Apple moves to 5-nanometer processor technology

In the fall of 2020, Apple plans to debug the full-scale industrial release of its A14 processors, which finally fully use the latest 5-nanometer technology (TSMC).

 

As part of the annual technology conference, the main details of both 5-and 3-nanometer processors were presented.

 

Based on the fact that Apple processors are manufactured only by Taiwan Semiconductor Manufacturing Corporation (TSMC), which is paid very good money by Apple to ensure that their latest products (iPhone and iPad, with the addition of iMac in the near future) use only the most modern technologies developed by TSMC.

At the annual technology conference TSMC Technology Symposium, the announcement of production process optimization was presented, which is already beginning to draw a General idea of what can be expected from the new Apple processors.

 

In any case, the main criterion for the performance of new Apple processors will be their architecture. At the same time, a very large role in this case is assigned, among other things, to production processes, which help to get an idea of how complex, large and productive such processor chips can be, what their power consumption level is at the specified clock frequency parameters, and how fast they can work over a long period of time.

In the context of announcements presented on the site Anandtech.com it is already possible to get a rough idea of the chips expected from TSMC for Apple: in terms of 5-nanometer extensions, the processor for Apple A12 was made using 7-nanometer technology, and in the production of A13 processors for iPhone 11, an expanded upgraded version of the second generation is used.

We expect Apple to debug the full-scale release of A14 processors made using 5-nanometer technology in the fall of this year.

 

It is also expected to release iMac computers with their own chips, also based on 5-nanometer TSMC technology.

 

What is the difference between processors based on 5 nanometer technology from TSMC? The developer claims that the latest technological process will make it possible to place 80% more logic circuits on the same substrate area, as well as reduce power consumption by 30% at the same clock frequency and increase performance by 15%.

 

Most likely, given the specified parameters increase even these optimistic data by 5-15% within development release 2021 module N5P, you can expect at least increase performance by 5% while reducing energy consumption by 10-15%.

 

However, the real increase in transistor density, we do not see – that is, to increase one jump substantial productivity gains in the fall of 2021 will have to increase the area of the chip, which will increase the cost of products.

 

Prospects for the development of a 3-nanometer technology processors at this time, presented the possibility of increasing transistor density by 70%, which will give a real performance increase by about 70% and the growth speed by 10-15% reduced energy consumption by 25-30%…

 

So Apple, as always, goes its own way, investing in technology and the future.

NVIDIA very quietly became the owner of ARM

While there are marketing wars between major players in the field of various high-tech products-NVIDIA has acquired ARM-the main developer of mobile processors.

 

The transaction became known from the information posted on the company’s official website. As it became known, the industry giant will need to pay the previous owner (SoftBank) 21 billion dollars in shares, as well as 12 billion in cash.

 

 

This gives SoftBank the opportunity to potentially earn another $ 5 billion from future earnings. In addition, SoftBank employees should receive a total of about $ 1.5 billion due to additional obligations under this transaction.

 

The final transition, after complying with all the rules and getting all the approvals from regulators in the UK, EU, USA and China, will take approximately 18 months.

What will this do for users of computers and other devices?

 

One of the co-founders of ARM, Hermann Hauser, has already expressed his concerns about the sale. He believes that such a deal with NVIDIA could cause the collapse of the industry.

The fact is that one of the main positions in the ARM business model is the ability to sell its products to every company that expresses such a desire and will be able to produce such chips strictly according to the technology.

 

SoftBank’s unique advantage so far has been that, while not being a chip manufacturer, it has been able to maintain both legal, manufacturing, and marketing neutrality in terms of distributing ARM products.

 

If ARM becomes part of NVIDIA, most of the companies that own licenses for the production of such chips will most likely have to look for alternative options…

 

At the same time, NVIDIA plans to maintain the open license approach model for the time being.

 

In addition to including ARM in its system, NVIDIA is also going to become an investor in the construction of a research center for the development of artificial intelligence – ARM’s experience should help with the integration of all the developments and technologies available in the overall system to create a leading computing company in the era of AI.

