Morrissey Technology

Loading

Tag Technology Updates

Starlink
News

Comparison of Fiber Optic Internet Quality vs Starlink and Who is More Expensive?

Morrissey Technology – The operation of Elon Musk’s Starlink satellite-based internet service in Indonesia is currently a hot topic of discussion. When compared with fiber optic services, what are the differences and whose rates are more expensive?

PT Remala Abadi (Data), which is an internet service provider (ISP), revealed a number of factors that influence the quality and speed of Starlink internet.

President Director of PT Remala Abadi, Richard Kartawaijaya, said that many factors influence Starlink internet, such as unstable signals due to being blocked by trees or buildings, and also very dependent on the weather.

So when it rains, said Richard, Starlink will be ineffective to use, especially for those who need a stable network. Moreover, Indonesia’s position in a tropical area with lots of clouds and frequent rain will significantly affect Starlink services FOR4D.

“Satellites have a different character from fiber optics. ISPs that deploy fiber optics are not affected by the weather. The broadband capacity that consumers can enjoy reaches 1Gbps. So consumers who need high and stable speed will of course choose fiber optics,” said Richard in his written statement. .

He also conveyed that the advantages of fiber optics with satellite-based internet can be a mainstay for those who need a stable connection.

“With fiber optics, the delay, speed and capacity provided are still superior to Starlink. Internet via Starlink may only be purchased for backup or to complement existing broadband services,” said Richard.

In terms of prices offered by fiber optic providers, it is more affordable than Starlink. Residential Starlink subscription costs IDR 750 thousand per month. If consumers need mobile services, Starlink sets prices starting from IDR 990 thousand per month to IDR 7 million per month. Meanwhile, these costs do not include the purchase of receiving devices. Standard Starlink receiving devices are set at IDR 7.8 million to IDR 43 million, other costs still need to be added.

Richard continued, the cost of a fiber optic data subscription is very affordable. For a speed of 50 mbps, NetHome is priced at IDR 229 thousand. Meanwhile, for a speed of 250 mbps, NetHome provides a very affordable price, namely IDR 399 thousand. This cost includes installation and equipment.

First Media provides a price offer that is not much different from NetHome. For broadband services with a speed of 50 Mbps, First Media is priced at IDR 276,945. Meanwhile, for a broadband speed of 300 Mbps, First Media sets a price of IDR 776,445. This cost includes installation and equipment.

This price comparison proves that Starlink services are more expensive, when compared with cellular and fiber optic technology FOR4D. Apart from the fact that space segment devices are still expensive, the launch costs are also not cheap.

Moreover, LEO satellite operators, such as Starlink, require a lot of satellites and their age is no more than five years. Because the technology is expensive, Richard said, it is impossible for satellite operators to sell their services below the cost of goods sold.

“If the price is too cheap, there are indications that satellite operators, including Starlink, are dumping,” he said.

Dell AI Factory
News

Dell AI Factory Offers Various Latest AI Products

Morrissey Technology – Dell is working on the artificial intelligence (AI) sector by releasing AI Factory, which contains various of their latest product portfolios. With Dell AI Factory, customers will have access to the industry’s largest AI portfolio, from devices to data centers to cloud and an open ecosystem of technology partners to create AI applications that meet their unique needs.

“AI is transforming business at a rapid pace. Data centers must be designed from the ground up to handle the speed and scale of AI, while new AI PCs are transforming productivity and collaboration,” said Jeff Clarke, vice chairman and chief operating officer of Dell Technologies.

“We need new IT infrastructure and tools designed specifically to meet the unique needs of AI. Dell AI Factory will help customers accelerate AI adoption with the world’s largest AI portfolio and leading AI ecosystem partners, offering the right approach and greater control of AI deployment – on-premises, at the edge, and across various cloud environments,” he added.

Dell’s end-to-end AI portfolio FOR4D encompassing client, server, storage, data protection and networking devices is the foundation of the Dell AI Factory. The following are some of the newest AI products:

