Processador Intel Future Roadmaps

[Nemesis11]

Intel Considers Options Such as a Foundry Split​

• Once-dominant chipmaker looks to longtime bankers for help
• Possible moves could include scrapping some factory projects

Intel Corp. is working with investment bankers to help navigate the most difficult period in its 56-year history, according to people familiar with the matter.
The company is discussing various scenarios, including a split of its product-design and manufacturing businesses, as well as which factory projects might potentially be scrapped, said the people, who asked not to be identified because the deliberations are private.

Morgan Stanley and Goldman Sachs Group Inc., Intel’s longtime bankers, have been providing advice on the possibilities, which could also include potential M&A, the people said. The discussions have only grown more urgent since the Santa Clara, California-based company delivered a grim earnings report this month, which sent the shares plunging to their lowest level since 2013.

No major move is imminent and discussions are still in early stages, the people cautioned. A representative for Intel declined to comment, while Morgan Stanley and Goldman Sachs didn’t immediately respond to requests for comment.

The various options are expected to be presented during a board meeting in September, the people said.

A potential separation or sale of Intel’s foundry division, which is aimed at manufacturing chips for outside customers, would be an about-face for Chief Executive Officer Pat Gelsinger. Gelsinger has viewed the business as key to restoring Intel’s standing among chipmakers and had hoped it would eventually compete with the likes of Taiwan Semiconductor Manufacturing Co., which pioneered the foundry industry.

But it’s more likely that Intel takes a less dramatic step before it reaches that point, such as holding off on some of its expansion plans, the people said. The company has already done project financing deals with Brookfield Infrastructure Partners and Apollo Global Management.

Intel’s Gelsinger is running out of time to pull off a much-needed turnaround. He’s been attempting to expand the chipmaker’s factory network at the same time that sales are shrinking — a money-losing proposition. The company suffered a net loss of $1.61 billion last quarter, and analysts are predicting more red ink for the next year.

“Expect big capex cuts from Intel over the next 12 months,” said Amir Anvarzadeh, market strategist at Asymmetric Advisors. “Intel’s model is effectively broken. It’s fighting fires on too many fronts.”

Gelsinger, an Intel veteran who left the company for more than a decade, took the helm in 2021 and promised to restore the company’s technological edge. Under previous CEOs, the chip pioneer had lost market share and its long-vaunted reputation for innovation.

But his comeback plan proved overly ambitious, and the company has had to scale back. When it reported earnings earlier this month, Intel announced plans to cut about 15,000 jobs and slash capital spending. The company even suspended its long-prized dividend.

“It’s been a difficult few weeks,” Gelsinger told investors at the Deutsche Bank Technology Conference on Thursday. The company tried to lay out a “clear view” of its next steps during its earnings report, he said. “Obviously the market didn’t respond positively. We understand that.”

Adding to the upheaval, director Lip-Bu Tan abruptly stepped down from the board last week. The semiconductor veteran, who was brought in two years ago to help with the comeback effort, cited scheduling commitments. But his departure removed one of the few directors with industry knowledge and experience.

Gelsinger’s comeback plan hinged on recasting Intel into two groups: one that designs chips and another that manufactures them. The production arm would then be free to seek business from other companies.

But the biggest client of Intel’s factory network is still Intel. Until the foundry business has more outside customers, it’s going to be challenged financially. It reported operating losses of $2.8 billion in its most recent quarter and is now on course to have a worse year than projected.

With a market value of $86 billion, Intel has fallen out of the top 10 largest chipmakers in the world ranked by that measure. It’s the second-worst performer on the Philadelphia chip index this year and suffers in comparisons with the stratospheric gains of Nvidia Corp., a company that’s on course to post double Intel’s revenue in 2024.

As recently as 2021, Intel was three times the size of Nvidia by revenue.

https://archive.is/FVBsu

Os próximos meses vão ser muito importantes para o futuro da Intel.

 

[Dark Kaeser]

A AMD e IBM já tinham percebido isso.

 

[Nemesis11]

Um artigo que vem contrariar um pouco o artigo anterior.

