Pentium anniversary: Overclockable Pentium Anniversary Edition Review: The Intel Pentium G3258

96 Comments

Overclockable Pentium Anniversary Edition Review: The Intel Pentium G3258CPU BenchmarksGaming and Synthetics on Processor GraphicsDiscrete GPU GamingPentium-AE Is A Processor We Want, But Not The Processor We Need

Many industries, both inside and outside of technology, are versed in the terminology ‘cheap and cheerful’. When enthusiasts were overclocking their CPUs at the turn of the century, this was the case – taking a low cost part, such as the Celeron 300A, and adjusting one or two settings to make it run as fast as a Pentium III 450 MHz. This gave a +50% frequency boost at the lower price point, as long as one could manage the heat output. The Pentium Anniversary Edition is a small nod back to those days, and to celebrate the 20+ years of Pentium branding, Intel is now releasing a $75 overclockable dual core Haswell-derived CPU.

The Pentium G3258, or Pentium-AE / Pentium-K

Since the initial announcement from Intel regarding the release of its newest low-cost overclocking processor since Clarkdale (2009/2010), a variety of names have been suggested. Here at AnandTech, I hypothesized that Intel would continue the K naming scheme for overclocking processors, and call this new part the Pentium G3420-K, or Pentium-K for short. However, to tie in with the Anniversary Edition theme, I have since heard from two separate Intel employees at industry events call the model ‘Pentium-AE’ for short, or the ‘Pentium G3258’ as the official name. In order to remain consistent with the naming, we will use the Pentium-AE or Pentium G3258; however other sources may use other monikers.

Intel’s mainstream product line starts with Celeron processors, with the name indicating dual core parts without hyperthreading, but with 2MB of L3 cache and DDR3-1333 MHz memory support. Pentium parts are similar to Celeron, with 3MB of L3 cache but can come with either DDR3-1333 or DDR3-1600 memory support. i3 processors are next, which feature hyperthreading and 4 MB of L3 cache, then i5 which are quad core, no hyperthreading but 6 MB of L3, then i7 which are quad core with hyperthreading and 8 MB of L3. There are also IGP adjustments through the line:

Intel Haswell Desktop CPU Classifications
	Celeron	Pentium	i3	i5	i7
Cores	2	2	2	4	4
Hyperthreading	No	No	Yes	No	Yes
L3 Cache	2 MB	3 MB	4 MB	6 MB	8 MB
L3 Cache / Core	1 MB	1. 5 MB	2 MB	1.5 MB	2 MB
L3 Cache / Thread	1 MB	1.5 MB	1 MB	1.5 MB	1 MB
Memory Support	DDR3-1333	DDR3-1333 or DDR3-1600	DDR3-1600	DDR3-1600	DDR3-1600
IGP	GT1	GT1	GT2	GT2	GT2
AVX / AES-NI	No	No	Yes	Yes	Yes

The Pentium G3258 falls on the lower end of the Pentium bracket. The frequency is high, at 3.2 GHz and matching the G3420, but it comes with only DDR3-1333 support (signified by the G32xx rather than G34xx naming). While this usually does not matter much for overclockers who will likely overclock the memory as well, it does have an effect due to the binning process. Enthusiasts already know that CPUs with DDR3-1600 support can use memory kits above 2666 MHz, but the use of DDR3-1333 on the Pentium-AE CPU may be limiting. It should, theoretically, mean that Intel has a lot more CPUs from the production line that fit into this category.

The aim of Pentium-AE can be considered two fold. As part of the overclocking community, the ‘cheap and cheerful’ mentality is what got a lot of us started in the first place – can we get top end CPU performance without paying top-end prices? With the CPUs being cheap, they could almost be considered disposable, allowing even ‘casual extreme overclockers’ (as much as that phrasing sounds weird) to try lots of processors and compete in a fun category.

The second part of the equation is aimed at gaming. One of the big reasons for growth in the PC industry of late is down to gaming, and the popularity of titles such as League of Legends or DOTA2, among others. These titles typically do not need the latest and greatest, and with the presence of pre-overclocked gaming systems from system integrators based on the Pentium-AE processor, parents who buy systems for their enthusiastic children might be able to start at these lower price points. The added benefit here is that Intel may end up encouraging these individuals to invest in a higher performance machine as they age and can afford it themselves.

