The Intel Processor Information Guide - <<READ BEFORE POSTING>>

[Serg]

The Intel Processor Information Guide V2.5

Always trying to help, and glad to!
Serg

UPDATE to the 29/11/2009
Feel free to post and leave suggestions, updates, impressions, thanks.

Index

1. Reason for the Post
2. Definitions and Architecture
3. CPU Classification and Nomenclature
4. Atom, ULV and LV series CPU - Netbooks, Ultra-thin, Ultra-portable, Tablets
5. P and T series CPU, Curious Fact T vs. P - Full range, from 12" to 17"
6. X,Q and QX series CPU - Desktop replacements, Workstations, Gaming Rigs
7. Mobile i7 (Nehalem) CPU - Desktop replacements, Workstations, Gaming Rigs
8. Arrandale 32nm (Westmere) CPU, Sandy Bridge Appears and 22nm is working - Mainstream, Portable, Thin-and-Light, Desktop Replacement
9. Conclusion
10. Author's Notes
11. Useful Links

[Serg]

1. Reason for the Guide

__First off, hi to everyone who stops by and for all of you newcomers, welcome to NBR, the most complete site to ask about notebooks and help in the matter. Whether you are looking for a new Notebook for you or you are having problems, or simply want to learn, NBR is the place to come and freely ask around and learn.
Well, I have been a member of NBR for quite some time now, and since day one, every time I've asked anything, someone has helped me very kindly. This is a small contribution to NBR and my way of thanking them. Therefore, I decided to give some information about CPUs for those who might need it.

__Unfortunately nowadays there are tons of CPU, and naming, numbers, speeds, it is hard to understand it, so I put together this Guide to help you understand a little more. Even more, as I rewrite now, we are between a major change in CPU architecture. Having both Merom (Core) and Nehalem (ix) at the same time, and shifting, things get a little confusing, so the object of this guide is to help you understand a little more about them, covering the basics and to help you know which one fits you the most.

__After all, at some point, we were all newbies on this matter, and a little extra knowledge is always welcomed. So next time you are wondering what CPU to choose, you can have swing by here, a little guide to help you thru that awful decision. And this confusing times and branding.

__This is a continuous work on progress, and everyone is invited to help and provide information and sources, contributions are always welcomed. It is always good to know more. Just shoot me a PM or leave a comment and I will update.

Back to Index

[Serg]

2. Definitions and Architectures

I) Processor Introduction
So, we all hear about the computer parts, and the CPU is one of the most important parts. We always hear about it, but what is really? In plain words, the Central Processing Unit is the brain of the computer, as the name suggests it. It is the one taking care of your tasks, when you give an order, the CPU is the responsible for that order, and making sure it is done properly.

In simple words, you give a command to the computer. The software sends the command all the over the CPU. The CPU processes the command and interprets. Does whatever the command said that should be done. Then it returns to the origin with the answer. And it is displayed on your screen. All of these things are done in matter of milliseconds.

Here is a common layout of a CPU:


I.1) Definition of a thread
This new area was created for those who are slightly confused on what a "thread" is. If you are familiar with the concept of a thread, then you can skip this.

A thread is a set of commands, a single order that goes into the CPU to be processed. Running an antivirus is one command, running Word is another, viewing a movie is another, and so on. But these commands have "sizes" and requirements of resources, that is why opening Word and running it causes no problem, it is a "light" command, "Open Word, Keep Word Open, Stand-By for further orders" and the CPU goes to idle if nothing else has to be done. Running an antivirus is "heavier", since it requires more resources to run (and that is why the CPU usage goes up to 70% or so, depending on your CPU), the command would be "Run AV, Search For Files, Search For Objects, Compare to Database, Show Count of Items, Search for errors, Report Errors, Report Objects found, etc". It is a more complex command, but only one nonetheless. To simplify it, a thread is a set of things to do. Referred to it as command.

II) CPU parts
This area will be summarized, since we could write entire guides and threads about each part.

