Intel has identified the problem and pulled down this information, but Tom’s Hardware has taken a screenshot and saved it.

Sadly, we currently know nothing about this mysterious architecture, but we can currently confirm that Panther Lake and its associated iGPU will be announced during the 22nd Intel Architecture Day. If the news is true, we should see more official descriptions by then.

According to the details that can be seen so far, Panther Lake and its iGPU will arrive after Meteor Lake, Arrow Lake and Lunar Lake. Considering that the latest Lunar Lake is scheduled to come out sometime in 2024, Panther Lake will not come out until around 2024 or 2025 at the earliest.

Wait until around 2024 or 2025, when Intel may use the Intel 20A process (the second half of 2024 if there is no delay) to produce this series of chips.

According to reports, Intel 20A will also use a new packaging architecture called RibbonFET and an interconnect technology called PowerVia. Although Intel has not released Intel 20A performance or energy efficiency data, there is no doubt that it will bring a higher performance advantage over existing 10nm products Alder Lake and Raptor Lake.

For reference, the upcoming Intel 4 for Meteor Lake promises a 20% performance-per-watt improvement over the existing Intel 7, while Intel 3, the optimized version of Intel 4, will deliver an 18% performance-per-watt gain over that.

The post Intel engineers reveal Panther Lake architecture, coming in 2024/25 at the earliest appeared first on TechGoing.

According to Taiwan Economic Daily, Intel 20A and Intel 18A have been flowing (design finalization). This is the company’s latest process technology progress below 2nm, making the market more concerned about the process technology competition between it and TSMC.

Rui Wang, Intel’s senior vice president and chairman of China, recently said, “The semiconductor industry recovers from every huge drop, and we have no doubt about our strategy for IDM2.0, whose biggest challenge is our internal execution. No one likes the (2022) Q4 Intel earnings report, and we can’t afford to make adjustments.”

Currently, around the IDM2.0 strategy, Intel wants to break out in three ways. The first is the development of the foundry business, Intel in accordance with the 4 years 5 nodes, plans to regain the process technology leadership in 2025, and through the “system-level foundry” to establish fab differentiation.

Intel’s process technology planning in recent years has gradually moved from Intel 10 to Intel 7, Intel 4 process, and then to Intel 3, Intel 20A, and Intel 18A process development.

According to Wang Rui, Intel 7 has been mass production, Intel 4 will be on the market in the second half of this year, and will be used for Core processors (Meter Lake), Intel 3 is advancing as planned, Intel 20A (2 nm), Intel 18A (1.8 nm) test chips have been flowing.

It is worth mentioning that the market had also rumored that Intel’s 3nm process technology development was delayed, but later denied by Intel CEO Kissinger. He stressed that the relevant plans are proceeding as expected, so far all the announced 3nm projects will be released in 2024.

According to the plan, Intel’s Granite Rapids and Sierra Forest series processors will be produced using its own Intel3 process. According to Intel’s previously announced information, Intel 3 can further draw on the advantages of FinFET optimization, enhance the use of the EUV ratio, and bring more improvements. Compared with Intel 4 process, Intel3 can provide about 18% performance improvement per watt, and is expected to gradually start production from the second half of this year.

Wang also stressed that Intel is very confident that it will be able to return to the leading position by 2025.

The post Intel 20A and Intel 18A process development has been a breakthrough appeared first on TechGoing.

Intel 4: is advancing in volume production; expected to flow with the production process of Meteor Lake processors in the fourth quarter

Intel 3: Still on track; Intel 3 and Intel 4 are the first process nodes to feature EUV lithography, which will deliver significant improvements in transistor performance per watt and compute density

Intel 20A and 18A: First internal test chips have flowed; internal test chips are being produced in the fab for a major potential foundry customer

For those familiar with the semiconductor industry, you may know that the Intel 20A process is roughly at the industry’s 2nm level. The other two advanced chip makers, Samsung and TSMC, are still focused on the 3nm process, and there has been no significant progress on the 2nm process. For example, Samsung has only announced some technical solutions for 2nm, and TSMC’s 3nm process has even been delayed and postponed. At this time Intel announced the Intel 20A first batch of internal test chips has been flowing news will undoubtedly cause some impact on the other two, so Intel can rely on the “Intel 20A” process to return to the dominant position in chip manufacturing.

Intel 20A process
Although the Intel 20A process is roughly at the industry’s 2nm level, the industry is currently even the same technology node (for example, both 2nm processes), and the gap between different manufacturers is still very large. And Intel’s process characteristics and the other two have a large gap, so the “Intel 20A” directly understands the 2nm process is still very inaccurate.

At present, the other two 2nm process is not obvious progress, but still can be combined with the process roadmap and other information for rough projections.

In terms of transistor density.

Samsung 2GAE (2nm) < Intel 20A < TSMC N2 (2nm) < Intel 18A

In terms of performance.

Samsung 2GAE (2nm) < TSMC N2 (2nm) < Intel 20A < Intel 18A

PPA Aspect.

Samsung 2GAE (2nm) < TSMC N2 (2nm) < Intel 20A < Intel 18A

Note: PPA is the abbreviation of Performance (performance), Power (power), and Area (size), and is a comprehensive evaluation of the chip manufacturing process performance.

Intel 20A is not the “king of bombs” depends on the release time
“Intel 20A” process has already started the internal test flow, but there is still a distance from full-scale production. So whether Intel 20A can become a “king bomb”, on the one hand, depends on the Intel 20A mass production time, on the other hand, depends on the TSMC and Samsung 2nm process mass production time.

However, it seems that Intel’s progress is still relatively smooth, so we assume that Intel 20A is released as scheduled, and based on this assumption, we will analyze it.

According to current news, the Intel 20A process is expected to be in mass production in 2024, while TSMC and Samsung’s 2nm process mass production is expected in 2025. If this is the case, Intel 20A process is almost a complete win compared to the other two 3nm processes.

In terms of transistor density.

Samsung 3GAP (3nm) < TSMC N3E (3nm) = TSMC N3P (3nm) < Intel 20A

In terms of performance.

Samsung 3GAP (3nm) < TSMC N3E (3nm) < TSMC N3P (3nm) < Intel 20A

PPA.

Samsung 3GAP (3nm) < TSMC N3E (3nm) < TSMC N3P (3nm) < Intel 20A

Conclusion
1, if Intel, Samsung, and TSMC’s 2nm process are released as scheduled, then Intel 20A is the “king bomb”, Intel’s process will remain in the “world’s first” position for a period of time.

2, Intel Intel 20A although only the “internal test chip has been flowing”, compared to the other two 2nm processes progress is still relatively fast.

3, in the 2nm technology node, TSMC and Intel are the first to use the new transistor structure (Intel for RibbonFET, TSMC for NanoSheet), so theoretically compared to the previous generation process will have a greater improvement.

The post Intel disclosed the Intel 20A process, whether to return to a dominant position appeared first on TechGoing.