There have been several concerns since the original Pentium-AE announcement however. Aside from the low core count which may restrict frame rates, the low amount of L3 cache and lower-speed DRAM memory support have both been noted as potential bottlenecks. A number of overclockers have since requested an unlocked i3 processor from Intel, perhaps fleshing out the range of ‘K’ processors which are currently limited to i5 and i7. Other users are also requesting AES-NI and AVX support (which would come with an unlocked i3 processor), as the Pentium range does not have support for these technologies used for CPU throughput or encryption.

From the CPU-Z screenshots, the only difference between the G3258 and the G3420 is the name string in the CPU firmware (memory support is not shown here), and moving up to the i3-4330, the AVX/AES support is listed along with hyperthreading.

Pentium-AE and Devil’s Canyon

In our review of the new Devil’s Canyon CPUs, we noted that even though the Pentium-AE processor is launched at the same time, and all three are aimed at the overclocking crowd, that the Pentium-AE is not a Devil’s Canyon processor. Back in that review, we discussed the two changes that Intel had made to Devil’s Canyon over standard Haswell processors – additional decoupling capacitors on the package and upgraded thermal interface material to lower processors. For the first change, a quick look at the rear of the G3420 (which was launched in 2013 with the original Haswell processors) and the G3258 processors shows no difference:

When Intel decided to release the Pentium-AE processor, they had two choices. Either adjust processors coming off the line and turn overclocking ‘on’, or actually make physical changes similar to Devil’s Canyon. If the physical changes were an all-or-nothing policy, then I would have to say that the ‘new’ Pentium-AE is just the ‘old’ CPU with a firmware switch enabled. However, the overclocking performance surprised me a little.

Overclocking Performance

The holy grail for Pentium-AE, as the processor was being announced, was to match the history of popular processors such as the Celeron 300A which gave a +50% overclock, or the Core 2 Duo E2160 which could go 100%+ in the right hands. Reaching anywhere near these percentages would be an impressive feat, given that or the past three generations of Intel processors, users have been achieving only +200 MHz (+5%) to +700 MHz (+25%), depending on how lucky they are with the silicon they purchased. Haswell is still known for having a wide swing in overclocking potential from CPU to CPU, so this is still a potential issue with the Pentium-AE processor.

At Computex, several companies were promoting ‘4.5 GHz’ as a magic number:

Aside from the awkward/inaccurate scaling on the slide shown, the 4.5 GHz number gives a +1.3 GHz over the base frequency, or ~40% increase in clock speed.

For our testing we started at 3.5 GHz and 0.900 volts and continued our normal overclocking procedure. If the setting is stable (POV-Ray benchmark and 5 minutes OCCT CPU load), the multiplier is increased, but if for any reason the system fails, the voltage is adjusted by +0.025V offset. This continues until the load temperature is too high, or the voltage jump is overly significant.

Our results are as follows:

Moving from 3.5 GHz to 4.4 GHz was very easy. For most jumps of +100 MHz, only +0.025 volts was needed. Above 4.5 GHz, the load temperature started to rise more significantly, as well as the system power. At 4.7 GHz, moving to 4.8 GHz was almost impossible – even with +0.150 volts in the CPU, the system would crash at any loading attempt.

This behavior is similar to what we saw with the original Haswell CPUs and also with Devil’s Canyon. As long as the user is not thermally limited, there seems to be a big bump where the amount of voltage needed to increase the frequency by 100 MHz is significantly higher than before. Note we are discussing air/water cooling, rather than sub-zero, which affords different properties.

The 4.7 GHz value is also eerily similar to that which we achieved with both the i5 and the i7 Devil’s Canyon CPUs. Note that the peak load temperatures at 4.7 GHz were ‘only’ 76C, but this is for two cores, whereas we saw 79C on the i5 for four cores.

The system power draw, increasing from stock to 4.7 GHz, gave +38W / 41% rise for a ~47% increase in frequency. The POV-Ray performance was also similar, with a 47.67% increase in performance for a 47% increase in frequency/41% increase in power.

