Chinese semiconductor company, Tianshu Zhixin, has launched the ‘Big Island’ GPGPU. This is China’s first domestically-developed 7nm GPGPU for the data centre and cloud market. The Big Island GPGPU has been designed for AI training, high-performance computing (HPC), and other industries, including education, Internet, finance, among many others.
According to Tianshu Zhixin (via TomsHardware), the Big Island GPGPU (general-purpose graphics processing unit) is based on the 7nm process node and makes use of 2.5DCoWoS (chip-on-wafer-on-substrate) packaging technology to accommodate 24 billion transistors. The foundry responsible for the chip’s production hasn’t been unveiled.
Cai Quangen, chairman of Tianshu Zhixin, said that the development process of the Big Island chip is a milestone to the company’s product development and to China’s HPC independent research and development. The development of this GPGPU had a duration of about two years, from the R&D phases to the mass-production.
The company claims that the Big Island GPGPU offers two times more performance than its competitors’ products while maintaining an attractive price-performance ratio. Featuring FP32, FP16, BF16, INT32, INT16, and INT8, and other multi-precision data mixed workloads, the Chinese 7nm GPGPU delivers an FP16 performance of up to 147TFLOPS.
Comparing the Tianshu Zhixin Big Island chip’s performance to the Nvidia A100 and AMD Instinct MI100, the competitors have an FP16 performance of 77.97 TFLOPS (312TFLOPs using RT cores) and 184.6 TFLOPS, respectively. Unfortunately, these numbers alone are not enough to decide if the new Chinese GPGPU will stand a chance against its competitors, but at least it gives us a glimpse of what to expect from them.
KitGuru says: Whether or not this particular chip makes its way to western markets remains to be seen. Do you think something like this would be successful in the market against Nvidia and AMD?
Become a Patron!
Check Also
Gigabyte’s RTX 3060Ti Gaming OC Pro graphics card is getting thinner
Gigabyte has launched a 2nd revision of the RTX 3060Ti Gaming OC Pro graphics card, featuring …
With two changes to the RHEL licensing, Red Hat probably wants to defuse its discontinuation of CentOS somewhat. On the one hand, the provider is expanding its developer program, with which individual developers can equip up to 16 systems with RHEL in the future. These can be used for test purposes, but it can also explicitly refer to systems that are in productive use are free. However, you must register with Red Hat with your own account or single sign-on with a third-party account. Support from Red Hat is not included; The provider also promises not to abuse the users of the program as an opportunity for sales.
Developer program for the entire department Furthermore, complete teams can be added to the developer program in the future, which also does not incur any additional costs and does not limit the options that can be selected. Systems at providers such as Amazon AWS or Microsoft Azure can also be equipped with the free RHEL via Red Hat Cloud Access.
The changes to the developer program are taking effect February 1st. In the announcement, Red Hat also promises to introduce additional programs for previous CentOS applications in mid-February. The regular CentOS will take its leave at the end of 2021, shortly after the step became known, CloudLinux announced a RHEL-compatible alternative.
In a blog post, CEO Shay Bannon justifies why Elasticsearch no longer appears as free software. Amazon is solely responsible for the step, whose behavior towards the developers of the search engine is unacceptable. Now, as the market leader, Elastic has to stand up to the group.
Specifically, it is about AWS offering Elasticsearch without Amazon cooperating with Elastic. The name Amazon Elasticsearch represents a trademark infringement, all attempts to clarify with the group have failed and ultimately made a lawsuit necessary.
Amazon Elasticsearch: Fork or no fork? Furthermore, AWS would repeatedly advertise its services with a cooperation with Elastic, which however does not exist. Again and again this confuses the community. Amazon’s own distribution for Elasticsearch caused even more irritation, with parts of the code coming from the developer’s commercial features.
Shay Bannon repeatedly emphasizes that nothing changes for the users – and he does I also do not speak out against cooperation with cloud providers, provided that this corresponds to the basic open source ideas of transparency, cooperation and openness. In fact, the CEO of Amazon publicly accused Amazon of violating these principles.