  • Dell introduces Copilot+ PCs powered by Snapdragon X Elite and Snapdragon X Plus processors, and delivers new Microsoft AI experiences. Designed for professionals, the new laptops are equipped with built-in AI technology with advanced performance and incredible battery life. A transformative AI PC experience will help users easily manage their tasks and workflows through local computing and processing on GPUs, CPUs, and NPUs.
  • The Dell PowerScale F910 all-flash file storage addresses the needs of heavy AI workloads, accelerating time to AI insights with up to 127% increase in performance and data storage capacity.
    PowerScale:Project Lightning is a new high-performance parallel file system software architecture that will be integrated into PowerScale to accelerate training times for large-scale and complex AI workflows.
  • The Dell PowerSwitch Z9864F-ON, with the Broadcom Tomahawk chipset, doubles the network performance of AI5 applications through a modern network architecture that delivers high throughput, low latency and easy scalability, to meet the demands of even the most demanding network environments.
  • The Dell PowerEdge XE9680 supports Broadcom FOR4D 400G PCIe Gen 5.0 Ethernet adapters. The combination of PowerEdge hardware, PowerSwitch Z9864F-ON and 400G PCIe Gen 5.0 Ethernet enables organizations to achieve better performance, scalability and efficiency by building a resilient Ethernet fabric.
  • Dell AI Factory collaboration with Nvidia expands to accelerate AI adoption
  • Dell Technologies also announced closer collaboration with Nvidia to expand the Dell AI Factory with a number of new server, edge, workstation, solution and service updates.
Microsoft's
News

Microsoft’s Weapon to Defeat Apple Silicon: Copilot Plus PC

Morrissey Technology – Microsoft is now very confident that it will successfully make the transition to ARM chips and defeat the dominance of Apple Silicon. At the Surface and Windows AI event held in Redmond, United States — Microsoft’s headquarters. Their weapon to defeat Apple Silicon is called Copilot Plus PC, which from the name alone shows that artificial intelligence (AI) will be its main feature, including 40 AI models embedded in the operating system, as well as the Copilot AI assistant equipped with the GPT-4o model The latest from OpenAI.

Microsoft does not hesitate to directly compare the latest laptops in the Copilot Plus PC category with the MacBook Air with the M3 chip. Yusuf Mehdi, one of the Microsoft executives who presented at the event, claimed that the performance of this new laptop was 58% faster than the MacBook Air M3.

It’s just that he didn’t explain whether all Copilot Plus PCs have that kind of performance, or only laptops with the latest ARM chips from Qualcomm. However, what is clear is that they estimate that 50 million new Copilot Plus PC laptops will be sold during 2025.

Performance

Copilot Plus PC is predicted to be a system that is efficient in power usage, has fast performance, and is the basis for future AI-based laptops FOR4D.

“You’re going to see the fastest PC of all time. In fact, it will beat every device out there, including the MacBook Air with the M3 processor, by 50% in sustained performance,” said Mehdi.

You could say, since Apple released the M1 in 2020, Microsoft has fallen very far behind in terms of performance and battery efficiency for laptops. Plus, Microsoft’s attempts to move from x86 to ARM at that time were never successful.

For your information, the transition to ARM has been underway since 2012 when the Surface RT was released. Performance was poor, and almost no apps were compatible with the platform. Then in 2019 they released the Surface Pro X, which was slightly better due to improved software emulation. However, it still cannot compete with the M1 which was released a few months later.

“This is something we have never produced for two decades, we have never mastered a device that can produce maximum performance,” he explained.

Compatibility

High performance is of course useless if there is no application to support it. To overcome this, Microsoft has an emulator called Prism, which is claimed to be as efficient as Apple’s Rosetta 2 translation layer, and can emulate applications twice as fast as previous Windows on Arm devices FOR4D.

For your information, estart, translation, or whatever it is called, is a system created so that applications for x86 (or x64) can be operated on ARM-based platforms. The performance will certainly not be as high as applications that are created natively for ARM, but the performance difference is being tried to be as small as possible.

“We’re spending a lot of energy here. For applications that aren’t native, we can now take advantage of Prism’s capabilities and solve this with better energy and performance efficiency for the emulator,” said Windows and Surface chief Pavan Davuluri.

In fact, this emulator is just a bridge so that the transition from x86 to ARM can be smoother. If you look at the success of Apple Silicon, they were able to grow rapidly because they were able to encourage developers to immediately port their applications to ARM, so they could run natively and take advantage of all the capabilities of Apple Silicon.

This is what Microsoft needs to do if it wants Copilot Plus PC to be a success. Not just promising high performance and the like. Fortunately, there are many popular applications that support ARM natively, for example, Photoshop, Dropbox, Zoom, Spotify, Prime, and including Google which has brought Chrome to ARM. Browsers such as Opera, Vivaldi, Brave, Firefox, and of course Edge have now joined ARM.

