Pentium 4

Pentium 4

File:Pentium4ds.jpg
General information
Launched	2000
Discontinued	2008
Common manufacturer	Intel
Performance
Max. CPU clock rate	1.3 GHz to 3.8 GHz
FSB speeds	400 MT/s to 1066 MT/s
Architecture and classification
Technology node	0.18µm to 65nm
Microarchitecture	NetBurst
Instruction set	x86 (i386), x86-64, MMX, SSE, SSE2, SSE3
Physical specifications
Sockets	Socket 423 Socket 478 LGA 775
Products, models, variants
Core names	Willamette Northwood Prescott Cedar Mill

The Pentium 4 brand refers to Intel's line of single-core mainstream desktop and laptop central processing units (CPUs) introduced on November 20, 2000 (August 8, 2008 was the date of last shipments of Pentium 4s). They had the 7th-generation microarchitecture, called NetBurst, which was the company's first all-new design since 1995, when the Intel P6 microarchitecture of the Pentium Pro CPUs had been introduced. NetBurst differed from the preceding Intel P6 - of Pentium III, II, etc. - by featuring a very deep instruction pipeline to achieve very high clock speeds (up to 4 GHz) limited only by maximum power consumption (TDP) reaching up to 115 W in 3.6–3.8 GHz Prescotts and Prescotts 2M (a high TDP requires additional cooling that can be noisy or expensive). In 2004, the initial 32-bit x86 instruction set of the Pentium 4 microprocessors was extended by the 64-bit x86-64 set.

The original Pentium 4, codenamed "Willamette", ran at 1.4 and 1.5 GHz and was released in November 2000 on the Socket 423 platform. Notable with the introduction of the Pentium 4 was the 400 MT/s FSB. It was actually based on a 100 MHz clock wave, but the bus was quad-pumped, meaning that the maximum transfer rate was four times that of a normal bus, so it was considered to run at 400 MT/s. The AMD Athlon was running at 266 MT/s (using a double-pumped bus) at that time.

Pentium 4 CPUs introduced the SSE2 and SSE3 instruction sets to accelerate calculations, transactions, media processing, 3D graphics, and games. They also integrated Hyper-threading (HT), a feature to make one physical CPU work as two logical and virtual CPUs. The Intel's flagship Pentium 4 also came in a low-end version branded Celeron (often referred to as Celeron 4), and a high-end derivative, Xeon, intended for multiprocessor servers and workstations. In 2005, the Pentium 4 was complemented by the Pentium D and Pentium Extreme Edition dual-core CPUs.

Architecture

In benchmark evaluations, the advantages of the NetBurst architecture were not clear. With carefully optimized application code, the first P4 did outperform Intel's fastest Pentium III, as expected. But in legacy applications with many branching or x87 floating-point instructions, the P4 would merely match or even fall behind its predecessor. Its main handicap was a shared uni-directional bus. Furthermore, the NetBurst architecture dissipated more heat than any previous Intel or AMD processor.

As a result, the Pentium 4's introduction was met with mixed reviews: Developers disliked the Pentium 4, as it posed a new set of code optimization rules. For example, in mathematical applications AMD's much lower-clocked Athlon easily outperformed the Pentium 4, which would only catch up if software were re-compiled with SSE2 support. Tom Yager of Infoworld magazine called it "the fastest CPU - for programs that fit entirely in cache". Computer-savvy buyers avoided Pentium 4 PCs due to their price-premium and questionable benefit. In terms of product marketing, the Pentium 4's singular emphasis on clock frequency (above all else) made it a marketer's dream. The result of this was that the NetBurst architecture was often referred to as a marchitecture by various computing websites and publications during the life of the Pentium 4.

The two classical metrics of CPU performance are IPC (instructions per cycle) and clock-frequency. While IPC is difficult to quantify (due to dependence on the benchmark application's instruction mix), clock-frequency is a simple measurement yielding a single absolute number. Unsophisticated buyers would simply associate the highest clock-rating with the best product, and the Pentium 4 was the undisputed megahertz champion. As AMD was unable to compete by these rules, it countered Intel's marketing advantage with the 'megahertz myth campaign.' AMD product marketing used a "PR-rating" system, which assigned a merit value based on relative-performance to a baseline machine.