In classical architecture, the CPU is mostly composed by 3 main parts, the regular numbers we hear when specs are mentioned:

- First of is the Core itself and the speed at which it is running. The core is the actual processing plant. It is the one that reads the information and interprets what to do, mostly responsible of what is going on your computer. The Core speed is noted in GHz normally. This is how many revolutions it is doing per second, the faster it turns, the faster it completes to process. You can have as many cores as you want and each core handles each command, also known as thread.

- Secondly is the Front Side Bus (FSB). FSB is the link between the Cores and the Northbridge. The wider or bigger the FSB, the faster the information is sent and received between these. This is commonly expressed in MHz. It is the wide at which the speed goes thru. Consider this as a bridge with traffic. The wider the bridge, the best it handles traffic, same goes for the FSB and info.

- Third is the cache. There are 3 types. The L1 is integrated in the core, the fastest of all memory types. The L2 Cache is a specialized memory unit that enhances performance by providing the CPU with data at speeds ten times faster than DRAM. Here bits of used information are stored for the CPU to quickly reach them, as little codes. It is shared. The 3rd type is L3, only present on Quad Cores. It is massive and works the same way. Lx Cache was created so that the CPU does not have to access the RAM or the HDD in search for instructions, saving time.

- Other parts to mention that play an important role, are the Northbridge, the center which distributes the information between the CPU and the peripherals (GPU, RAM) and the Southbridge, which connects to PCI ports, USB ports and SATA (HDD and Optical Drive Bay).

Unfortunately for those who don’t know much, the thing gets a tad more complicated now. Intel has finally released Core i7 mobile CPUs, which introduce a completely new architecture. Just go up, re-read about the Northbridge, now forget that. Northbridge is gone on i7 CPUs, and if you remember, the FSB served as a bridge to the CPU, now that is gone too. What Intel did was get rid of the Northbridge to accelerate data transition between the parts, accelerating the entire process described above. So if you are searching for a new processor, what you need to know is the following.

- Instead of FSB, Intel developed what they call Quick Path Interconnect (QPI). The QPI is, in its simplest form used to connect the processor to the Input/Output Hub, for example, connect an Intel Core i7 to an X58 chipset. It is a point-to-point processor-interconnect. To make it simpler, it is a connection that is much faster and saves time, it is basically all you need to know about it.

- The appearance of the Massive L3 Cache as shared by all Cores, speeding the search. And L1 and L2 are core-dedicated, meaning they are built-in the core itself.

Here is an image of how the Meron layout compared to Nehalem layout. Notice that in contrast to the classical architecture, Nehalem is simpler, as it has lesser parts where the information has to go thru.



As said above, the Northbridge, or information distribution center, was eliminated, and the FSB disappeared along with it. And this is only the hardware (physical) changes.

As you can see, the design is simpler, but by no means is this bad. What other things are included? Intel made sure you had a headache when trying to understand.

So to let you know what CPU is based on what architecture, either Merom (Core Architecture) or Nehalem Architecture, take a look at this graph:


III) Architecture
So far it is a bit complicated to understand, so I will now give a small explanation on these parts and how they work on each different architecture. What you need to know is that there are two major families on the market, Core Architecture and derivates, and mobile i7 Nehalem. I will not enter into naming, such as Merom and Clarksfield, to keep it simple.

- Core Architecture.
This is the one we all know. It covers most CPUs in the market. They are what I refer as Classical Architecture, which I referred previously. The design includes more parts, but we are all used to it. What is so interesting about Core Architecture? These CPUs carried the ability to run 64bit software, and introduced a fully functional virtualization support. Really there is not much to explain, unless you want to go into technical stuff. Something important to show about this architecture is the flexibility it had to adapt into today’s Ultra Low Voltage (ULV) CPUs.

- Mobile i7 (Nehalem) Architecture.
Just launched, the Mobile i7 Architecture is based on the desktop similar line. Designed with lower power consumption and high performance in mind, this new architecture introduced a plethora of new things into play. They are all Quad-core natives, with some little surprises.