 

Taking into account all the directions of development of corporations such as Microsoft and Apple in terms of their use of ARM technologies in their products, more large-scale purchases and mergers are possible in the coming years…

Hardware vulnerabilities: 63% of companies they already know in practice what it is…

According to a published report by Dell, penetration into the cyber environment using vulnerabilities at the hardware level has recently become quite popular among hackers.

Not to say that this was such a closed topic – these vulnerabilities have been known for many years; but only now, with increased protection at the software level and implemented everywhere corporate systems for tracking security loops-such options as the use of loopholes in the hardware of both computers and network equipment have become more relevant.

 

 

 

 

At the same time, despite the critical nature of the level of attacks at the hardware level, only 59% implemented a hardware protection strategy.

And this is despite the fact that 63% of companies said that they had experienced at least one incident of information leakage during the year due to vulnerabilities in hardware protection.

 

With the increase BIOS memory – and, respectively, with the increase in there code, plus appeared in recent years, the possibility of software updates from the local network and the Internet – and, of course, the appearance of the ready channels of communication with the motherboard as computers and laptops, and all networking equipment has its own operating system (and not only – it printers, scanners, multifunction devices, SIM cards and more – importantly, to present OSes)…

 

Of course, network equipment manufacturers try to keep up with the times. But often their strategy is not so much in patches (patching “holes” in software, but rather in providing a paranoid level of secrecy in everything that concerns the functioning of their equipment – up to receiving encrypted update files via email). Of course, this brings quite a lot of benefits – without having information about the system, it is quite difficult to remotely collect the necessary amount of data to organize hacking with high-quality hardware protection. But, as usual, the weakest link in the protection of information is a person, and therefore the frequent cases of hacking companies that specialize in protecting information are extremely indicative. After all, they have a lot of data on the basis of which the security audit of many companies is conducted with a full description of hardware and software, schedules and security strategies…

 …From the report data: 47% of the surveyed company representatives said that over the past year they have received at least two incidents at the hardware level, and most of them were carried out with the help of external attacks – 29%, phishing  -43%, when exploiting software vulnerabilities – 41%, as well as web applications-40% or mobile malware – 38%…

 

Consequences of this policy of company management: these violations can lead to harmful consequences, including

– loss of confidential data (52%);

– financial losses related to system downtime (39%);

– slow it recovery time (36%);

– and failures that affect systems served by customers (35%).

 

However, it is encouraging to see that almost everyone has declared hardware security in their IT systems as a top priority for the coming year.

By investing in stricter security measures, organizations will see significant benefits. Companies reported major benefits, such as increased overall security of their businesses (55%), reduced equipment costs (39%), increased business continuity (44%), and accelerated digital transformation initiatives (42%).

 

This is where security at the chip level comes out in the first place – and it becomes clear that corporations-users of IT products want to understand what they are buying and someone has a company that produces a product in our time is not ready to be responsible for the entire production cycle, such a product is always the fruit of complex technologies, with a load of various patents and factories scattered around the world for the production of various components – and it is not always possible to understand the level and number of “engineering inputs” to the chip, intended most often for testing equipment, and how this will affect the ultimate security of a particular device.

 

By 2020, cybersecurity is becoming more sophisticated due to several factors-starting with the pandemic, the transition of many companies to remote office mode, and the undoubted improvement in the quality of software implemented during the pandemic to ensure the security of remote activities of corporate employees around the world.

 

 

As another new component – with the increase in free time, a new hobby appeared – deepfakes based on artificial intelligence. Adding this type of entertainment to the potential of the soon widespread introduction of 5G networks, especially with the developing and waiting for the speed of the Internet of things – we should expect more and more ways of hacking at various levels. Some deepfakes with fake voices and videos with executives can bring a completely unforgettable experience with completely practical consequences in the form of reputational and financial losses…

 

Thus, the concept of a single direction and security standard comes out in the first place, can at the level of the blockchain and, if possible, two – factor authentication of both a person and devices for correct activity both at the level of access and privileges in systems.