See also: Megahertz myth

A Pentium 4, 2.4 GHz

At the launch of the P4, Intel stated NetBurst was expected to scale to 10 GHz (over several fabrication process generations). However, the NetBurst architecture ultimately hit a frequency ceiling far below expectation—the fastest retail Pentium 4 never exceeded 4 GHz. Intel had not anticipated a rapid upward scaling of transistor power leakage that began to occur as the chip reached the 90 nm process node and smaller. This new power leakage phenomenon, along with the standard thermal output, created cooling and clock scaling problems as clock speeds increased. Reacting to these unexpected obstacles, Intel attempted several core redesigns ("Prescott" most notably) and explored new manufacturing technologies. Nothing solved their problems though and in 2005–06 Intel shifted development away from NetBurst to focus on the cooler-running Pentium M architecture. In March 2006, Intel announced the Intel Core microarchitecture, which puts greater emphasis on energy efficiency and performance per clock. The final NetBurst-derived products were released in 2006, with all subsequent product families switching exclusively to the Intel Core microarchitecture.

Processor cores

The Pentium 4 has an IHS (Integrated Heat Spreader) that prevents the die from accidentally getting damaged when mounting and unmounting cooling solutions. Prior to the IHS, a CPU shim was sometimes used by people worried about damaging the core. Overclockers sometimes removed the IHS on Socket 478 chips to allow for more direct heat transfer. However, on LGA775 chips the IHS is directly attached to the die with epoxy, meaning that the IHS cannot be removed easily.

Intel Pentium 4 processor family
Original Logo	New Logo	Desktop			Laptop
		Code-named	Core	Date released	Code-named	Core	Date released
Original Pentium 4 logo	New logo as of 2006	Willamette Northwood Prescott	(180nm) (130nm) (90nm)	Nov 2000 Jan 2002 Mar 2004	Northwood	(130nm)	Jun 2003
Pentium 4 logo, M-Variation (P4M)					Northwood Pentium 4-M	(130nm)	Apr 2002
Hyper-threading (HT)
Original logo for the Intel Pentium 4 with Hyper Threading Technology	New Intel Pentium 4 with Hyper Threading logo	Northwood Prescott Prescott 2M Cedar Mill	(130nm) (90nm) (90nm) (65nm)	May 2003 Feb 2004 Feb 2005 Jan 2006	Northwood Prescott	(130nm) (90nm)	Sep 2003 Jun 2004
Pentium 4 HT Extreme Edition logo	New logo as of 2006	Gallatin XE Prescott 2M XE	(130nm) (90nm)	Sep 2003 Feb 2005
List of Intel Pentium 4 microprocessors

Willamette

Pentium 4 Willamette for socket 423.

A 'Willamette' core Pentium 4 processor

Willamette, project code name for the first Pentium 4 architecture implementation, experienced long delays in completion of its design process. The project was started in 1998, when Intel saw the Pentium II as their permanent line. At that time, the Willamette core was expected to operate at frequencies of around 1 GHz, maximum. However, Willamette release delays saw the introduction of the Pentium III prior to its completion. Since the radical differences in these architectures meant Intel could not market Willamette as a Pentium III, it was named Pentium 4.

In November 2000, Intel released the Willamette-based Pentium 4 at speeds of 1.4 and 1.5 GHz. Most industry experts regarded the initial release as a stopgap product, introduced before it was truly ready. According to these experts, the Pentium 4 was released because the competing Thunderbird-based AMD Athlon was outperforming the aging Pentium III, and further improvements to the P-III were not yet possible. This Pentium 4 was produced using a 0.18 micrometer (180 nm) process and initially used Socket 423, with later revisions moving to Socket 478. These variants were identified by the Intel product codes 80528 and 80531 respectively.