The license change will take place with the upcoming version 7. 11, from which the Elastic License and the Server Side Public License (SSPL) are available. All allegations against AWS can be found in the blog entry from Elastic. iX asked Amazon to comment on the allegations, which was not yet available at the time the report was published.
Amazon Web Services (AWS) has released the second major version of its Software Development Kit (SDK) for the Go programming language. According to the publisher, user feedback from the first version was incorporated and new features of the language were taken into account.
Modularize service clients The revised modularization is new: Version 2 of the SDK packages each AWS Service Client as an independent Go module. This should enable developers to model service dependencies in their applications. Feature updates and the service client for app control can thus be handled separately and the size of individual modules shrinks. Go modules have been around since Go 1. 13, the innovation in the SDK follows the general versioning strategy for Go.
Configuration and API client methods are simplified or unified: Instead of numerous ways to call an API, version 2 offers a single API client to call services . Service operations can be called directly on the client, with each method taking the API input along with the context and then returning the response or the error.
Less Pointer for API parameters required The pointers for API parameters have also been streamlined. In the new version, the SDK requires fewer pointer references than before. Types can be converted into values whenever this is possible according to the API model data. Apparently this also applies to the map and slice data. Enums are now part of the type string aliases, can be generated as constants and can be used directly in the code.
According to the AWS team, this simplifies the type support for enums in the IDE ( Integrated Development Environment). With a code example, the team illustrates how value types for the API filters of the DescribeInstance in the Amazon Elastic Cloud let set. Filters and Values become value types instead of pointers (as in version 1 of the SDK):
result, err: = client.DescribeInstances (context.TODO (), & ec2.DescribeInstancesInput {Filters: types.Filter {{Name: aws.String (“instance-state-name”), Values: string {“running”, “pending”, “stopped”,},},},}) Configuring with a single config type The new SDK version for Go sees a single Config – Type before what the interaction between session – and Config – type simplified. The SDK developers have moved the configuration flags specific to each service to the area of individual service client options. According to the blog entry, this is intended to make it easier to find configuration values that the Service Client supports.
With the command LoadDefaultConfig developers can obviously define the standard configuration values in the configuration package and load configuration values from external sources. It could look like the following example code – details on this can be found in the developer guide for the AWS SDK for Go.
// Load SDK configuration with shared config profile. cfg, err: = config.LoadDefaultConfig (context.TODO (), config.WithSharedConfigProfile (“profile-name”),) // Create Amazon S3 API client using path style addressing. client: = s3.NewFromConfig (cfg, func (o s3.Options) {o.UsePathStyle = true}) Cross-page pagination Call-back functions are no longer used in pagination with a single context across all paginated service calls, which made error handling more difficult in the past. Instead, the new SDK version provides the context page by page and allows iteration over API results that can span several pages. According to the publisher, performance should also be increased, so the AWS team refers to its own benchmark program and a half-cut CPU load.
The AttributeValue and Expression packages are also new Package from Amazon DynamoDB and Amazon S3 Transfer Manager for transferring large files in smaller portions, using goroutines.
More information can be found in the migration guide. Further code examples and more detailed information can be found in the release notification in the AWS blog.
Continuous Integration / Continuous Delivery – the automated construction, testing and rollout of software – is an important component of agile software development in order to ensure and improve software quality. In the two-day online course Introduction to Continuous Integration with Jenkins, you will learn in a realistic training environment how to install, configure, operate and integrate Jenkins into your infrastructure.
You will practice creating CI / CD pipelines for automated software tests and gain insight into typical use cases as well as the connection to the cloud and the use of Docker. You will also learn how to integrate Jenkins into the common developer tools Gitlab, Github, SVN, Redmine and Jira.