In a few years they will finally be published Apple’s own video cards.

At the WWDC 2020 conference, Apple released information about the plan to switch over the next two years on computers with chipsets from Intel to products of its own design and production. The architecture, according to the company’s previous statements, is ARM.

In addition, AMD’s products (in terms of graphics accelerators) will also lose their relevance in the light of Apple’s own solutions.

 

 

However, it is unlikely to expect the announced integrated graphics accelerators from Apple in the near future – according to official plans of the company, even in 2021, MacBook Pro 16 laptops and iMac monoblocks will use Central processors from Intel and AMD Radeon graphics will be used.

The latest computer models, at the same time, will receive relatively powerful integrated solutions in terms of graphics accelerators. The only question is whether the overall performance of such a bundle will meet the needs of users.

Historically, a similar approach took place in Intel and AMD, when they tried to create an alternative to the desktop version of integrated GPUs in laptop computers, and then in desktop systems. Experience has shown the inferiority of this approach, this niche has worked in the budget variant.

 

Of course, Apple’s undoubted advantage is decent code optimization for its own hardware and high autonomy promised in its new ARM processors, which in theory should offset the disadvantages of analogues identified when implementing such platforms in Intel and AMD.

In General, presumably, Apple will completely abandon Intel Central processors and AMD video cards in its Mac computers around the end of 2021 or in the middle of 2022. Of course, it is unlikely that the proposed new graphics will immediately be able to compete with the top video cards from AMD, but the company does not set this goal, at least not so quickly…

Also, clearly not earlier than 2022, but still will have to appear and discrete video cards from Apple.

 

According to tests, at this time, “the Apple a12z Bionic single-chip system in OpenCL tests bypasses the embedded graphics in AMD Ryzen 5 4500U and Intel Core i7-1065G7 chips. This year’s upcoming 5-nm a14x Bionic chip for future iPad Pro tablets promises to be much more powerful — according to some estimates, it will be on par with the 8-core Intel Core i9-9880H. There are rumors that the first 12-inch MacBook based on ARM, expected this year, will get a 12-core processor — it will be interesting to see what performance such a system can offer.”

The Antimonopoly service of Russia recognized Apple as a violator

The Russian Antimonopoly regulator FAS (Federal Antimonopoly service) has declared Apple a violator – the regulator said that the company abused its dominant position in the iOS app market by restricting third-party apps in the App Store using non-competitive methods.

Similar lawsuits against the IT industry giant have been filed in many other countries in Europe and Asia. Russia also issued a court decision on a complaint filed last year by Kaspersky Lab in relation to the parental control program Kaspersky Safe Kids (KSK).

Interestingly, the FAS of the Russian Federation made such a conclusion from the review of the materials that Apple reserved the right to reject the application of a third-party developer even if it fully meets the company’s requirements.

Quote from the FAS Russia press release:

 

“The case (the case of signs of violation of part 1 of article 10 Of the law on protection of competition) was initiated on the application of Kaspersky Lab about the unjustified rejection of versions of the parental control program Kaspersky Safe Kids (KSK) by Apple, as a result of which the next version of KSK lost a significant part of its functionality. At the same time, Apple introduced its own Screen Time app to the market in iOS version 12, which coincides with parental control apps in its capabilities.”

Also, the FAS concluded that Apple, which occupies a “dominant position with a 100% share in the market for distributing mobile apps on the iOS operating system,” has been following unfair practices since October 2018. The company, as stated on the Agency’s website, “limited tools and capabilities for developing parental control applications, resulting in the loss of most of the functionality of third-party applications.”

 

The FAS of Russia should soon issue an order to Apple to eliminate violations.

Pacemakers and the level of access: threat to national security

Today, there are many hidden threats to access levels in various IoT devices – this has been widely discussed recently.

But few people imagine the real state of Affairs: all those small devices that have a real operating system, which the average person does not even know that it can be here – in cars, planes, space satellites, as well as in printers, cellular cards, Bank card chips, and, finally, in pacemakers – can be subject to hacking and unauthorized access, and not only to them, but also to those systems where these devices have access by default. This can be both personal computers of specialists and company servers, as well as cloud services.