On the test bench, the Willamette was somewhat disappointing to analysts in that not only was it unable to outperform the Athlon and the highest-clocked Pentium IIIs in all testing situations, it was not clearly superior to even the budget segment's AMD Duron. Although introduced at a price of US$819 (in 1000 unit quantities), it sold at a modest but respectable rate, handicapped somewhat by the requirement of relatively expensive Rambus Dynamic RAM (RDRAM). The Pentium III remained Intel's top selling chip, with the Athlon also selling slightly better than the Pentium 4.

In January 2001, a still slower 1.3 GHz model was added to the range, but over the next twelve months, Intel gradually started reducing AMD's leadership in performance. April 2001 brought the 1.7 GHz P4, the first one to provide performance clearly superior to the old Pentium III. July saw 1.6 and 1.8 GHz models and in August 2001, Intel released 1.9 and 2.0 GHz Pentium 4s. In the same month, they released the 845 chipset that supported much cheaper PC133 SDRAM instead of RDRAM. While SDRAM was much slower than RDRAM and severely hampered the bandwidth-hungry Pentium 4, the fact that it was so much cheaper caused the Pentium 4's sales to grow considerably. The new chipset allowed the P4 to displace the Pentium III virtually overnight, becoming the top-selling processor on the market.

The Willamette code name is derived from the Willamette Valley region of Oregon, where a large number of Intel manufacturing facilities are located.

Northwood

A 'Northwood' core Pentium 4 processor (P4A)

In October 2001, the Athlon XP regained a clear lead for AMD, but in January 2002, Intel released Pentium 4s with their new Northwood core at 1.6, 1.8, 2.0 and 2.2 GHz. Northwood (product code 80532) combined an increase in the secondary cache size from 256 KB to 512 KB (increasing the transistor count to 55 million, up from 42 million) with a transition to a new 130 nm (0.13 micrometer) fabrication process. By making the chip out of smaller transistors, chips can run at higher clocks or at the same speed while producing less heat.

A 2.4 GHz P4 was released in April 2002, and the bus speed increased from 400 MT/s to 533 MT/s for a 2.26 GHz, 2.4 GHz, and 2.53 GHz part in May, 2.66 GHz and 2.8 GHz parts in August, and a 3.06 GHz Pentium 4 arrived in November. With Northwood, the Pentium 4 came of age. The battle for performance leadership remained competitive (as AMD introduced faster versions of the Athlon XP) but most observers agreed that the fastest Northwood P4 was usually ahead of its rival. This was particularly so in the summer of 2002, when AMD's changeover to a 130 nm production process did not help the "Barton" and "Thoroughbred" Athlon XP CPUs clock high enough to overcome the advantage of P4s in the 2.4 to 2.8 GHz range.

The 3.06 GHz processor acquired Hyper-Threading technology that first appeared in Xeon, enabling multiple threads to be run together by duplicating some parts of the processor in order to let the operating system believe that there are two logical processors.

In April 2003, Intel launched new 800 MT/s FSB variants, ranging from 2.4 to 3.0 GHz. This was meant to help the Pentium 4 better compete with AMD's Opteron line of processors. However, when Opteron was launched, due to its server-oriented positioning motherboard manufacturers didn't initially build motherboards with AGP controllers. Because AGP was the primary graphics expansion port at the time, this missing feature prevented the Opteron from encroaching on the Pentium 4's market segment. AMD did boost the Athlon XP's bus speed from 333 MT/s to 400 MT/s, but it wasn't enough to hold off the new 3.0 GHz P4. A 3.2 GHz Pentium 4 Northwood variant was launched in June and the final 3.4 GHz version arrived in early 2004.

Overclocking early stepping Northwood cores yielded a startling phenomenon. When core voltage (Vcore) was increased past 1.7 V, the processor would slowly become more unstable over time, before dying and becoming totally unusable. This became known as Sudden Northwood Death Syndrome, which is caused by electromigration.