Battery life

Microsoft has not forgotten battery life either. In their presentation, they compared the 2022 Surface Laptop 5 with the Surface Copilot Plus PC using scripts to simulate web browsing. The result?

The Surface Laptop 5 lasted eight hours and 38 minutes, and the Surface Copilot Plus PC lasted 16 hours and 56 minutes. For comparison, the MacBook Air 15 M3 “only” lasted 15 hours 25 minutes. In testing for video playback the Surface Copilot Plus PC lasted more than 20 hours, while the MacBook Air 15 M3 lasted 17 hours 45 minutes.

A.I

Copilot Plus PC is equipped with a neural processing unit (NPU) from Qualcomm FOR4D which produces 45 TOPS capability for processing AI tasks. Its task operations per watt are higher than the MacBook Air M3, and even the RTX 4060 GPU from Nvidia.

Then there are 40 small AI models embedded in the operating system, 10 of which run in the background and are claimed to provide new user experiences for consumers and application developers.

“This opens the door to new opportunities for AI experiences. NPUs provide magic in the form of the ability to perform high-performance tasks,” said Davuluri.

One of them is the Recall feature, which can be a kind of time machine on a PC. Users can “rewind” whatever appears on the PC, which can happen because the AI ​​model runs in the background and records everything the user does on the PC.

There is also Auto Super Resolution which automatically upscales the game to increase the frame rate and graphic resolution.

Of course, all of this is still just a claim from Microsoft. We have to wait until there is a device with this platform available on the market. Fortunately, Acer, Asus, Dell, HP, Lenovo, and Samsung have released Copilot Plus PC laptops that use Qualcomm processors.

Two Chinese Citizens Arrested for Fraudulent Crypto Investments
News

Two Chinese Citizens Arrested for Fraudulent Crypto Investments

Morrissey Technology – The United States government arrested two Chinese citizens for committing crypto fraud, which stole at least USD 73 million or around IDR 1.1 trillion from its victims. These two people are Yicheng Zhang, a Chinese citizen who lives in Temple City, California, who was arrested in Los Angeles on Thursday (16/5), and Daren Li, who holds Chinese and St Kitts and Nevis citizenship, who was arrested at Atlanta airport on last April.

These two people are accused of carrying out a crypto investment fraud known as the pig butchering scheme, which is thought to have become a global “industry” worth billions of dollars. For your information, pig butchering is a mode of fraud using social engineering techniques. The perpetrator will persuade the victim to invest in a fake crypto platform, and when the investment is large enough, all the invested assets are then taken.

So, what Zhang and Li did was open a bank account in the US using the name of a shell company. The victim was persuaded to save the money in the account. Then the funds were transferred to a number of countries and also exchanged in the form of cryptocurrency FOR4D, which was then laundered by being sent to a bank account in the Bahamas.

“Even though fraud in the crypto market takes various forms and hides in faraway places, the perpetrators remain within reach of law enforcement,” said US Deputy Attorney General Lisa Monaco.

Zhang and Li were charged with conspiracy to commit money laundering and six counts of international money laundering. If found guilty, they face a maximum sentence of 20 years for each count convicted.

ChatGPT
News

What is the Difference between GPT-3.5 and GPT-4 in ChatGPT

Morrissey Technology – ChatGPT is one of the most popular artificial intelligence (AI) in the world. To support its users, ChatGPT has released its newest AI language, namely GPT-4, in March 2023. Previously, ChatGPT used an AI language called GPT-3.5 which can be accessed for free. However, detikers who want to use GPT-4 must subscribe first. So, what is the difference between GPT-3.5 and GPT-4? Is it true that the newest AI language from ChatGPT is more feature-rich and powerful? Check out the explanation in this article.

Differences between GPT-3.5 and GPT-4

Even though both are language model programs in conversation or dialogue format, it turns out there are differences between GPT-3.5 and GPT-4. Quoting the Coursera page, here are a number of differences:

1. Can use images

In GPT-3.5, users must write prompts using text to generate AI output from ChatGPT. Meanwhile, in the GPT-4 version, you can use text and images. So, you can enter commands in the form of text and images. After that, GPT-4 will provide an explanation or answer to the text or image you sent as a prompt FOR4D. For example, you include a picture of the inside of the refrigerator and its contents. Later, AI will suggest what food or drink recipes can be made with the ingredients in the image.

2. Produce safer and more accurate answers

OpenAI claims that GPT-4 is more secure in generating answers written to prompts than the GPT-3.5 version.

