Intel is set to retire its long-used Pentium and Celeron CPU brands and has instead opted for a new, unified naming structure for Intel processors.
Starting in 2023, Intel notebooks will be powered by an Intel processor that spans multiple processor technology families.
“Intel is committed to driving innovation for the benefit of users, and our entry-level processor families have been instrumental in raising the standard of PCs at all price points,” said Josh Newman, the company’s interim general manager of Mobile Client Platforms.
“The new Intel processor branding will simplify our offering so users can focus on selecting the right processor for their needs.”
An Intel Pentium chip was first used in a consumer PC in 1993, and many generations have come since then. The Celeron chips were introduced in 1998.
Intel is looking to simplify the entire naming structure, with “Intel Processor” set to join the flagship Intel Core, Intel Evo, and Intel vPro brands.
The idea behind this change is to make it easier for consumers to buy products. However, mostly only the branding will change, since the product lines that the chipmaker plans to serve remain largely the same.
With the chip bottlenecks in the Corona crisis, months of waiting for washing machines and cars became just as normal as the futile search for new Playstation consoles. How long will this go on?
Source: dpa
Some experts are cautiously optimistic: Industry analyst Alan Priestley from the IT market researcher Gartner can imagine that overcapacities will occur again in the coming year. At the chip group Intel, however, company boss Pat Gelsinger warns that problems with production capacity and the availability of necessary machines are likely to last at least until 2024.
Expert Priestley is counting on the large-scale expansion of production soon bearing fruit. The industry has been investing in new factories since the first signs of bottlenecks at the beginning of the corona pandemic. “Therefore we will probably have overcapacity in 2023 or 2024,” he predicted. At least until the increasing demand eats them up.
This interplay is in itself typical for the chip industry, stressed Priestley. “As soon as demand and supply are in balance, investments are made in order to have capacities for the next surge in demand.” The current chip crisis was so severe because in some areas not enough was invested in expansion – and at the same time demand increased by leaps and bounds with the increased work and learning in the corona pandemic.
Different chips require different production processes
It makes everything even more complex that there is not just one semiconductor industry. Different chips are required for different tasks – and for these there are different production processes that cannot be arbitrarily interchanged. An important feature is the so-called structure widths – roughly speaking, how finely the circuits are worked. The narrower the structure widths in modern production processes, the more powerful and efficient the processors are.
For something like a transformer or other simple tasks, old production processes are sufficient – which are also less lucrative, so that the capacities for them had leveled off at a low level. And then came the Corona boom. It was often the lack of a few simple components that prevented cars or computers from being completed.
In addition, chip factories not only cost billions and have to be built over years – they are also not necessarily flexible. “If I build a factory for chips with a structure width of seven nanometers, I cannot easily convert them to five nanometers,” says Priestley. And it is not as easy as in other areas to simply let production rest. That’s why companies have to plan their expansion decisions with particular foresight – and new capacities in one area are not necessarily helpful for other parts of the industry.
Machines for chip production are still scarce
Another bottleneck are machines for chip production. With super-narrow structure widths, the only way forward is with UV lithography – and the European manufacturer ASML is practically the only supplier for this. He is accordingly fully booked.
Stories of desperate measures are doing the rounds in the industry. At the presentation of the latest quarterly figures, ASML boss Peter Wennink talked about his conversation with the management of a “very large industrial group”, whose name he kept to himself. “They actually told me that they buy washing machines to rip out semiconductors and put them in industrial modules.” Of course, one could say that this is just an anecdote, admitted Wennink. “But that happens everywhere.”
Another problem for the industry came with the war in Ukraine. The lasers of the lithography machines require neon in addition to other gases. And that’s a common by-product of steelmaking. The huge Azovstal steel works in Mariupol, which has been shelled and bombed by Russian troops in their war of aggression against Ukraine for weeks, was an important supplier of neon. The loss of shipments from Ukraine, while disruptive, will be something the industry will have to digest, Priestley said. “There are also other sources” – including the large steel mills in South Korea.
The Russian chips are not yet suitable for everyday use in companies, but there is progress.
The ongoing coronavirus pandemic has shown us how fragile our supply chains have become. Many products are sometimes only manufactured in certain regions of the world, which is why microchips have become so rare in the last 2 years. For this reason, too, Russia would like to buy more products from the domestic market.
Therefore, a small technology sector of its own has developed in the country in the past, in addition to software products, the associated hardware should also be “Made in Russia”. The MCST Elbrus-8C processors show that this is not yet the case with processors.
SberTech, the technology division of Russia’s largest bank, has compared the Elbrus-8C with systems with Intel Xeon Gold 6230 and had to realize that the technology is making great strides, but at the same time it is not yet suitable for everyday tasks within the company.
SberTech is not alone in this. Russian chips have already been installed in many places in the past, but when it comes to the critical infrastructure, one is still dependent on companies such as AMD and Intel. In the test field set up by SberTech, the Elbrus-8C failed in 84 percent of the functional test cases, and the chip could not or only partially withstand other tests.
Nevertheless, all hope is not lost. Experts say that although the chip has many disadvantages, it is still possible to work with it. In addition, adjustments in the software can optimize the current behavior and thus get the maximum out of the processor.
Another positive aspect is that such processors already exist, even if the development of new generations still takes far too long. Years can quickly pass before a concept is converted into a finished chip. During this time, companies such as AMD and Intel have long overtaken the progress that was then achieved. So it remains to be seen how chip development “Made in Russia” will progress in the future.