Mobile Pentium 4

The Mobile Intel Pentium 4 Processor [1] was released to address the problem of putting a full Pentium 4 desktop chip into a laptop, which some manufacturers were doing. The Mobile P4 still used 70 W of power, which let it bridge the gap between the full Pentium 4 (using about 82 W), and the Mobile Pentium 4 M (using about 35 W).

Mobile Pentium 4 M

Also based on the Northwood core, the Mobile Intel Pentium 4 Processor - M [2] was released on April 23, 2002 and included Intel's SpeedStep and Deeper Sleep technologies, and Hyper-Threading in some models. Intel's naming conventions made it difficult at the time of the processor's release to identify the processor model.There was the Pentium III mobile chip (or the PIII-M), the Mobile Pentium 4 M (or the P4-M), the Mobile Pentium 4 (or the Mobile P4), and then just the Pentium M which itself was based on the Pentium III. Its TDP is about 35 Watts in most applications. This lowered power consumption was due to lowered core voltage, and other features mentioned previously.

Gallatin (Extreme Edition)

In September 2003, at the Intel Developer Forum, the Pentium 4 Extreme Edition (P4EE) was announced, just over a week before the launch of Athlon 64, and Athlon 64 FX (AMD64 FX). The design was mostly identical to Pentium 4 (to the extent that it would run in the same motherboards), but differed by an added 2 MB of Level 3 cache. It shared the same Gallatin core as the Xeon MP, though in a Socket 478 form factor (as opposed to Socket 603 for the Xeon MP) and with an 800 MT/s bus, twice as fast as that of the Xeon MP. An LGA775 version is also available.

While Intel maintained that the Extreme Edition was aimed at gamers, some viewed it as an attempt to steal the Athlon 64's launch thunder, nicknaming it the "Emergency Edition". With a price tag of ~$1000, it was also referred to as the "Expensive Edition" or "Extremely Expensive". Many condemned Intel for cannibalizing the Xeon line, but no such complaints were aimed at AMD's Athlon 64 FX-51, which was merely a repackaged Opteron 148.

The effect of the added cache was somewhat variable. In office applications, the Extreme Edition was generally a bit slower than the Northwood, owing to higher latency added by the L3 cache. Some games benefited from the added cache, particularly those based on the Quake III and Unreal engines. However, the area which improved the most was multimedia encoding, which was not only faster than the Pentium 4, but also faster than both Athlon 64s.

A slight performance increase was achieved in late 2004 by increasing the bus speed from 800 MT/s to 1066 MT/s. Only one Gallatin-based chip at 3.46 GHz was released before the Extreme Edition was migrated to the Prescott core. The new 3.73 GHz Extreme Edition had the same features as a 6x0-sequence Prescott 2M, but with a 1066 MT/s bus. In practice however, the 3.73 GHz Extreme Edition almost always proved to be slower than the 3.46 GHz version.

The 'Pentium 4 Extreme Edition' should not be confused with a similarly-named later model, the 'Pentium Extreme Edition', which is based on the dual-core Pentium D.

Prescott

Pentium 4 LGA775

On February 1, 2004, Intel introduced a new core codenamed "Prescott". The core used a 90 nm process for the first time, and "[it] is also a major reworking of the Pentium 4's microarchitecture—major enough that some analysts are surprised Intel didn't opt to call this processor the Pentium 5". Despite this overhaul, the performance difference was debatable. Some programs benefitted from Prescott's doubled cache and SSE3 instructions, whereas others were more crippled by its long, inefficient pipeline. The Prescott architecture allowed it to be set at slightly higher clock-rates, but not nearly as high as Intel had expected. (See Overclocking.) The fastest mass-produced Prescott-based processor was clocked at 3.8GHz.