Performance Overview

It can be hard to pinpoint where the typical Pentium user might lie. If we consider a household machine for family use, it is not going to be overclocked. Similarly for simple office use, either for regular Office software or remote connecting, overclocking is not going to be a feature. Overclocking lies at the heart of both the enthusiast and the gamer, which is where we should look first. One could argue that there are HTPC usage points for an overclockable Pentium; however the increase in power draw and temperature for an overclocked processor might not be appropriate.

For our testing, our main comparison point at stock would be the Pentium G3420. This Haswell processor’s only difference is the DDR3-1600 memory support, meaning that we test the G3420 at DDR3-1600 and the G3258 at DDR3-1333. We also compared the overclocked Pentium G3258 with an i3-4330, showing the effect of hyperthreading.

Intel Pentium G3258 Comparison
Benchmark	G3420 vs G3258	G3258 OC vs G3258	i3-4330 vs G3258 OC
3DPM-ST	0%	47%	-25%
3DPM-MT	1%	47%	33%
WinRAR 5. 01	3%	12%	12%
FastStone 4.9	2%	33%	-33%
Xilisoft VC 7.5 Skyfall LQ	0%	25%	23%
Xilisoft VC 7.5 BBB 4K 60	2%	30%	-7%
PovRay 3.7 beta	1%	47%	-17%
HandBrake v0.9.9 Skyfall LQ	1%	41%	-23%
HandBrake v0.9.9 BBB 4K 60	3%	32%	-4%
Agisoft PS v1.0 Stage 1	3%	19%	27%
Agisoft PS v1. 0 Stage 2	9%	16%	3%
Agisoft PS v1.0 Stage 3	4%	27%	-17%
Agisoft PS v1.0 Stage 4	2%	30%	-20%
Agisoft PS v1.0 Mapping	3%	17%	11%
TrueCrypt 0.7.1a	1%	46%	339%
7-Zip MIPS	1%	35%	4%
Average	2%	32%	19% (-2% w/o TrueCrypt)

For pure CPU performance, the biggest defects from not having DDR3-1600 support are in Photoscan, a 2D to 3D modeling conversion tool. The 47% increase in frequency to 4.7 GHz gives more of a boost, ranging anywhere from 12% in our WinRAR test to 47% in anything that was fully multithreaded with no memory limitations (3DPM, POV-Ray), for a 32% average.

The situation between the i3-4330 and the overclocked G3258 is a story of two tales. While the average score gives +19% to the i3, any single threaded benchmark (3DPM-ST, FastStone) gave an advantage to the G3258 OC while most of the multi-threaded benchmarks looked at the i3. There are clear exceptions to this – video conversion and POV-Ray. For the video conversion tests, especially with the smaller frame sizes of the low quality videos, the benefits of the high single thread speed outweighed the benefits of more threads. It should also be worth pointing out the encryption boost of 339% in favor of the i3 – this is due to AES-NI support on the i3. Without this result in the mix, the average result is actually in favor of the G3258 at 4.7 GHz by 2%.

Turning to gaming, and we tested in single and dual GTX 770 formats with six games at 1080p and all the settings turned up.

Gaming Average Frame Rates: GTX 770 at 1080p
Benchmark	G3258	G3420	G3258 at 4.7 GHz	i3-4330
F1 2013	70.9	72.2	86.5	112.7
Bioshock Infinite	87.1	89.6	93.6	91.4
Tomb Raider	49.8	49.0	48.6	49.1
Sleeping Dogs	55.1	58.9	57.2	56.2
Company of Heroes 2	37.7	38.0	42.7	43.2
Battlefield 4	53. 7	54.1	56.3	59.3
Gaming Average Frame Rates: 2x GTX 770 at 1080p
F1 2013	64.6	69.1	83.9	101.1
Bioshock Infinite	126.3	131.0	144.1	145.7
Tomb Raider	95.5	96.3	86.7	96.4
Sleeping Dogs	67.2	67.6	85.5	89.0
Company of Heroes 2	36.0	37.1	42.3	42.6
Battlefield 4	51.3	53.6	65.7	93. 4

Comparing the G3420 to the G3258 at stock speeds, the effect of DDR3-1600 vs. DDR3-1333 is felt most in Sleeping Dogs with one GPU and F1 2013 in dual GPU mode, giving +7% to the G3420.