Exclusive: Intel CEO to pitch board on plans to shed assets, cut costs, source says​

• Intel CEO Pat Gelsinger and other executives to present plans to trim down company
• CEO will present ideas at mid-September meeting,
• Plans could include selling Altera programmable chip unit, sources say
• Capital spending cuts may include German factory expected to cost $32 billion, source says
• Intel has retained Morgan Stanley and Goldman Sachs to advise on asset sales

The plan will include ideas on how to shave overall costs by selling businesses, including its programmable chip unit Altera, that Intel can no longer afford to fund from the company’s once-sizeable profit.

The proposal does not yet include plans to split Intel and sell off its contract manufacturing operation, or foundry, to a buyer such as Taiwan Semiconductor Manufacturing Co., according to the source and another person familiar with the matter.

Intel has already broken off its foundry business from its design business, and has been reporting its financial results separately since the first calendar quarter of this year.

The company has erected a wall between the design and manufacturing businesses to assure that potential customers of the design division would have no access to technology secrets of customers using Intel’s factories, known as fabs, to manufacture their chips.

The proposal Gelsinger and others will present is likely to include plans to further reduce the company’s capital spending on factory expansion. The pitch may include plans to pause or altogether halt its $32 billion factory in Germany, a project that has reportedly been delayed, the source said.

In August, Intel said it expects to cut capital spending to $21.5 billion in 2025, down 17% from this year, and issued a weaker-than-expected third-quarter forecast.

One potential unit the company may look to unload is its programmable chip business, Altera, which Intel acquired for $16.7 billion in 2015. Intel has already taken steps to spin it out as a separate but still wholly owned subsidiary and has said it planned to sell a portion of its stake in an initial public offering in the future, though it has not set a date.

But Altera could also be sold entirely to another chipmaker interested in growing its portfolio, and the company has quietly begun exploring whether a sale would be possible, according to one source familiar with its advisory plans and one of the sources familiar with the plans to cut businesses.

Infrastructure chipmaker Marvell is one potential buyer for such a transaction, according to one of the sources.

https://www.reuters.com/technology/...shed-assets-cut-costs-source-says-2024-09-01/

Já agora, duvido que governos e reguladores aceitassem a venda do Foundry business à TSMC ou mesmo à Samsung.

 

[Dark Kaeser]

Parece confirmar-se mais um chip mobile "Panther Lake"

https://twitter.com/i/web/status/1788648028909298117

aparentemente haverá também uma variante PTL-U.

EDIT:

Intel Panther Lake to feature up 16 CPU cores and 12 “Celestial” Xe3-Cores​

https://videocardz.com/newz/intel-panther-lake-to-feature-up-16-cpu-cores-and-12-celestial-xe3-cores

https://twitter.com/i/web/status/1830222240681431356

 

[Nemesis11]

Hoje foi lançado o Lunar Lane. Um CPU para o mercado mobile.
https://www.servethehome.com/intel-core-ultra-200v-series-lunar-lake-launched/

Porque é que ele tem 9 SKUs? Em 4 dos SKUs, a diferença é só suportarem 16 GB de RAM (Porque é que em 2024 se lançam SKU limitados a 16 GB de RAM???) e o de topo, com mais 100 Mhz no CPU e 50 Mhz no iGPU, tem quase o dobro do TDP em default. Deve ser o SKU para ficar bem nas reviews.
Mesmo os restantes SKUs, as diferenças são pequenas.

Este processador podia ter 2 SKUs. Um com 12 MB L3, 8 Xe Cores e 17 W de TDP, configurável para 8 - 30 W e outro SKU com 8 MB L3 e 7 Xe Cores e configurável entre 8 - 17 W. Os dois com suporte de 32 GB de RAM, que por si já é um limite baixo.

 

[lmikotas]

A questão da RAM prende-se por a mesma ser on-board, vem mesmo integrada no chip (como a Apple faz), daí haver SKU diferentes.

Portanto, se a OEM quiser oferecer um PC com 16GB de RAM tem de escolher essa SKU, se quiser um com 32GB de RAM escolhe a outra.