The course will take place on March 2nd and 3rd as an online workshop in the browser, so that You can participate comfortably from your own desk. It is limited to 15 participants in order to ensure intensive support for the participants. Speaker Johannes Kastl is a Linux consultant and certified trainer with a focus on system administration, system and configuration management and project management at B1 Systems. When booking by 1.2. you get 10% early bird discount.
Entry into the Continuous integration with Jenkins, online course, 2.-3.3. 2021: Further information and registration (odi)
In Germany, IBM will give almost 1000 employees the termination. The group has informed its supervisory boards and the responsible works council committees accordingly in the last few days with maintaining competitiveness and realigning the organization and skills. This was announced by the service union ver.di. In order to implement the layoffs, negotiations on redundancy plans are to be started with the responsible bodies of the companies belonging to the German IBM group . It is also still unclear which parts of the group are affected.
Restructuring and profits Last autumn, IBM expanded its infrastructure Services split off into a new company in order to focus even more on cloud services. In this area, IBM 2018 had the open source provider Red Hat converted for 30 Billion euros taken over. Red Hat is now part of the Cloud & Cognitive Services division. Here IBM was able to record a growth in turnover, while the income of the other services declined. Nevertheless, in the third quarter 2020, IBM was still able to post slightly increasing profits of 1.7 billion US dollars, the equivalent of almost 1.4 billion euros.
IBM CEO Arvind Krishna announced a growth target of five percent for the group in October. To this end, IBM wants to further develop its strategies and implement measures to simplify and improve the business model.
Worldwide, IBM has around 350. 000 Employees, in Germany the number of positions is over 10.000 estimated. The ver.di union protests against the layoffs and calls on IBM not to implement the plans. Information events and protests are being planned.
The next installment of the popular Hitman series will debut on the market tomorrow. On the eve of the premiere, a flood of reviews appeared, according to which the Danish studio IO Interactive on the new adventures of the Agent 47 with a barcode on the back of the head did a great job. At the moment, the game has received the highest ratings not only among all parts of the World of Assassination trilogy, but also the entire series, starting in 2000 of the year of Hitman: Codename 47 and ending with the last game of 2018 year. Hitman 3 is the culmination of this trilogy and, as critics’ opinions suggest, a goodbye. Selected reviewers mention that, with the previous two installments, it is one of the most complete trilogies in the history of games. On the other hand, there are also voices saying that this is more of an addition than a full-fledged installment.
According to reviews, Hitman 3 is one of the best stealth games in recent years. This is so far the highest-rated game in the series of adventures about the Agent 47, which offers .in. even better level designs or greater emphasis on the story than in the last two installments.
Hitman 3 for PlayStation VR in share. IO Interactive studio presents new gameplays from the third part of the Agent trilogy 47
Hitman 3 gets very high ratings. The average rating in the Metacritic and OpenCritic aggregators was at the time of writing 88 and 87% based on over
respectively and 60 Reviews . It is even higher than in the case of Hitman 2: Silent Asasssin, which is considered – next to Blood Money – to be the best part of this stealth series. It may not be possible to maintain such a high average after the premiere, but fans of the series should not be disappointed, especially those who played the game from the last two installments. According to critics, Hitman 3 offers a slightly developed, sandbox formula developed in Hitman from 2016. For fans of the series it will be “more of the same”, but apparently the devil is in the details, and IO Interactive studio has done its homework in this respect.
Hitman 3 – IO Interactive studio has revealed all six locations from the Agent game 47. I have to admit that they are atmospheric
In the third part we get access to six extensive maps with a different climate (the gloomy Dartmoor mansion in England, a skyscraper in Dubai or Chinese streets in Chongqing). Some of them are the best of the entire trilogy. Each mission, except the last one with a more linear structure, can be performed in many different ways, encouraging repetition. Players who missed a greater emphasis on the story will be pleased that it is more extensive in Hitman 3, and the goals and the plot around them have become more intriguing. However, there are not enough changes in the game that would encourage people who do not like this formula to play. And the general content, devoid of, among others online modes, makes Hitman 3 seem to offer less than the previous installments of the trilogy. We will see it 20 January 2021 of the year when the game will be available for sale on PC (temporarily exclusive to Epic Games Store) and PlayStation 4, PS5, Xbox One and Xbox Series X / S consoles (on Nintendo Switch is to work via the cloud) in the English language version (there will not even be Polish subtitles).