First off, there is Turbo Boost (TB). Turbo Boost is the name for the native capability of Nehalem to overclock. Of course, TB has some conditions. How does this work? You have four cores in your CPU; in the Core Architecture, they would always be running, at minimum, but still running. On i7, when you have 4 cores, but you only need one, the remaining 3 are shut down until needed. So now you have a single core running, this is when TB kicks in. Since there is no more power consumption, the only core running can get a little extra aid, self-overclocking until the CPU considers it is safe, and a 1.6GHz core turns into a 2.8GHz automatically. Also applied when 2 cores are running, but overclock will be lesser, remember that it has to maintain within a safe temperature. When 3 or 4 are running, TB might give a small boost, but not as noticeable, considering the max TDP of the CPU. Follow this link for a graphic explanation (image was too large).



Second major change is the return of Hyper Threading (HT). If someone remembers the day of Pentium 4 when HT was present, basically the same, but greatly improved. For those who do not know what Hyper Threading is, HT is the capability of the CPU to run two threads per core. In the Core Architecture it was limited to one-core-one-thread, with Nehalem you can run two threads per core, so this means that a quad-core can deal with eight threads at the same time.



This is something that makes multi-tasking easier and better, since HT and TB can run together. Remember I said you had one core self-overclocked? That overclocked core is doing two things at the same time, and if you are using two cores, it means you can do up to four things at the time.

NOTE: In the task manager, and in marketing, they are referred as “cores” because the system recognizes them as such, but they are “virtual cores”, which means, they are not there, simply is the same core that does two things at a time.

As some might have imagined, all these options help to achieve the objectives of Nehalem. When CPU is idling the consumption drops, saving battery life; and when stressing the CPU the consumption rises drastically to feed such a powerful CPU.

Back to Index

[Serg]

3. CPU Classification and Nomenclature

We are now entering into the heavy zone, where numbers and letters start mixing up, to give name to each CPU. I will try to make it simple, but remember, there are many variants on the mobile market.

First let us start by the letter in front of the number. What does this letter stand for? It stands for the Thermal Design Power (TDP), which represents the maximum amount of power the cooling system in a will require to dissipate the heat. It is noted in Watts [W].

1. QX - Mobile Quad-Core Extreme Performance (>40W)
2. Q - Mobile Quad-Core Performance (>40W)
3. X - Mobile Dual-Core Extreme Performance (>40W)
4. T - Mobile Highly Energy Efficient (30-39W)
5. P - Mobile Power Optimized Energy Efficient higher performance (20-29W)
6. L or LV - Mobile Highly Energy Efficient (12-19W)
7. U or ULV - Mobile Ultra Low High Energy Efficient (5.5-11.9W)
8. SP - Mobile Small package Power Optimized - Energy Efficient higher performance (20-29W)
9. SL - Mobile Small package Highly Energy Efficient (12-19W)
10. SU - Mobile Small package Ultra Low High Energy Efficient (5.5-11.9W)

But CPU naming is not limited to a letter, which designates the microprocessor’s classification, but it is also accompanied with four numbers.

-___The first number classifies the CPU by L2 Cache size.

o In full-sized CPU (LV/P/T/X/Q/QX) the 9xxx stand for 6MB of L2 Cache, 8xxx and 7xxx stands for 3MB L2 Cache, 6xxx stands for 2MB L2 Cache.

o In ULV CPUs, the thing changes; 9xxx, 8xxx and 7xxx are 3MB L2 Cache, they differ in clock speed. And 4xxx has 2MB L2 Cache, but with a slower clock, and named Pentium ULV.

-___The second number designates performance. The higher the number, the faster the clock speed, but this only applies after you have taken into consideration the letter classification, and first number Cache-Size-Designator. For example a T9900 is faster than a T9600, and you can tell this by looking at the same number, but note they are both T-classified and 9-series. But a P8700 is not faster than a T9600. Note they are P and T-classified, and 8 and 9-series.

-___The third and forth numbers designate the version of the CPU. The higher means the newer the refresh or revamp the CPU is, and that it the speed is higher than the original same chip.

Back to Index

[Serg]

4. Atom, ULV and LV series CPU

I will start from the smaller, less power-hungry ones, and scale all the way up to the bigger ones. So, starting the count up we have the Atom CPUs, made for the smallest of all notebooks, the netbook. Following up there are the ULV CPUs followed by the bigger LV ones.