That is, pacemakers, insulin pumps for diabetics, hearing AIDS with cloud-based speech recognition-and other medical systems-are potentially dangerous in terms of computer security.

Of course, designers of such devices strive to make them as compact as possible, energy-efficient for greater autonomy, and at the same time sufficiently productive for reliable performance of their direct functions. However, this also applies to other systems, such as the widely discussed Autonomous video surveillance cameras (they need autonomy to record events in the event of such incidents as, for example, a power outage, a failure of cellular communication to transmit data, or, for example, deliberate suppression of the network signal using special devices – when the camera still writes to its own memory and then when restoring communication, transmits the recorded data to the server).

Such incidents affect the device’s autonomy, performance, and service life. That is, if the same pacemaker must constantly monitor the extreme critical parameters of the heart, the so – called peak values, and record them in memory, and then transfer them to the specialist’s computer when visiting a specialist cardiotechnical center-there is, in the absence of basic security measures, a very high probability of obtaining unauthorized access to both the pacemaker itself and its connection infrastructure.

It doesn’t make sense to consider such scenarios as viral infection of the pacemaker yet, since it will happen in any case if there is interest.

For the first time, public attention was drawn to this threat in 2017, but the practical conclusions of researchers have not yet found a response from manufacturers of medical equipment. On the one hand, this is understandable: no one will remove devices from the human body for such reasons. If we also take into account the fact that the age of patients in clinics is often over 60 years old, when the regeneration processes are significantly slowed down and the immune system works in a reduced mode, the risks of replacing the pacemaker are high.

Therefore, to the credit of doctors and technicians, recognizing the problem and the need to improve security, medical centers went to update the firmware of devices. About 446,000 pacemakers were voluntarily recalled for firmware.

In General, it is estimated that from 22% to 45% at this time are subject to such threats, and developers are working to achieve the proper level of security.

Power is canceled indefinitely: MacBook moves to ARM processors …

Potentially, changing one processor platform to another, in this case – from Intel to ARM architecture – is an unpredictably lengthy process that includes both technological and financial and marketing aspects of the company’s activities.

 

Of course, the main question is why is this transition happening? Why is Apple switching to ARM, especially its own production?

 

The new name of Apple’s ARM processors is “Apple Silicon”. This, as well as the fact of avoiding the use of processors from Intel, was announced at WWDC 2020. The transition is planned to be implemented gradually, over 2 years.

 

The decision to change the processor platform was officially made to strengthen the position of autonomy and management of the bundle of hardware and software platforms of Apple, and to improve overall performance.

 

However, there are alternative versions. In particular, there is an opinion of some former Intel employees about problems with the quality of the Skylake family of chips.

Historically, this will be the third generation of processors used by the Apple platform.

In all cases, the motive for the transition was the principle of increasing the autonomy and performance of the architecture (that is, Motorola 68000, then PowerPC, and finally Intel).

Learn more about Apple’s modern ARM:

ARM Holdings is the name of a UK company that develops processors and other electronic components.

ARM issues licenses for its products – that is, other companies can purchase a license to create their own chips on this architecture, create processors to implement the list of arm commands.

At this time, the processors used by Apple are the leaders in performance – both in mobile phones and tablets, clearly the best in their segments in this regard. An important aspect is also the control over the combination of hardware and software.

Also, however, the performance of ARM processors that are supposed to be installed in Apple computers is clearly much lower than those currently used, and the autonomy is significantly higher.

 

In addition, there are other difficulties that, of course, the company’s management will have to overcome in the very near future: first of all, the use of the translator program “Pozetta” (which was written to convert code written for PowerPC for Intel processes).

Then the process is repeated – you will need to convert the code for ARM (which will require, respectively, “Pozetta-2”). Thus, as long as a translator is needed, there will inevitably be losses of about 30-40% of the processor power for the code translation process itself…