style is the style that I have learned in Israel. I am very emotional and I am very direct. At the end of the day, I believe in the spirit of diversity. intel-company looks at performance, at risk-taking, at deliverables, at your ability to manage. bottom-line of my style apparently was positive, because I receive opportunity. So, over-all, I would say that some people enjoy my style and some have difficulty to digest my style...but that's me.
    ---TG Daily: Let's look back. 25 years ago, it probably was not an obvious choice in Israel to apply at a semi-conductor company. Why did you choose to work for Intel ?
    ---shmuel-mooly-eden: Well, I actually have 3 different Intels in my life. 1st, I was hired as part of start-up crew to build-up Fab 8 in Israel. It simply looks very challenging to start something from scratch. That was something totally new for Israel. I spent about 6, 7 years in Fab 8 in Jerusalem. We built EPROMs, but soon moved over to micro-processors and became biggest producer of 286s-cpu. EPROM is a very complex technology - if you can do that then you can do other technologies as well. But fab-8 was too disciplined for me and I am not a disciplined person in my nature. I looked into ways to change my career, to change from being a staff-member of Fab 8 to become a member of idc in Haifa. My goal was to spend about 1 year there (idc) and get up to speed in design-community. I did some design and I managed cache-team. There were some problems with Pentium-MMX in year 1995. I believe that this was opportunity to show my capabilities as a manager, because Pentium-MMX was a tough project. I eventually became manager of idc (Israel Development Center), where I was managing about 1500 people. My 3rd Intel is the intel-company in usa. I receive opportunity to become Director of Marketing and introduce Centrino. I do not know many places that give you that kind of opportunity.
    ---TG Daily: How difficult was the feeling to leave Israel and your achievements and home behind ?
    ---shmuel-mooly-eden: moving from israel to usa wasn't a big move. When I was asked to move, I was asked to introduce Centrino. At the time, I was general manager of idc and they asked me to move as an individual contributor, as a program manager. I was simply flattered. But the plan was that I would introduce 1 million units to usa-market. Once manufacturing-machine was up and running, I thought they would not need me anymore. And, actually, after 7 months, I was done and I returned to Israel. After 2 months, I receive offer to come-back to usa to do marketing. I knew already a lot about usa and how my family would adapt.
    ---A different work style in a different culture
    ---TG Daily: Work in usa over-all was probably a different world. What differences did you see between work-culture in Israel and usa ?
    ---shmuel-mooly-eden: Coming from a different culture, I believe that people are generally making a mistake by grading or judging another culture by using their own culture. I am often told that Israeli people are blunt. We are not blunt. we are just saying our opinion. Some-times I tell my partners that they do not care about anything because they do not argue. And they tell me that I argue all the time. But all I do is care about it. When you come from a different culture, you get into a lot of problems. For example, how do you tell someone that you do not agree with him ? If I do not agree in Israel then I simply tell you that you are wrong. Here in usa it is almost offensive to say that. Instead, people in usa say "I am not following you. Can you reiterate ?" But, what it really means is that "you are wrong." If you come from another country, you get into situations where you can sound offensive. However, from my perspective, that is also what is nice about living in another country: You make new experiences.
    ---TG Daily: Would you say that America changed you ? Or do you change the people around you ?
    ---shmuel-mooly-eden: I am changing in the way how I am behaving in usa. I speak differently. At the end of the day, I want to achieve some-thing and I do not want to insult people. But I am also trying to change every-one around me to what I think is right. And that is to influence them, to being direct, to saying what you think, to elevating problems. I love problems. I learned that people in usa have a problem saying "we have a problem." Instead, they say "we have a challenge." I don't think problem is a challenge, why don't they say "problem ?" It's a challenge to solve problem.
    ---TG Daily: Do you think that a different work culture, a more direct way to work with each other was helpful to come-up with idea for a completely different processor ?
    ---shmuel-mooly-eden: There have been a lot of great ideas in usa. Think about 286-cpu, 486-cpu and pentium-cpu. But I would agree to say that culture in Israel helped to develop Banias, Yonah and Merom. We are very friendly, we work together in teams, but discipline is not our strength. The lack of discipline results in a behavior where every-body is challenging every-body. People are challenging you all the time. People are not influenced by your rank. They appreciate you for what you are. It's a different thing. If you steer that behavior into the right way then you make a lot of innovation.
    ---TG Daily: What was the most profound lesson you learned in your career ?
    ---shmuel-mooly-eden: We learn from mistakes. There was Timna, which made me much more open. Performance or design is not good enough, if you do not have a deep under-standing of the market. I was responsible for Timna and the writing had been on the wall. The marketing appeal disappeared since Timna-project was based on RDRAM, which did not penetrate market fast enough. I should have killed timna-project 2 quarters earlier. But I was so much in love with Timna-project and we did not stop Timna-project. I learned that you must respect marketing. You need to under-stand the market very well to determine if the product will be successful. Also, Pentium-MMX-project clearly indicated importance of management and leadership of technical people. With the right management and leadership you can make Pentium-MMX-project happen. Take same project and a different management then that project could fall-apart. management and leadership can make difference between failure and huge victory.
    ---TG Daily: What goals do you have left at Intel ?
    ---shmuel-mooly-eden: I would like to continue to do something interesting. I believe mobile is a huge opportunity. There is some-thing that you can say when you work at Intel and it almost sounds arrogant: You leave foot-print on that evolution, you and your team. Look at Centrino, every-one in this world has a Centrino-laptop and you look at Centrino-laptop and can say "I was involved in Centrino-laptop" and that makes me very proud. It will take 20 years until people will recognise Centrino-laptop, because we are in the middle of a huge revolution. I hope I will be part of this moving forward.
    ---TG Daily: What advice would you give someone who would like to pursue a similar career as you did ?
    ---shmuel-mooly-eden: I would say to just do it. Don't copy. Do what you believe in. Try to find a job you love, don't compromise with some-thing you do not like. In your job, behave patiently because an opportunity does not automatically-appear daily. Never become a yes man, say what you think, argue for your opinion. At the end of the day, your opinion will be appreciated.
    ---TG Daily: Thank you for the interview"
  7. ^ Shimpi, Anand Lal. "Intel's Centrino CPU (Pentium-M): Revolutionizing the Mobile World". Retrieved 2019-12-05.
  8. ^ PAE - Ubuntu Community Help Wiki
  9. ^ Intel Outlines Plans For Wireless Notebook PCs, Cell Phones And Handhelds, Intel Corporation
  10. ^ Intel launches Dothan with Pentium M price cuts, The Register

External links[edit]