Upon release, many reviewers mistakenly concluded that the Prescott generated approximately 40% more heat clock-for-clock than the Northwood, and almost every review of it was negative, earning it the sobriquet PresHot. In reality, the core temperature sensor of the Prescott gives higher readings than the Northwood core temperature sensor, meaning that the increase in heat generated for CPU work done is believed to be around the 10% range. Overclockers mistakenly believed that the Northwood was a better choice for overclocking, while in reality the Prescott would outperform the Northwood and generate less heat in overclocked conditions. At stock speeds (or less ambitious overclocks) however, Prescott was generally still hotter and slower than Northwood. Retailers like Newegg.com were selling Northwoods at slightly higher prices than equally-clocked Prescott counterparts because the Northwoods were in higher demand. A shift in socket type (from Socket 478 to LGA775) was expected to reduce the heat to more acceptable levels, but in fact proved to have the opposite effect, with power requirements increasing by a further 10%. However, the LGA775 reference cooler and mounting system were somewhat better designs, so average temperatures were slightly lowered. Subsequent revisions to the processor by Intel engineers were expected to reduce average temperatures, but this never happened outside of the lowest speed grades. Prescott Pentium 4s were given the product codes 80546 (Socket 478) and 80547 (LGA775).

Finally, the thermal problems were so severe, Intel decided to abandon the Prescott architecture altogether, and attempts to roll out a 4 GHz part were abandoned, as a waste of internal resources. Intel realized that it would be wiser to head towards a "wider" CPU architecture with a lower clock speed to keep heat levels down while still increasing the throughput of the CPU. Also of concern was the fact that a review showed that in games, it took a 5.2 GHz Prescott core to soundly beat the performance of a 64-bit Athlon FX-55 that clocked at 2.6 GHz.[3] Considering Intel boasted at launch the Pentium 4 architecture was intended to support up to 10 GHz operation with further reductions of core size, this can be seen as one of the most significant, certainly most public, engineering shortfalls in Intel’s history. Overclockers did not break the 8 GHz barrier until the end of the Pentium 4 line on 3.0-3.6 GHz CPUs, which by then had a dwindling enthusiast user base. This also meant that while Northwood ultimately achieved clockspeeds 70% higher than Willamette did, Prescott only managed a 12% rise over Northwood.

The Pentium M instead became the internal reference layout for Intel’s design teams, and P4 development was essentially abandoned. To this extent, the little-funded Israeli design team that produced the Pentium M core took over the much larger desktop development project.

Originally, two Prescott lines were released: the E-series, with an 800 MT/s FSB and Hyper-Threading support, and the low-end A-series, with a 533 MT/s FSB and Hyper-Threading disabled. Initially there were big problems with people who installed Windows XP Service Pack 2 on systems with these processors as an incompatibility with the BIOS, processor and SP2 coding led to systems unable to boot. Microsoft and Intel worked on a solution; users with this problem can find out how to install SP2 on a Prescott machine.

LGA775 Prescotts use a rating system, labeling them as the 5xx series (Celeron Ds are the 3xx series, while Pentium Ms are the 7xx series). The LGA775 version of the E-series uses model numbers 5x0 (520-560), and the LGA775 version of the A-series uses model numbers 5x5 and 5x9 (505-519). The fastest, the 570J and 571, is clocked at 3.8 GHz. Plans for 4 GHz processors were axed by Intel in favor of dual core processors, although some European retailers claim to be selling a Pentium 4 580, clocked at 4 GHz.

The 5x0J series (and its low-end equivalent, the 5x5J and 5x9J series) introduced the XD Bit (eXecute Disable) or Execute Disabled Bit [4] to Intel's line of processors. This technology, first introduced to the x86 line by AMD and called NX (No eXecute), can help prevent certain types of malicious code from exploiting a buffer overflow to get executed.

Intel also released a series of Prescotts supporting Intel 64, Intel's implementation of the x86-64 64-bit extensions to the x86 architecture. These were originally released as the F-series, and only sold to OEMs, but they were later renamed to the 5x1 series and sold to the general public. Two low-end Intel64-enabled Prescotts, based on the 5x5/5x9 series, were also released with model numbers 506 and 516.