The biggest change? If you buy a Lunar Lake laptop, it won’t have separate memory sticks or chips! Lunar Lake now bakes 16 or 32GB of LPDDR5X memory into the package itself, with no ability to connect more RAM. It’s a change that reduces the power consumption of moving data through the system by approximately 40 percent, according to Intel. For those who need more memory, Hallock says a separate Arrow Lake architecture is coming to laptops later this year.

https://www.theverge.com/2024/6/3/24169115/intel-lunar-lake-architecture-platform-feature-reveal

 

[Nemesis11]

A questão da RAM prende-se por a mesma ser on-board, vem mesmo integrada no chip (como a Apple faz), daí haver SKU diferentes.

Portanto, se a OEM quiser oferecer um PC com 16GB de RAM tem de escolher essa SKU, se quiser um com 32GB de RAM escolhe a outra.


https://www.theverge.com/2024/6/3/24169115/intel-lunar-lake-architecture-platform-feature-reveal

Obrigado pela correcção. Não reparei que a RAM LPDDDR5X está no package do SOC. Sendo assim, podiam ter reduzido o SKUs a 4. A diferença dos Ultra 5 e Ultra 7 é praticamente 200 Mhz de Turbo.

Tendo em conta que a RAM está no SOC, fui ver qual é o Bus Width total da RAM, porque tinha alguma esperança de ser acima de 128 bits.
O máximo é mesmo os "comuns" 128 bits e além disso, reparei que há outra diferença muito substancial entre as versões com 16 e 32 GB de RAM.

Se repararem no fundo da imagem, onde está a tabela, a versão 32 GB tem 128 bits de largura de Bus (2x64), mas a versão 16 GB tem apenas um Bus de 64 Bits (1x64).
Óptimo para a Intel porque pode usar os mesmos módulos de LPDDR5X em todos os SKUs, mas péssimo para o consumidor.
Se comprarem algum produto com o Lunar Lake, não comprem a versão com 16 GB de RAM. Tem um hit de performance devido a este "detalhe".



Noutras notícias. A Intel cancelou o processo de fabrico Intel 20A, mas a notícia é dada de uma forma positiva. Se é ou não assim tão positivo, tenho algumas dúvidas, porque por muito bom que esteja a ser o desenvolvimento de Intel 18A, também passa a ideia que não tinham clientes externos para o processo 20A.

Continued Momentum for Intel 18A​

Progress on lead product designs and process readiness is enabling us to bridge from Intel 20A earlier than we’d planned.

Since releasing the Intel 18A Process Design Kit (PDK) 1.0 in July, we have seen positive response across our ecosystem and are encouraged by what we’re seeing from Intel 18A in the fab. It’s powered on and booting on operating systems, healthy, and yielding well – and we remain on track for launch in 2025.

One of the benefits of our early success on Intel 18A is that it enables us to shift engineering resources from Intel 20A earlier than expected as we near completion of our five-nodes-in-four-years plan. With this decision, the Arrow Lake processor family will be built primarily using external partners and packaged by Intel Foundry.

The journey to Intel 18A has been built on the groundwork laid by Intel 20A.

It enabled us to explore and refine new techniques, materials and transistor architectures that are crucial for advancing Moore's Law. With Intel 20A, we successfully integrated both RibbonFET gate-all-around transistor architecture and PowerVia backside power delivery for the first time, and these learnings have directly informed the first commercial implementation of both technologies in Intel 18A. This points to the iterative nature of semiconductor innovation, and we’re excited to bring these advancements to all Intel Foundry customers.

Focusing resources on Intel 18A also helps us optimize our engineering investments. When we set out to build Intel 20A, we anticipated lessons learned on Intel 20A yield quality would be part of the bridge to Intel 18A. But with current Intel 18A defect density already at D0 <0.40, the economics are right for us to make the transition now.

Ben Sell is vice president of Technology Development at Intel Corporation.

https://www.intel.com/content/www/us/en/newsroom/opinion/continued-momentum-intel-18a.html

 

[lmikotas]

Exclusive: Qualcomm has explored buying pieces of Intel chip design business​

• Qualcomm has explored buying pieces of Intel’s design business, sources say
• Qualcomm is interested in Intel’s PC business, a source says
• Intel explores sale of various units ahead of a board meeting in September, sources say

https://www.reuters.com/technology/...-chip-design-business-sources-say-2024-09-06/