The State Social Court of Thuringia has obliged the job center to take over the purchase of an internet-enabled PC with accessories for an eighth grader whose family receives Hartz IV. The equipment that should enable the student to take part in the corona-related online lessons include a computer, screen, keyboard, mouse, printer and three printer cartridges. The judges determined that the costs should be a maximum of 500 euros. According to the Social Security Code II, they represented an “unavoidable, ongoing additional requirement” (file number L 9 AS 862 / 20 B ER).
In the statutory requirement for Hartz IV recipients, such expenses are not taken into account – but would be under the current pandemic circumstances this is no longer “realistic”, the court found according to the announcement. Since there have been no face-to-face classes since December, the acquisition is rather necessary to guarantee the applicant’s right to education and equal opportunities. In the household of the schoolgirl’s family there is only an internet-enabled smartphone that is not suitable for using the Thuringian school cloud. The school also does not provide any equipment.
No right to “best possible care” However, the court restricted that the The schoolgirl is not entitled to the device she has chosen for the price of 720 euros. It is not about “the best possible care”, but only about the satisfaction of simple and basic needs. An inexpensive and possibly used device that is sufficient for the purpose must be sufficient. The job center can now fulfill its imposed obligation by assuming the cost of an acquisition or by providing the applicant with a device and accessories.
The mother had previously the student made a corresponding application at the job center, but it was rejected. The Nordhausen Social Court, which was subsequently involved, had also denied a claim. The family then went to the regional social court, whose interim order to the job center is incontestable according to the notification.
Pointing the way, but not binding Social courts in Germany are repeatedly faced with the question of the extent to which the job center has to help families out with hardware for the school – and the judges do not always see an obligation to assume the costs as in this case. A court spokesman stressed to the MDR that it was a case-by-case decision. Although this is groundbreaking, it is not binding for job centers where other Hartz IV recipients in a comparable situation also submit such applications.
The decision was made in the urgent procedure, the judgment in the main proceedings is still pending. In this way, the court takes into account the fact that a decision should only be made when children can go to school normally again, the court spokesman continued.
The e-mail provider mailbox.org of the Berlin company Heinlein Support GmbH is to provide business e-mail inboxes for Berlin teachers in the future. Teachers can also use other services such as calendar, address book, task management, Office Suite, file storage and video conferencing.
Departure from private e-mail accounts According to a decision by the Berlin Senate Department for Education, Youth and Family (SenBJF), Berlin teachers should no longer use their private e-mails for school communication. Access mail addresses. According to the SenBJF press release, mailbox.org was chosen as the service provider, as it fulfills “the necessary requirements in terms of data security, encryption technology and user-friendliness”.
As in In times of Corona, data protection-compliant communication can be implemented in schools is currently the subject of many debates. For example, the Baden-Württemberg Ministry of Culture’s plan to use commercial solutions such as Microsoft Office 365 is very controversial.
Mailbox.org also for private users The fee-based e-mail provider mailbox.org also offers its services for private individuals and advertises to be safe, anonymous and ad-free. In addition to office functions and cloud storage, the company also offers confidential video meetings.
Asus is currently in the process of updating its homebrewed graphics overclocking utility, GPU Tweak, from version 2 to version 3 (or, because Roman numerals are still cool apparently, from version II to version III). Overclocking and tuning is a great way to eke a bit more performance, or potentially lower temperatures, from the best graphics cards, and you can typically boost your fps by 5-10 percent, inching up the GPU benchmarks. GPU Tweak 3 just released it to the public with an open beta and features a slew of UI and feature updates, including a refreshed UI and Nvidia’s OC scanner.