When the G3258 is overclocked, most of the benefits are given in multi-GPU modes. Here, three of our titles recorded ~30% frame rate increases, with another two around 15%. The biggest beneficiaries of an overclocked Pentium-AE are going to be in the multi-GPU bracket.

However, the crux of the situation shows that even with an overclocked Pentium, an i3-4330 can get a lot more out of your graphics cards. In single GPU, we see a 30% increase in frame rates for F1 2013, moving towards 120 FPS. In dual GPU modes, BF4 was the biggest beneficiary with a +42% increase. Moving from 65.7 FPS to 93.4 FPS for dual GTX-770s is a no brainer.

Overall, the main performance benefit from having an overclocked Pentium is going to be from all the ‘need-now’ types of CPU tasks that rely on response time. Having a 4.7 GHz Pentium feels fast when opening tabs or documents, but the minute any serious load is applied, the overclocked Pentium will feel ‘on average’ like an i3 at stock. The only exception is AES-NI or AVX workloads that will fly on the i3. Putting this into perspective, this lands the overclockable Pentium at the feet of ‘a bit of fun’, rather than anything serious.

Motherboards for Pentium-AE

Overclocking is a targeted market for motherboard manufacturers; especially those that invest time and money into research to let their users get better overclocking performance. They cannot expect users to purchase a $70 CPU then a $250-$400 motherboard. There will be overclocking enthusiasts that will do that, but the larger market of end-users who want some free performance will most likely be purchasing a motherboard similar in price to the CPU, or perhaps reaching into the $110-$120 range at most.

To this end, at least one of the motherboard manufacturers has specifically released new motherboards aimed for Pentium-AE users, and all the others have directed me to models they already produce. ASRock launched its Z97 Anniversary and Z97M Anniversary models at Computex this year:

MSI sent me its Z97 Guard Pro while I was testing the G3258, their main motherboard for Pentium-AE. ASUS call upon its Z97-A, and GIGABYTE has the UD3H and lower models as well.

CPU Benchmarks
Overclockable Pentium Anniversary Edition Review: The Intel Pentium G3258CPU BenchmarksGaming and Synthetics on Processor GraphicsDiscrete GPU GamingPentium-AE Is A Processor We Want, But Not The Processor We Need

Tweet

PRINT THIS ARTICLE

Review: Intel Pentium Anniversary Edition G3258 — CPU

Remember the days when you could take a cheap processor, overclock it to rude levels using basic air cooling, and then achieve the kind of performance normally reserved for far more expensive chips? Those days are largely gone, more’s the pity, because AMD’s base CPU performance isn’t great while Intel limits overclocking on cheaper non-K processors by locking the maximum all-core multiplier and base clock ratio to the chip’s default level. Want to go faster, spend the extra and buy a K-series chip.

In a somewhat surprising move disclosed earlier this month, Intel decided to launch a budget processor with K-class overclocking credentials. Said chip is the Pentium Anniversary Edition, based on the same Haswell architecture as the latest Core series, but now with an unlocked multiplier that paves the way for end-user tinkering.

Launched to coincide with the 20th anniversary of the first Pentium processor, this new budget chip is actually known as the Pentium G3258. Imbued with two cores and able to process a total of two threads, G3258 is nominally clocked in at 3.20GHz and backed with 3MB of cache. It isn’t the fastest Pentium G-series chip on paper; that honour goes to the G3450, which is able to run at a maximum of 3.40GHz. Here’s how select Intel processors line up against one another.

Better than it looks on paper

The dual-core Pentiums sip on less juice than the Core i5-4670K that most enthusiasts will be familiar with. A leaner architecture — both in terms of CPU and baked-in GPU — means that Pentium Gs are rated to just 53W. Most enthusiasts would likely think the G3258 too low rent for a decent PC, but the very fact that it is unlocked, and has the potential to run at way past 4GHz with a bit of novice tweaking, may have them thinking again, especially with regards to gaming.

Too good to pass up?