I.___Atom CPU
Atom is the brand name of a line of extremely low voltage CPUs used mainly in netbooks (ultra-portable laptops with more limited performance) and nettops (very small desktops). Until now, there have been two main series: The first was the Z series, code-named Silverthorne. It was followed by the N series, code-named Diamondville. The N270 by far the most commonly used CPU in netbooks in the past year.

The Z and N series processors are single-core with 512 KB of L2 cache, and have a TDP from 2 to 5 W, significantly lower than same-generation Pentium or Core 2 Duo CPUs. They all support hyper-threading. For the complete family of Atom CPU, follow this link.

The current chips are the following:
CPU____Cache____Clock___FSB___TDP
Z550___512KB L2 2.00GHz 553MHz 2.4W
Z540___512KB L2 1.86GHz 553MHz 2.4W
Z530P__512KB L2 1.60GHz 553MHz 2.2W
Z530___512KB L2 1.60GHz 553MHz 2.0W
Z520PT_512KB L2 1.33GHz 553MHz 2.2W
Z520___512KB L2 1.33GHz 553MHz 2.0W
Z515___512KB L2 1.20GHz 400MHz 1.4W
N270___512KB L2 1.60GHz 553MHz 2.5W


Comparison of N270 Diamondville with N450 Pineview

The N450 is 1.66GHz and the N470 is 1.86GHz. Both Atom have the following characteristics:

1. 45nm Single Core CPU
2. Supports Hyper Threading
3. Integrated GPU
4. 64-bit Capable
5. Supports DDR2 RAM
6. Designed for fan-less Devices

UPDATE: Newer Intel Atoms seem to carry a huge improvement finally in the graphic area, where the GMA3150, will power the new Atom D410 and D510. These new improvement is based in the old GMA X3100 Intel chipset, but with a slightly up in performance, so after all, we might see some not-so-low-end-graphics on a netbook, though still the X3100 was not good, it is a major improvement over the G945GC they used to carry. More info provided by Fudzila here.

UPDATE 09/11/2009:
Intel Corp. will reveal more details about its next-generation Intel Atom platform for netbooks and nettops on the 21st of December ’09 and computers makers will launch actual systems powered by Intel Atom N450 and derivatives on the 4th of January 2010, just in time to show them on the Consumer Electronics Show.

“Intel is planning for a fast transition to Pine Trail. To generate excitement for the platform ahead of launch, Intel is planning a press release in late December publicly disclosing the details of the platform,” an Intel document seen by X-bit labs reads.

Intel Pine Trail platform includes next-generation Atom “Pineview” processor with improved performance, integrated graphics core, built-in DDR2 memory controller, x86-64 capability, a new core-logic set as well as a Broadcom video decoding processor

According to sources close to Intel, there will be several next-gen Atom “Pineview”chips for netbooks and nettops unveiled:

• Intel Atom N450: single-core with Hyper-Threading support, 1.66GHz, 512KB cache, x86-64, BGA437 package, $63 price-point;
  Intel Atom D510: dual-core with Hyper-Threading support, 1.66GHz, 1MB cache, x86-64, BGA437 package, $63 price-point;
  Intel Atom D410: single-core with Hyper-Threading support, 1.66GHz, 512KB cache, x86-64, BGA437 package, $43 price-point;

We can expect this new Atom models by March with the launch of some new Asus models at CES 2010.

II.___Ultra Low Voltage CPU
As said above, ULV stand for Ultra Low Voltage, or also known as Consumer Ultra Low Voltage (CULV). Largely underestimated, the ULV CPU was designed for mobile users who perform regular tasks and play some multimedia content on-the-go. Designed with ultra-thin clients in mind, the ULV was conceived to perform good enough for most users, and capable of giving long battery life (with even some models reaching the 10 hours mark). With around only 5.5W of power consumption (TDP), the ULV is the most power-efficient Core-Architecture-based CPU currently on the market. They are what some might call “stripped-down” versions of the full-sized CPUs, same architecture, same softwares enabled on board, but slower clocks and smaller L2 Cache.