The most striking update to GPU Tweak 3 is its newly refreshed UI. Mostly gone is the ROG themed text, instead replaced by a more traditional font that looks like Arial. Placement of the sensor info and slider adjustments have been significantly altered as well. Now the sensor info resides along the top of the UI, and the adjustment sliders are placed more horizontally in the middle and lower sections of the app.
For monitoring your GPU, Tweak 3 includes both a refreshed monitoring utility and an Asus skinned version of GPU-Z built right into the program. Tweak 3’s monitoring utility gets a refreshed UI as well and features much more visible information on-screen without scrolling thru all the sensor data.
Overall Asus has made Tweak 3’s UI more optimized and easier to control.
Image 1 of 2
(Image credit: Future)
Image 2 of 2
(Image credit: Future)
Regarding tools, Asus has updated GPU Tweak 3 with more tools at your disposal. Asus gives you the Xplit Gamecaster for free (same as GPU Tweak 2), a mining program called Quantum Cloud — which is just another mining program to make you a few cents thru GPU mining. (If you’re serious about mining, there are far more profitable tools available.) Finally, you get a direct download to ROG’s FurMark utility, which is Asus’s modified version of the vanilla FurMark stress testing tool. (You can download it separately if you don’t use GPU Tweak 3).
Asus has also updated and added a few other things including a cleaner OSD, better fan controls, support for Nvidia’s OC Scanner, and a new profile connect program that will allow you to change OC profiles based on what game/app you are using.
If you want to try GPU Tweak 3 out for yourself, you can download it here. GPU Tweak 3 just released to open-beta so don’t expect a bug-free experience. Still, the previous iteration has been around for many years so the core functionality should be fine. GPU Tweak 3 can also work with non-Asus cards, though functionality varies.
This week PC users subscribed to Xbox Game Pass will be able to redeem Control at no additional cost. Microsoft announces it, confirming the imminent debut of the Remedy title in the Windows catalog 10: that’s when you can install it.
by Pasquale Fusco published 19 January 2021 , at 16: 21 in the Videogames channel Xbox Microsoft
Last December Microsoft announced big news for its acclaimed Xbox Game Pass . The gaming on demand service inaugurates the 2021 expanding the offer reserved for its users, introducing new games for console, PC and Android catalogs.
Among the new entries we also find Control , acclaimed action developed by Remedy Entertainment , which will soon be redeemable by Game Pass users on Windows 10.
Control lands on Xbox Game Pass PC on 21 January 2021
The debut of Control on Xbox Game Pass for PC is planned for next time Thursday, 21 January . This is confirmed by Microsoft itself through a fun teaser released on Twitter a few hours ago: “Things you can control in Control: furniture, debris, enemies , severity “.
Things you can control in Control: see below
Things you can ?? t control: your hype until it ?? s available in 4 days. pic.twitter.com/SfPGGqx 18 u
?? Xbox Game Pass For PC (@XboxGamePassPC) January 17, 2021
This is how the latest game of is described ) Remedy (Max Payne, Alan Wake): “After a secret agency in New York is invaded by an otherworldly threat, you become the Director struggling to regain control. This third-person action / adventure will challenge you to master a combination of supernatural abilities that will keep you on a tightrope “.
We remind you that, at the time of writing, the subscription to Xbox Game Pass for PC is offered at 9. 99 euro per month, with the possibility of purchasing the first month for 1 euro .
For the richest plan Ultimate we find a ‘similar offer: users can get the first three months for 1 euro and will pay 12. 99 euro for the following months. Xbox Game Pass Ultimate includes the subscription Live Gold – with related offers – and that EA Play , which adds an additional collection of Electronic Arts-branded games to the on-demand catalog.
Most importantly, with Ultimate it is possible to play via the cloud with your Android device . As announced by Microsoft, this feature will also be available on PC and iOS devices: a Beta version will be released during the spring 2021.
Microsoft wants to be the next technology company to get involved in the self-driving car business. The Windows giant is participating in a two billion dollar financing round of the robotaxi company Cruise of the auto company General Motors, as the company announced on Tuesday.