5x0, 5x0J, and 5x1 series Prescotts have incorporated Hyper-Threading in order to speed up some processes that use multithreaded software, such as video editing. The 5x1 series also supports 64 bit computing.

Prescott 2M (Extreme Edition)

Intel, by the first quarter of 2005, released a new Prescott core with 6x0 numbering, codenamed "Prescott 2M". Prescott 2M is also sometimes known by the name of its Xeon derivative, "Irwindale". It features Intel 64, the XD Bit, EIST (Enhanced Intel SpeedStep Technology), Tm2 (for processors at 3.6GHz and above), and 2 MB of L2 cache. However, any advantage introduced by the added cache is mostly negated due to higher cache latency, and the double word size if using Intel 64 mode. Rather than being a targeted speed boost the double size cache is intended to provide the same space and hence performance for 64-bit mode operations.

6xx series Prescott 2Ms have incorporated Hyper-Threading in order to speed up some processes that use multithreaded software, such as video editing.

On 14 November 2005, Intel released Prescott 2M processors with VT (Virtualization Technology, codenamed "Vanderpool") enabled. Intel only released two models of this Prescott 2M category: 662 and 672, running at 3.6 and 3.8 GHz, respectively.

Cedar Mill

The final revision of the Pentium 4 was Cedar Mill, released in early 2006. This was simply a straight shrink of the 600-series core to 65 nm, with no real feature additions. Cedar Mill had a lower heat output than Prescott, with a TDP of 86 W. The Core Stepping of D0 in late 2006 reduced this to 65 watts. It has a 65 nm core and features a 31-stage pipeline (just like Prescott), 800 MT/s FSB, Intel 64, Hyper-Threading and Virtualization Technology. As with Prescott 2M, Cedar Mill also has 2 MB of L2 cache. It was released as Pentium 6x1 and 6x3 (product code 80552) at frequencies from 3.0 GHz up to 3.6 GHz. Overclockers managed to exceed 8 GHz with these processors. None of the 6x1 range (631, 641, 651, and 661) has Virtualization Technology support. As of March 2007 it has not been possible to obtain 6x3 nor have Intel any records of this product line on their homepage.

To distinguish Cedar Mill CPUs from Prescott CPUs with the same features, Intel added 1 to their model numbers. Thus, Pentium 4 631, 641, 651 and 661 are 65 nm microprocessors, while Pentium 630, 640, 650 and 660 respectively are their 90 nm equivalents.

Successor

Main article: Intel Core 2

The original successor to the Pentium 4 was Tejas, which was scheduled for an early-mid-2005 release. However, it was cancelled a few months after the release of Prescott due to extremely high power consumption (a 2.8 GHz Tejas consumed 150 W of power, compared to around 80 W for a Northwood of the same speed, and 100 W for a comparably clocked Prescott) and development on the NetBurst architecture as a whole ceased, with the exception of the dual-core Pentium D/Extreme Edition and Cedar Mill.

Since May 2005, Intel has released dual-core processors based on the Pentium 4 under the names Pentium D and Pentium Extreme Edition. They represent Intel's shift towards parallelism and their intent is to eventually make the bulk of their main processor line dual-core. These came under the code names Smithfield and Presler for the 90 nm and 65 nm parts respectively.

In 2006, Intel had plans to work further on the Pentium 4 Cedar Mill architecture to develop a 9 GHz Pentium 4. These plans were eventually scrapped in favour of developing the Intel Core microarchitecture.

The ultimate successors to Pentium 4 are the Intel Core 2 processors using the "Conroe" core based upon the Intel Core microarchitecture, released on July 27, 2006. Intel Core 2 processors have, so far, only been released as dual and quad core processors. Single core counterparts are present in the Intel Core line, primarily for the OEM market.

See also

• Intel Core 2
• List of Intel Pentium 4 microprocessors

External links

• The future of Prescott: when Moore gives you lemons... at Ars Technica
• Prescott vs. Northwood Pentium 4 Review
• Intel Documentation
• Inside Pentium 4 Architecture