We have seen many laptops coming out with this Ultra Low Voltage CPU, they are all aimed at mobility and high battery life in thin and light laptops, where the very small enclosed chassis is not the best environment for a lot of heat. They provide extremely good performance for the aimed market

The current chips are the following:
CPU____Cache____Clock___FSB___TDP
SU9600___3MB L2 1.6GHz 800MHz 10W
SU9400___3MB L2 1.4GHz 800MHz 10W
SU9300___3MB L2 1.2GHz 800MHz 10W
SU3500___3MB L2 1.4GHz 800MHz 5.5W
SU3300___3MB L2 1.2GHz 800MHz 5.5W

III.___Low Voltage CPU
Following the classification, the next in the line is the LV series. Very much like the ULV but these were conceived before. ULV is the latest family to be added and fully implemented. LV on the other hand was conceived some time ago, but has not seen much light on the market. These CPU are in the middle of the scale, offering a great performance compared to power consumption. Originally designed to make laptops smaller and keep on par with larger notebooks, these CPUs had a slightly higher price tag than the ULV, and expansion of these was very small.

Some very famous models to carry these are the Apple MacBook Air (MBA) which debuted some time ago, and the Lenovo IdeaPad U110. Both had the problem of high price tags, unfortunately.

The current chips are the following:
CPU____Cache____Clock___FSB____TDP
SL9600___6MB L2 2.13GHz 1066MHz 17W
SL9400___6MB L2 1.86GHz 1066MHz 17W
SL9380___6MB L2 1.80GHz 800MHz 17 W
SL9300___6MB L2 1.60GHz 1066MHz 17W
L7700____4MB L2 1.80GHz 800MHz 17W
L7500____4MB L2 1.60GHz 800MHz 17W
L7400____4MB L2 1.50GHz 667MHz 17W
L7300____4MB L2 1.40GHz 800MHz 17W
L7200____4MB L2 1.33GHz 667MHz 17W

Back to Index

[Serg]

5. P and T series CPU

Right on the middle of the entire CPU family we find the most common CPUs around. The P and the T series are the best known around. They range from Budget to High-End (reference to the Intel Processor Family Guide). Known by most of us, and by almost everyone looking for a laptop, they are the most powerful CPUs that available for regular consumers. Here is a CPU comparison.

I.___P-series CPU.
Released when the Core Architecture got a refresh, passing from Santa Rosa Platform to Montevina Platform, they were the last to be added to the long list of Intel CPUs that are available and running two cores. What characterizes these is the lower TDP of 25W when fully stressed vs. the 35W of a similarly clocked T-series.

The P series are designed to provide a full and powerful performance, and trying to preserve battery as much as possible. You will notice that the specifications on both P and T are the same, except for Thermal Design Power.

NOTE: Although they carry a smaller TDP, when idling, both P and T consume pretty much the same.

Laptops of every type and manufacturer have carried these CPUs, and they are highly recommended due to their lower power consumption and heat emission.

The current chips are the following:
CPU____Cache___Clock___FSB___TDP
P9700___6MB L2 2.8GHz 1066MHz 28W
P9600___6MB L2 2.66GHz 1066MHz 25W
P9500___6MB L2 2.53GHz 1066MHz 25W
P8800___3MB L2 2.66GHz 1066MHz 25W
P8700___3MB L2 2.53GHz 1066MHz 25W
P8600___3MB L2 2.40GHz 1066MHz 25W
P8400___3MB L2 2.26GHz 1066MHz 25W
P7550___3MB L2 2.26GHz 1066MHz 25W
P7450___3MB L2 2.13GHz 1066MHz 25W
P7350___3MB L2 2.00GHz 1066MHz 25W
SP9600__6MB L2 2.53GHz 1066MHz 25W
SP9400__6MB L2 2.40GHz 1066MHz 25W
SP9300__6MB L2 2.26GHz 1066MHz 25W

The SP-series are the smaller design, made to fit thinner laptops, such as the Lenovo T400s.