Cruise was given a total of 30 Billion dollars. The company has been testing its vehicles in San Francisco for years and has already presented the prototype of an autonomous taxi without space for a driver. Among other things, Microsoft wants to bring services from its cloud platform Azure to the partnership – which Cruise will mainly use in the future. In addition to the parent company GM, the Japanese car manufacturer Honda is also participating in the financing round.
Straggler Microsoft The other big tech companies are already very active in this area: The Google sister company Waymo applies As a leading developer of Robotaxi technology, Apple is also working on systems for autonomous driving and its own car, Amazon bought the robot car developer Zoox and placed it in its device division. Cruise got approval last year to put cars on the streets of San Francisco without a security driver behind the wheel. Waymo and Zoox had also previously been given permission to drive on public roads without people behind the wheel. Cruise originally announced the start of a Robotaxi offer for 2019, but then postponed the date indefinitely.
The Gesellschaft für Informatik (GI) is concerned that many schools are using “proprietary software solutions that are questionable under data protection law” against the background of the corona crisis and the current level of suffering in homeschooling. Above all, programs that transfer data from those affected to third countries such as the USA are questionable in view of the “good alternatives” available. There were “powerful and easy-to-use cloud solutions and learning management systems” that allowed data protection-compliant, legally secure operation on European servers.
Resources required The GI names the freely available programs Moodle and Ilias, which support interactive cooperative forms of work with plugins such as H5P and provide a “suitable basis for the IT infrastructure of schools” represented. In order to be able to operate such solutions smoothly and to avoid failures, the federal states would have to provide the necessary human and technical resources. This is the only way to “continuously improve” the open-source solutions in order to further develop conditional and competitive applications.
In general, the association advocates the digital Strengthen the sovereignty of students, schools, teachers and the IT location through open source software. This makes it possible to show transparently “which data is processed how and by whom”. At the same time, “the dependence on large corporations will be reduced”. The public view of the source code allows vulnerabilities to be found more quickly and security gaps to be closed. In this way, everyone involved could “act independently, self-determinedly and securely in the digital world”.
“The current political decisions for digital infrastructures will shape our education system for years,” emphasized GI Vice President Ulrike Lucke . “Transferable standards and formats” should be promoted “in order to avoid lock-in effects”. It has been proven that schoolchildren stayed with a digital tool they had got to know well beyond their school days. The GI therefore welcomes the fact that in the dispute over the digital education platform in Baden-Württemberg there is a broad alliance against the planned introduction of Microsoft Office 365 as a solution for all schools.
Currently high hurdles Two thirds of the students, parents and teachers experience In digital teaching during Corona times, Initiative D 21 meanwhile has a representative, in July through the Market research institute Kantar conducted survey found. The problem most frequently mentioned with 42 percent was the inconsistent approach to how and where teaching material is made available: the teaching material and how to fill it in of questionnaires mostly remained analogous.
37 percent found the self-organization required of them to be stressful, 27 percent complained about inadequate explanations on how to use the technology . There was a rarer barrier in the infrastructure: 16 Percent of those affected stated that they did not have enough equipment 14 percent complained about a poor internet connection. “Insufficient data protection” hits eight percent badly. 78 Percent support that teachers should receive compulsory digital training. 74 percent demand that schools try out new forms of learning and working methods more often. 60 percent fear that the current situation will exacerbate educational injustices.
What does the year 2021 look like in terms of the most significant smartphone technologies of recent times?
The smartphone market has become quite saturated in recent years, and the past year, helped by the corona, has indicated that the rise in the price level of high-end phones in particular is beginning to reach its peak. Many in the industry are of the opinion that too little is really being seen in the new phones, but on the other hand, the proliferation of new technologies such as 5G and the folding screen has not been particularly well received by everyone. Given the potential of folding display technology to change the structure and usability of smartphones as a whole, or the impact of 5G on the evolving potential of mobile devices, criticisms of the lack of development in the sector seem somewhat contradictory in some places.