We already know that the Pentium G3220, priced at £40, offers excellent value for money. Increasing the budget by a tenner, to this model, is sure to provide significantly more performance once it is adequately overclocked. It’s an easy chip to recommend, too, as cheap LGA1150 boards are plentiful, dual-channel memory is widely available, and constructing a base platform hardly breaks the bank.

Putting ourselves in the position of the reader, we’d like to know if default-clocked performance is adequate for a mainstream PC, and whether overclocked performance gives vastly more expensive Core processors a good run for their money. Let’s answer both with a slew of benchmarks.

Intel’s 35th Anniversary: ​​Little Known About the Known

Intel’s 35th Anniversary: ​​Little Known About the Known

According to Webster’s Dictionary, a name is «a word or words by which an object is characterized and distinguished from others.» When the founders of Intel chose a name for their corporation and its products 35 years ago, no one could have imagined how important these names would become over time. Let’s look at today. A recent survey by Business Week magazine showed that the Intel brand is one of the top ten most popular brands in the world. That name alone is currently worth about $31 billion. Where did the name «Intel» come from? How did the «Intel Pentium» brand come about? And what do Intel’s musical callsigns have in common with the Intel Inside program? Thanks to the Intel Museum, you can answer these questions.

Company name
On July 18, 1968, the late Bob Noyce and Gordon Moore incorporated a new business entity, NM Electronics. It didn’t sound very appealing, but at the time Noyce and Moore’s list of priorities didn’t include choosing a company name. Someone even suggested «Moore, Noyce Electronics», which sounded even worse, because it was consonant with «More Noise» («more noise»). In the end, they settled on another option — «Integrated Electronics» («Integrated Electronics»), since it indicated the scope of the future activity of the new commercial enterprise. However, Noyce and Moore failed to approve this or ten other names in the California State Corporations Commission. «Then we decided to take the first syllables of the words «Integrated» and «Electronics» and made them «Intel,» Noyce later recalled. To avoid conflict with other companies with similar names, Intel acquired the rights to use the word «Intel» from Intelco «We thought paying $15,000 would be easier than figuring out a new name,» says Moore.0008 New company begins operations in Mountain View, California. 12 employees were located in one room, and the amount of income for the company for the first year of existence was $2. 672. In 1971, Intel Corporation was transformed into an open joint stock company, and its shares began to be sold on the stock exchange at a price of $23.5 per share. One of its shares, issued in 1971, became equivalent to 38 shares in 1988.

Why is the «e» omitted from the Intel logo?
Dropped «e» logo was adopted shortly after the «Intel» name was approved. Public relations consultant John Hall suggested this design as a way to show that «Intel» is a compound word, consisting of the initial syllables «Integrated» and «Electronics». The idea was accepted with a bang.

Origins of the Intel Inside Program
In 1989, Intel’s marketing manager Dennis Carter formed a small group and for the first time in history began a program to market not an end product but a component, in this case the microprocessor. The target audience for the 386SX processor was information technology managers who purchased business PCs. D. Carter’s initiative was a resounding success. On the advice of its advertising agency, Dahlin, Smith and White, Intel adopted the slogan «Intel. The computer inside.» It was later shortened to «Intel Inside».
In 1991, Carter launched the Intel Inside program, which by then had evolved into a joint advertising program with personal computer manufacturers. For this purpose, a special fund was created at Intel Corporation, which was formed by deducting a certain percentage of the profit from the sale of processors for advertising purposes.

Intel has asked its customers to share the cost of creating and publishing PC ads with the Intel Inside logo. The addition of the Intel Inside logo not only made OEMs more cost effective on advertising, but also served as a kind of guarantee that their systems were based on the latest technology. The implementation of this program began on July 1991
In early 1992, Intel made its television commercial debut, emphasizing the speed, power, and affordability of microprocessors. The commercials were produced at George Lucas’s Industrial Light Magic studio, using state-of-the-art special effects that allowed viewers to take a walk through the «innards» of a personal computer. In conjunction with its color television ads, Intel used a three-second animated music piece (known as its branded AV screensaver) that displayed the logo and played a five-note melody.