II.___T-series CPU

The most known of the plethora of CPUs out there and the most mainstream of them all, the T-series CPUs have been around for quite a long time, all the way since Core Architecture debuted, after Pentium. They are cheaper than a comparable P series, due to it not being power-optimized, but in real life, they are basically the same.

Performance is almost top-notch, and for someone on a budget, this is the best you will get, and what you should look for if you want full performance for multitasking, gaming, or even regular activities.

NOTE: Something I want to leave clear, the now-called Pentium CPUs are in fact Core 2 Duo with smaller L2 Cache and a slower FSB, but same Core Architecture.

Most laptops, from 13 to 17+ inches have carried or carry these CPUs, they are the most mainstream of them all, and very easy to find. Every single manufacturer is using or has used this on their models.

The current chips are the following:
CPU___Cache___Clock___FSB___TDP
T9900___6MB L2 3.06GHz 1066MHz 35W
T9800___6MB L2 2.93GHz 1066MHz 35W
T9600___6MB L2 2.80GHz 1066MHz 35W
T9550___6MB L2 2.66GHz 1066MHz 35W
T9500___6MB L2 2.66GHz 800MHz 35W
T9400___6MB L2 2.53GHz 1066MHz 35W
T9300___6MB L2 2.50GHz 800MHz 35W
T8300___3MB L2 2.40GHz 800MHz 35W
T8100___3MB L2 2.10GHz 800MHz 35W
T6500___2MB L2 2.10GHz 800MHz 35W
T6400___2MB L2 2.00GHz 800MHz 35W
T4300*__1MB L2 2.10GHz 800MHz 35W
T4200*__1MB L2 2.00GHz 800MHz 35W

*Pentium CPU. As said above, based on same Core Architecture, with smaller L2 Cache than a similar Core 2 Duo.

As said above, these are the most common, and offer a good bang-for-the-buck performance. They produce a tad more heat and consume more when fully stressed compared to the P series, but for the price/performance ratio, it is a trade off most end-users can accept.

III.___Curious Fact - T vs. P

Posted by K-TRON, a fellow NBR Forums Member, which is a very experienced person in the hardware are, I have found this pieces of information, and under K-TRON's consent I am posting this here to help people clarify one of the biggest doubts that exists when choosing a CPU for your laptop. Thanks again K-TRON for the information provided.

Extracted from a recent thread where it was asked between a T-series CPU and a P-series CPU about performance. Contrary to what we are used to think, a fellow NBR Member has help us and proved something very important. The info is next and quoted:

The T-series falls in the 35 watt TDP, but that doesn't mean it uses 35 watts.
The P-series falls in the 25 watt TDP, but that doesn't mean it uses 25 watts.

While being stressed the P should be around 23W of TDP in use, while a similarly clocked T-series should be near to 27W.

This is because processors that do not fall in the less than 25 watt TDP zone get bumped into the 35 watt TDP bracket.
So a 26 watt processor will fall in the 35 watt TDP bracket. In reality, they will use close to the same power, and run the same temperatures.

After reading this, I asked how could this be, and here is the reply, proving what it is above.

I tried an indentical clocked T9600 and P9600 series chip in the same laptop - HP DV5
The same heatsink, thermal material, and fan were used for this test, so basically the only variable is the CPU.
The real life differences I saw under full load between the T9600 and P9600 was 3C under full load.

Thanks to this piece of information I think the P vs T thing is a bit more clear.

Back to Index

[Serg]

6. X, Q and QX-series CPU

Reaching the top of the line in the Intel Mobile Core Architecture we find the Extreme editions and the Quad cores. Over these CPUs, we will be changing Architecture, since this is the limit of the Core Architecture.

I.___X-series CPU
Considered by most people the best Core 2 Duo, this is the most overclockable too, according to some users and members of NBR. Its power draw is quite high though, and heat emission is also high, it is not strange we only find these CPUs on high-end fully-featured desktop replacements.

Consuming a whooping 40+W these are not to be found on thin and light laptops, most likely only on gaming rigs and mobile workstations, which can deal with their requirements a lot better.

The current chips are the following:
CPU___Cache___Clock___FSB___TDP
X9100___6MB L2 3.06GHz 1066MHz 44W
X9000___6MB L2 2.80GHz 800MHz 44W

They are a very limited series, since more than this is basically not needed.