The first folding display phones were already on the market in the first half of the year 2019, but the year 2020 brought more technical solutions more mature, although there is still a long way to go for a mass market-friendly implementation. . Samsung has so far been the most prominent of the manufacturers pushing folding display technology to the phone market, and the company has announced plans to launch more and cheaper folding display phones this year. So far, there is no more detailed information on how low Samsung can push the price level, but the fact is that the high price level has so far been one of the biggest brakes on the spread of new display technology. As long as the price level is lowered and options that appeal to ordinary consumers are introduced to the market, it will only be possible to see whether the folding screen has enough bangs to create a new trend in the smartphone market. However, during the year 2021, a significant increase in the use of folding telephones is unlikely to be seen. According to preliminary data, folding screen phones from other major manufacturers are also promised this year, so the supply will certainly increase
. Like the foldable display technology, the first smartphones with 5G support were already seen in the year 2019, but it wasn’t until last year that 5G made a strong case for mid-range devices in addition to high-end phones. The 5G technology used in phones developed by the year 2020 by far the fastest energy, but there is still room for improvement in this area compared to 4G. 5G devices also went well in trade, although not necessarily primarily with user needs, but with supply and market control. In other words, especially in the second half of the year, for those who buy a phone over 300 as a whole, the best options were almost entirely 5G models, even though there was no need for a new generation connection yet. . In addition, sales of 5G models in particular were boosted by very attractive offers. On the subscription side, Finnish operators shifted their emphasis to 5G subscriptions through both pricing and selection. By the end of the year, virtually all of the faster subscription options had been converted to 5G, and the fastest 4G subscriptions were priced so that choosing a 5G subscription was very attractive.
In 2021 5G support will become more widespread in more affordable price ranges and will be seen in phones less than 144 phones . The spread of 5G is being accelerated by system circuits with 5G modems aimed at increasingly lower price ranges, which are currently Mediatek’s cheaper Dimensity models and Qualcomm’s Snapdragon 690 and 480 models. Phones with 4G connections alone are fast becoming the exception and mainly the feature of the cheapest price range.
From the average user’s point of view, the proliferation of 5G will not provide a major sudden upheaval
From the average user’s point of view, the proliferation of 5G will not provide a major sudden revolution, but will mainly improve the performance and speed of mobile networks as technology becomes more common. It is also worth remembering that there is still a kind of transition going on, where 5G NSA (non-standalone) network technology also partially utilizes the old 4G networks. “Full-blooded” 5G networks based on cloud-based 5G Core technology are still in the testing phase in Finland and operators have not announced their exact launch schedules, but they can be expected to be completed within 1-2 years, according to io-Tech. Only Standalone 5G technology brings with it the most significant benefits of the new network technology, such as lower delays and much more versatile traffic management, which enables e.g. slicing of network capacity on a per-user basis
To answer the title question, both 5G and folding screens are still in the maturing stage, but their future prospects are quite different. 5G will become the new norm in the next few years, and while the overall level of service on mobile networks will improve, many basic users will not pay much attention to it. Folding screens are also in the process of maturing, but their future is so far a full question mark, and the year ahead is unlikely to determine anything lasting about the future of technology.
Other smartphone features already seen last year, but becoming more widespread this year, will include adaptive high-refresh rate displays as well as flagship cameras with four different focal lengths. Over the past year, most high-end phones have already found a high refresh rate display 90 and 144 Hertz fork. However, only in a few cases was the refresh rate a content-adaptive variety. This year, at least 144 Hertz screens will become more common in more expensive end phones even more and an adaptive refresh rate will be available for more and more models. 90 and 120 Hertz displays will also become more common in mid-range models.