How the Intel Pentium microprocessor got its name
In the late sixties, P-channel MOSFETs were assigned numbers that looked like 1xxx, and bipolar products — 3xxx. The second digit in these numbers was 1 (denoting random access memory), and the last two digits were the serial number of the product. Thus, the first bipolar RAM was called 3101. The numbering system was quite simple, and numbers were added as new products were developed. Marketers also had a hand in this, deciding to add appeal to these numbers by adding fours and eights.
Soon, however, Intel had problems with this naming system. In principle, Intel Corporation had every right to these numbers, since it was it that introduced such a naming system, but it did not formalize its rights to it, i.e. did not prevent other companies from using these names to refer to their own processors. The rights of Intel Corporation remained unrealized — after all, now these numbers no longer denoted the trademark of the Intel processor, but the class of the processor. A definition was issued that allowed Intel Corporation to use integers as a name on a general basis. This, however, did not extend to 286, 386 and 486, which were already widely accepted and used as «household marks» by the public and press.
In light of all this, a competition was announced among Intel employees for the best name for the new processor. As a reward, the winner was promised a trip for two to Hawaii. Many proposals were received, in connection with which it was decided to hire two agencies: NameLab and Lexicon, who were instructed to select the best applications. None of them had the «Pentium» name in them, but two Intel employees were rewarded with trips to Hawaii because their options eventually helped professional marketers develop a proper name to use on March 19.93, Intel introduced the Pentium processor. It became the first microprocessor in history whose name was not associated with a digital circuit. To this day, many people ask what this name means. The answer is it doesn’t mean anything. Nevertheless, the Intel Pentium processor has gained enormous, truly worldwide popularity, and sometimes this name is used as a household name for personal computers with a strong reputation for being the most powerful and reliable computing devices.

Snakes, kitchen vigils, and handwritten reports
During the 1970s, snakes sometimes crawled across the floor of Intel’s manufacturing facility in Penang, Malaysia, where the Intel 1103 memory chips were assembled, and Intel employees in Israel periodically broke away from the design microchips to wash dishes in the common dining room kitchen. The word is for those who have been working at Intel Corporation since those times.
Arik Shemer, 8087 Test Engineer, Intel Israel Design Center: «At 1978 I was looking for a job. Wandering around the Haifa tech park, I noticed the blue Intel logo hanging on the door of an old warehouse. Hmm, I thought, Intel… Didn’t I read about it in my last year of study as an engineer in an article about a crazy company from the USA that introduced a computer on four microcircuits? I knocked on the door, went in and asked if they needed new shots. I was interviewed immediately, followed by a second interview a week later, and before other companies even read my application, I got a job at Intel. The microchip simulation was done on the only minicomputer we had in the building. To get the opportunity to work on it, I had to pre-register in the queue. No more than 20 minutes were allotted for each modeling task. If you fail to meet the allotted time, other users start to get angry. It got to the point where some of them got impatient and inserted their punched cards into the computer to reset it, and as a result, all the work you had done was erased!»
Karen Oh Swee Keen, Intel Malaysia: “Back in the 70s, we were assembling 1103 memory chips. «Most of us started our careers as solderers. Operators of semi-automatic soldering machines had to point a navigation light at the contact pad of the crystal. At the push of a button, a soldering tube would come out and make a wired connection at the indicated place. The eyes hurt very much from such work.»
Maxine Fassberg, director of Intel Israel Fab 18, joined Intel in 1983 and became one of Intel’s first female plant managers. Fassberg owns the patent for the polyamide process. She recalls the early days: “We wrote weekly reports with pen and paper, after which I made copies of them and handed them over to my boss. One day they wrote a huge book — “Statistical Process Control” — and everything was by hand. turned in to the printers. Today, these reporting methods seem antediluvian.»