II.___Q-series CPU
The Q-series CPU were born when Intel decided to put two Core 2 Duo together and link them via a massive L3 shared Cache. They are on the top of the performance when multitasking. The only trade off for these Q-series is the high clock speed in exchange of four cores. They are normally found on 15+” laptops, with some exceptions going down to 13", like the MSI EX300, since they produce a lot of heat and consume a respectable amount of your battery, 45W of power draw in fact. They are very limited in numbers, much like the X-series.

Quad cores were introduced in August 2008, they were designed to handle massive compute and visualization workloads enabled by powerful multi-core technology. They are Core 2 Quad processors are built on 45nm Core micro architecture.

The current chips are the following:
CPU___Cache___Clock___FSB___TDP
Q9100___12MB L2 2.26GHz 1066MHz 45W
Q9000____6MB L2 2.00GHz 1066MHz 45W

III.___QX9300 CPU
And we have reached the top of the Core Architecture. Above this there is nothing else. We have arrived to a chip that is classified as extreme and quad. This processor is the top-of-the-line on mobile computing. It sure proves to be extremely powerful, as well as expensive.

Only to be found on gaming rigs and mobile workstations, this CPU is not your best friend if you want to save battery, consuming 45W, heat is also quite high. Some notebooks might be able to get this inside, but I do not recommend it, it is massive, it is powerful, it is expensive, it is the top-notch from Intel in the Core Architecture.

Currently there is only one chip,
CPU____Cache____Clock___FSB___TDP
QX9300___12MB L2 2.53GHz 1066MHz 45W

We have reached the end of the Core Architecture. And now we are evolving into the recently launched Mobile i7. Continue to discover what this bears.

Back to Index

[Serg]

7. Mobile i7 (Nehalem) CPU

Just launched some time ago, we finally see the Mobile i7 hit the market. It is also known as Clarksfield, Calpella platform, Nehalem Architecture, but I will refer to it as i7 only, to avoid confusion with the Desktop Clarkdale. As said above in the Architecture part, Core i7 is a totally different architecture. It brings tons of new things into play, including the afore-mentioned Hyper Threading and Turbo Boost.



Built in the same 45nm (nanometers) in which the Core Architecture is manufactured, it is a completely new world of things. It gives a better performance when needed, and power management has greatly improved.
They are native quad-cores with up to eight possible threads running at the time. And the base clock speed is a reference, since in fact these cores run in a range, thanks to Turbo Boost.

As noted in the Chapter 2. Architecture, they design incorporates more parts on the chips:


This CPUs are limited to 3 at the moment, and are only to be found on 15+” laptops, which were launched very recently. Some of these are the Dell Studio 15 and Studio 17, the Dell XPS 16, the Alienware M15x refresh, the HP Envy 15, the Sager NP8690.

The current chips are the following:
CPU____Cache_____Clock_____TDP
i7-920XM 8MB___2.00-3.20GHz 55W
i7-820QM 8MB___1.73-3.06GHz 45W
i7-720QM 6MB___1.60-2.80GHz 45W

Here is some more detailed information on how TB kicks in and boosts speed on the cores:
Details on Mobile i7

Do not let the low clock numbers scare you and make you think they are underpowered; in fact the i7-Q820QM is on par with the QX9300 mentioned above, thanks to Turbo Boost. And even though it has all that power, when running on battery, it saves a lot more.


Note how the QX9300 and the i7-820QM are almost on par.


Note the difference between Core Architecture (Penryn) and Nehalem Architecture (Clarksfield)

We are right in the point where we have both technologies at the same point, and are getting ready to pass from Core Architecture to the future fully implemented Westmere 32nm CPUs, the next to come, based on the same architecture but reaching every level of the market, in the dual-core version.

Back to Index

[Serg]

8. Arrandale 32nm (Westmere) CPU

So we arrive at the "what is to come" area, where we have some leaked info, but still waiting for these to come. What is new? Westmere is manufactured with the introduction of the 32nm micro-architecture, which will enable processors to be even lighter, faster and even less power-hungry. Technology is not walking, it runs. Just in August 08 the 45nm was introduced, a year later the 32nm is going into production.