On the camera side, rear camera assemblies with four different focal lengths were still quite rare last year. This year, they may be seen on higher end phones more than before. A typical combination of four cameras is likely to be a combination of ultra-wide-angle, wide-angle, short-zoom (normal / telephoto) and long-camera, which can already cover a fairly wide focal range. With this, of course, the size of the camera nodules will also increase even more. The number of megapixels for image sensors will also increase, as Samsung, for example, is rumored to release a 200 megapixel sensor. Increasing image processing performance will allow for faster processing and utilization of larger image data as well as data from multiple cameras simultaneously.
On the battery side, charging speeds will certainly increase even further, but the benefits for the consumer are already beginning to be questioned in many places, as already significantly lower charging powers have already achieved quite fast charging speeds. Manufacturers would therefore like to focus more on features that extend battery life, such as limiting charging power and setting a charge level limit, which are currently only found in phones from a few manufacturers. The omission of chargers from the sales package, a trend that is becoming more common based on the latest flagship releases from Apple, Xiaom and Samsung, brings its own spice to the soup. The solution will certainly reduce unnecessary electronic waste, but will also increase sales of accessories and make it more challenging for the basic consumer to find the full potential of the phone.
io-Tech’s goal is also to handle all major smartphone publications published in the Finnish market in the year 2021 as soon as possible. As launched last year, the most significant smartphone releases will be covered in the form of fresh first-time video, and the actual test article will be published in writing as soon as possible on the site side. The aim is to better optimize the publication time of articles compared to last year through pre-planning and a shorter lead time (time from the arrival of the test device to the publication of the article).
We also selected the best smartphones of the year 2020 in four different price categories during io-Tech’s mobile delivery. In the price range below 200 our choice was Nokia 5.3, below 450 in the euro price range Samsung Galaxy A 51 5G, less than 700 in the OnePlus 8T price category and over 700 in the OnePlus 8 Pro flagship category. In the video below you can see a more detailed justification for the choices as well as the worst challengers in each price range.
China has taken another step on the way to becoming an independent semiconductor industry: According to its own statements, Shanghai Tianshu Zhixin Semiconductor has designed a GPU accelerator developed entirely at home, of which the first test chips are running in the laboratories. It is intended to be similar to Nvidia’s Ampere-GPU A 100 for data centers with a focus on machine learning, specifically for training AI algorithms. The actual GPU accelerator from Shanghai Tianshu Zhixin Semiconductor belongs to the “Big Island” family and includes 24 Billions of transistors, paired with four memory stacks of high-bandwidth memory (HBM, probably HBM2 or HBM2e). The company relies on a manufacturing process with structure widths of 7 nanometers – the packaging process “2.5D-CoWoS” speaks for contract work from TSMC.
China can apparently own Designing GPUs with an AI focus, but not yet producing them with modern exposure technology. The Chinese chip order manufacturer SMIC currently produces small quantities with 14 – nm structures.
High computing power On paper, the GPU accelerator presented can be displayed with 147 FP 16 – TeraFlops for a debut work. “Big Island” can also be used with the integer formats INT8, INT 16 and INT 32 and Bfloat 16 and FP 32 bypass. For comparison: Nvidia’s A 100 – card creates 78 FP 16 – TFlops over the shader cores and 312 FP 16 – TFlops using the integrated tensor cores (624 with sparsity, which removes zeros from the matrices). AMD’s Instinct MI 100 reaches around 185 FP 16 – TFlops.
The hardware is only the half the battle – the drivers, especially with GPUs, take up a lot of work. In the case of machine learning, the software stack is also important, where Nvidia and Co. have years of experience.
Huawei has with the Ascend 910 has already presented its own AI accelerator, but in the form of a processor without a graphics unit, which simplifies development. Before the trade war between China and the USA, Huawei had the chip manufactured by TSMC with 7 nm technology; meanwhile the manufacturer had to stop production. Alibaba, in turn, has its own 12 – Nanometer AI chip Huanguang 800 presented for the Alibaba cloud; But this is an AI inferencing processor.
(mma)
We use cookies on our website to give you the most relevant experience. By clicking “Accept”, you consent to the use of ALL the cookies.
This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.
