The university plan is a very good job. When you encounter a problem that you don't understand well, you can directly send an email to a professor who is in a field between Cow A and Cow C. Intel's 22nm 3D technology cow, in the end, to what extent, in the end has a real impact on the industry, he is not very clear. No, an email is all done. The following is his reply, for children who want to know more technical details:

Dr. Xie (Kevin:)

The shock of this news is very big, its significance is:
(1) First of all, the world's second in the Process, Technology, Fab, compared to IBM has been far behind by Intel: 2 years. Two years is of great significance to semiconductors. Large Fab foundries such as TSMC, GlobalFoundry, and Samsung have become increasingly incapable. This is why TI and AMD have decisively abandoned FAB.
(2) Due to Intel's advancement in technology, more and more speculations have emerged. At the same time, the industry also believes that Intel's Fab will sooner or later go to OEM, that is, Intel's Fab is open to produce chips for Fabless, such as Xilinx. . Only in this way can the world enjoy the super compuTIng brought by the latest technology. At present, Intel has already done a 22nm OEM for an FPGA Startup (possibly acquired by Intel because the reconfigurable CompuTIng of the attached FPGA is a trend). I originally sent this news to everyone. The second target may be Apple, which means that Apple's chips will be processed by Intel in the future.
(3) The emergence of the 3D Tri-gate greatly extended Moore's Law. Various alternative technologies such as CNFET (carbon nanotransistor) and NanoPrint (an alternative lithography) are expected to be discarded at 22 nm. The 22nm ultra low power implementation is achieved through: 1. The high-k metal grid appearing at 45nm has greatly reduced the gate leakage; 2. The 3D gate structure greatly enhances the gate-to-channel control, and the full depletion of the channel greatly reduces the leakage of BTBT; 22nm ultra-high speed implementation Is passed: 1. The use of Intel's stress GeSi technology greatly enhances the drift speed of the channel carriers; The use of a 3D gate allows the transistor's Threshold Voltage to be strongly controlled. Wait

Answer your question now: Is it related to ARM? We talk about it in two ways:

(1) Architecture: Intel's x86 architecture is different from ARM architecture. Intel's is for general compuTIng, the depth of the pipeline, various complex Cache considerations, etc. will definitely determine that his architecture is a complex architecture for super computing, which will bring a lot of overhead. Unlike ARM, it is an embedded, ultra-low power design. So from an architectural point of view: ARM is a low-power architecture superior to Intel x86. This is why Intel's Atom is adjusted, and can't do the ultra-low power consumption of the ARM chip. I mean the same technology node, everyone uses 65nm or 45nm technology.
but:
(2) The ARM-based chip cannot be manufactured using Intel's fab, that is, he cannot enjoy the speed increase and power consumption brought by the technology aggressive scaling. That is, ARM chips are currently manufactured with up to 45nm or less mature 32nm technology. But Intel's Atom can be manufactured in 22nm. So the 22nm ATOM should pose a huge challenge to ARM's ultra-low power consumption.
If Intel decisively abandons the x86 architecture on the embedded CPU ATOM architecture, then he can benefit from both Architecture and Technology. This is also the basis of the intel tick-tock model.

In the field of embedded CPUs, Intel and ARM are forever rivals, and there is no partnership.

In the field of chip manufacturing, Intel is at most an indirect partner of ARM. Note: ARM never manufactures chips, which means that ARM is not a Fabless company which has and sells its fabricated chips. It is just an IP licensor. The indirect cooperation between Intel and ARM was generated by ARM's IP purchaser. For example, Apple's products use ARM as part of the SOC. Apple hopes that Intel will make Apple's SOC chip. At this time, Intel indirectly has a relationship with ARM.
Apple's rumored Apple uses Intel to do OEM work. This foundry business is likely to make Apple's own designed chips, without ARM, because Apple has already acquired a CPU company in the early years, the latest version of the Mac and smart terminals. May use the legendary Apple's A5 Processor.
In addition, once Intel opened its foundry business, the impact on TSMC will definitely be great. Because in terms of technology, TSMC is not an opponent of Intel. But this impact should not be instantaneous. Xilanx, Quacomm, Broadcom and other fabless companies are familiar with TSMC's technology. It takes time to digest the new Intel technology.

N-BK7 Window

The window is one of the basic optical components in optics. It is an optical flat panel that protects electronic components/sensors/semiconductor components in the optical path.
K9 (BK7) has excellent mechanical properties and very low bubbles and impurities, and has good scratch resistance. Due to its stable chemical properties, special grinding and polishing are not required. It is used in many fields such as optoelectronics, microwave technology, diffractive optical element and so on.
The window does not change the optical magnification and only affects the optical path in the optical path. Mechanical properties such as light transmittance, surface accuracy, thickness, parallelism, and substrate material are important parameters of the window. You can choose the window with the appropriate parameters according to the specific application. Such as zinc selenide (ZnSe) windows, germanium (Ge) windows, silicon (Si) windows, barium fluoride (BaF2) windows, calcium fluoride (CaF2) in the infrared (NIR/IR) spectral band Windows; Fused Silica windows used in the ultraviolet (UV) spectral band; K9 windows used in the visible light (VIS) spectral band, etc.


N-Bk7 Window,Optical Window,Laser Protection Lens Window,Window Wedge

Bohr Optics Co.,Ltd , https://www.bohr-optics.com