Have you ever dreamt with an i7 on a laptop? It will be possible by 2010 when the 32nm reaches full mass production. Another advantage of the Westmere line, which is called Arrandale for laptops and Clarkdale for desktops, is that they integrate the graphics and the memory controller. Named i7 the high end, i5 the mainstream and i3 the thin and light.



More details are found here. But a small summary is that the Arrandale while be able to reach 4GHz, and switchable graphics as a standard. It was reported even a 25% better performance than previous Intel GMA chips 4500MHD. As a little update, it seems this new IGP is called 5700MHD.

Right from Intel, the confirmation of this. Here.

For all of you, here you can see what to expect:









It is confirmed that Arrandale chips have the memory controller integrated and an IGP comes with the package as a standard.


EDIT: Information confirmed here:

Source Tom's Hardware

UPDATE 15/10/2009:

8.1. Sandy Bridge 32nm CPU Debuts

As if Intel's i7 were not enough change, Sandy Bridge, the next "Tock" in Intel's line of CPUs, has been shown, and it is working, based on a 32nm build process. Called the "true" 32nm, it is a new architecture based on the Westmere shrinking that is slated to go to 22nm. Arrandale is good-to-go, and in late 2009 or early 2010 it might come out as standard, but just for all of us that like technology, we have been shown that 32nm Sandy Bridge is now working. And the 22nm shrinking is due to 2011.

What is so great about this? Remember the Northbridge gone in i7 architecture? Well, that was not enough for Intel, the Southbridge is disappearing too. It will integrate a new graphics core into the chip die itself, improving the performance of integrated video significantly compared to the existing Nehalem's separate graphics. These designs should also incorporate Advanced Vector Extensions, a new set of extensible instructions that can handle 256-bit data (versus 128-bit for SSE4) and process four tasks at once. To see it, click here.





Sandy Bridge, is the next big change, and it is already working. It takes the graphics and joins them to the CPU, and it is done.



UPDATE 29/11/2009:


UPDATE 17/12/2009:
First images of the Arrandale:

Back to Index

[Serg]

9. Conclusion

As technology runs advances at huge steps, mobile computing is becoming more embraced by everyone. Laptops are becoming more and more the main PCs at some homes. Desktops are being replaced by their smaller cousins, and that is what pushes companies like Intel to be changing all the time.

As a teacher of mine once said "Qui non progredi, regredi est", which would translate to he who does not goes forwards, is going backwards. The same applies to technology, and as time passes, more advances are made, and in the laptop world it is very noticeable. Trying to keep up with this trend is very hard, almost impossible, but it sure is rewarding to know that you have the best laptop among your friends.

Processors will only get smaller, faster, stronger, more energy-efficient and more powerful over time. We just have to get used to seeing new things coming up all the time, which is very interesting.
Hope you can find the ideal processor for you.

10. Author’s Notes

I started this so I could let people now a bit more about what's powering their laptop. Many don't have an idea, or rush to buy anything without knowing, so I hope this is helpful for all of them.

This is a small contribution to a large community such as this, so I hope this was useful and helpful for you. I'm open to edit this anytime and I will gladly accept suggestions and updates on what's going on out there.

If you can provide some good sources of information, feel free to PM me at any time, I am always around, and I will update the post with the info.

Back to Index

11. Useful Links

Intel Processor Family Guide by Serg
AMD and Intel Cache architecture by Serg
The Ultimate CPU Guide by Lithus
Intel Mobile Processor Compatibility Guide by Commander Wolf
The Layman’s Guide to 64-Bit by Lithus
Sticky Hardware General Info Sticky Index by Chaz
IDF 2009 by Jayayess1190

11.1 External Links

Intel CPU PDF by Intel
ARK | Your Source for Information on Intel Products by Intel
Intel Microprocessor Quick Reference Guide by Intel
Intel Atom Processor: Intel's Smallest Chip by Intel
Intel Pine Trail (Next-Gen Atom Platform) by PC Perspective