Alex Kornhauser, Vice President of Intel Israel, was an employee of Intel Israel #40 and initially worked in a warehouse. When Alex was not busy developing chips, he was on duty in the kitchen. Today, Kornhauser is vice president of the Intel Technology and Manufacturing Group, corporate head of Israel, and general manager of the Fab 18 plant. equipment, including an HP calculator that I still keep.0008 This was the personal computer of the time… At the design center were two computers, each capable of executing 1 million instructions per second. Today, the processing power of a computer that is on the desks of our employees is thousands of times greater … A typical project involved five to seven engineers and the same number of template developers. The design was not automated, so a lot had to be done by hand.»
Lisette Uzan, Intel Israel: «I started working at Intel’s Haifa office on October 1977 years as employee number 35. My first project was the 8080b processor, which had about 5,000 transistors. And today we produce microcircuits with 50 million transistors and more!» : business

Intel Celebrates 30th Anniversary of Microprocessor

Digitization
Business applications

14 Nov 2001 12:11 14 Nov 2001 12:11

|

Share

In 1968, moviegoers were amazed and fascinated by a device called the HAL, a powerful computer featured in the science fiction film 2001: A Space Odyssey. But the creators of the film that has become a classic could not imagine the role that personal computers and the Internet play today. Shortly after the release of Space Odyssey, Intel engineer Ted Hoff developed a general-purpose logic device for the Japanese company Busicom, a manufacturer of calculators — this was the Intel 4044 microprocessor with a clock speed of 108 kHz, whose appearance in 1971 marked the beginning of the development of the PC and became the harbinger of a new way of life for people around the world.

Busicom originally acquired the rights to the microprocessor by paying Intel $60,000. Realizing the enormous potential that the «intelligent» chip contained, Intel decided to return this money to Busicom in order to regain the rights to the microprocessor. Busicom agreed, and on November 15, 1971, Intel introduced the 4004 to the world. It sold for $200 apiece. The 4004 was the first to implement the idea of ​​a programmable device, pre-existing chips only performing certain «fixed» functions.

Consisting of 2,300 transistors in an area less than a fingernail, the Intel 4004 microprocessor had the same processing power as the first electronic computer ENIAC, which was created in 1946 and occupied an entire room. The Intel 4004 microprocessor found its first practical use in systems such as traffic light controllers and blood analyzers. It was also used in NASA’s Pioneer 10 probe, which was launched by NASA 29 years ago and, by September 1, 2001, was 11.78 billion km away from Earth.

Much has changed since the introduction of the 4004 microprocessor. Today, more than 40 varieties of microprocessors work in the home of the average American family. (If we also take into account the computer and personal belongings, then their number will increase to 50). Variations of this technological marvel are found in household bathroom scales with digital displays, in irons with circuit breakers, in increasingly common electronic toothbrushes with 3,000 lines of code, and, of course, in home PCs.

The processing power of today’s microprocessors has also increased markedly. The modern Intel Pentium 4 processor performs two billion cycles per second and outperforms the 4004 microprocessor by about 18,000 times. The growth of processor performance stimulates the development of new trends in the use of computer technology in the home, including digital video, sound, photography, information exchange, three-dimensional graphics, and games.

In 1965, Gordon Moore, one of the founders of Intel, predicted that the number of transistors on a chip would double every one and a half to two years. This prediction of his was called Moore’s Law, which continues to this day: to move from the speed of 108,000 periods per second, which the 4004 «developed», to 1 billion periods per second (1 gigahertz) — the clock frequency of the Pentium processor III, it took Intel 28 years. On the other hand, it took only a year and a half to go from 1 GHz to 2 GHz (this barrier was overcome in August 2001, when the corresponding Pentium 4 model was introduced). The most recent offspring of the 4004-derived genus is the Intel Pentium 4 desktop processor running at 2 GHz, about 18,000 times the speed of the 4004 microprocessor.

Alexander Golubchikov, MegaFon: Cloud solutions and cybersharing are gaining popularity to expand the SecaaS concept

Safety

The latest achievements of Intel scientists and engineers related to miniaturization of transistors and the process of manufacturing chips give reason to say that by 2007 there will be a new generation of universal Intel microprocessors, the speed of which will be 20 GHz. Computing power at this new stage can be, in particular, directed to full-fledged speech recognition and voice control of home PCs, which, as experts suggest, will make it possible to talk to your computers and receive an immediate response from them.

---

Intel is a leading provider of the building blocks of the Internet economy. As a manufacturer of microprocessors, Intel is also a manufacturer of personal computer hardware, networking and communications products.