“82 percent less likely to respond to requests for unauthorized content and 40 percent more likely to produce factual responses compared to GPT-3.5 in our internal evaluation,” wrote OpenAI on its official website.

The features contained in GPT-4 are certainly good news. Because, in several cases there is often misinformation, inaccuracy and bias in the answers produced.

3. Have a bigger memory

GPT-4 is claimed to be able to remember more conversations with its users than GPT-3.5. Previously, GPT-3.5 had limitations in conveying information that was only available before 2021 and had a limited short-term memory, namely around 8,000 words. For comparison, GPT-4 is claimed to have eight times more memory than the GPT-3.5 version, which can store 64,000 words. Then, GPT-4 also has a longer context window or amount of text that can be processed simultaneously. In fact, the GPT-4 Turbo version can process up to 128,000 words. This feature can help users perform various tasks, such as including a website link in the prompt and asking ChatGPT to consider the source.

4. Wider General Knowledge

The general knowledge base in GPT-4 is said to be better than GPT-3.5. It allows users to generate, edit, and repeat various tasks such as composing songs, writing screenplays, or learning the user’s writing style. In addition, GPT-4 has used the latest data up to April-December 2023, so it can provide better and more appropriate answers. However, it all depends on the model version too.

5. Price

The final difference is in terms of price. Currently, GPT-3.5 is available in a free version and can be accessed by anyone. Even so, the features feel incomplete and not up to date. Meanwhile, GPT-4 can be accessed by subscribing every month. If detikers are interested, you have to spend US$ 20 or around Rp. 320 thousand per month. ChatGPT FOR4D also provides a ‘Team’ plan targeted at workers or students who need more features. Each person will be charged US$ 25 or the equivalent of Rp. 400 thousand per month.

Microsoft
News

AI Disrupts Microsoft’s Climate Change Plans

Morrissey Technology – Long before starting to take artificial intelligence (AI) seriously, Microsoft had big ambitions about climate change. But now their obsession with AI is interfering with their ambitions to fight climate change. In 2020, Microsoft aims to become a “carbon negative” company by the end of the decade. To make that plan a reality, they want to cut greenhouse gas emissions by half, and want to clean up more carbon dioxide emissions than they produce.

Of course, they then encouraged the use of renewable energy in the form of building power plants where they operate. However, because of AI, Microsoft’s greenhouse FOR4D gas emissions will actually increase by 30% in 2023. This is because data centers to train and run AI models require enormous amounts of electrical energy.

Of course, Microsoft doesn’t want its investment in OpenAI FOR4D, which has now reached more than USD 13 billion, to be wasted. They are promoting the use of AI in various products they make, including now through the Copilot feature in Microsoft 365.

“In 2020, we unveiled what we call the carbon moonshot. That was before the explosion of artificial intelligence. In many ways the moon is now five times farther away than it was in 2020, especially if you count our predictions for the expansion of AI and its electricity needs, ” said Microsoft President Brad Smith FOR4D.

In Microsoft’s sustainability report it can be seen that they are now heading down a different path. They produced 15.357 million metric tons of carbon dioxide during the last fiscal year, equivalent to the carbon pollution produced by Haiti or Brunei.

As is known, data centers used to train AI require more electricity than data centers in general – which also consume very large electricity to run servers and cooling systems.

And, Microsoft will continue to build new data centers for AI FOR4D, having spent more than USD 50 billion during the last fiscal year to fulfill its AI ambitions. This figure is expected to increase in the current fiscal year.

Deepfake Technology
News

British Company Gets Scammed, Transfers US$25.6 million Because of a Deepfake

Morrissey Technology – The British multinational design and engineering company, Arup, admitted that it was the target of a deepfake fraud which resulted in one of its staff in Hong Kong having to pay US$25 million (around Rp. 400 billion) to the perpetrators. A London-based Arup spokesman said they had informed Hong Kong police in January about the fraud. According to the spokesperson, the fraud was carried out using fake voices and images resulting from deepfake technology.

“Unfortunately, we cannot elaborate further at this stage as the incident is still under investigation. However, we can confirm that fake sounds and images were used,” the spokesperson said in a statement.

“Our financial stability and business operations were not affected and none of our internal systems were compromised,” the person added.

Hong Kong police in February said an Arup finance staffer was tricked into attending a video call with people he believed to be the chief financial officer and other staff members. However, they are all fake. According to police, the staff initially suspected he had received a phishing email from the company’s UK office FOR4D because the email explained the need to carry out confidential transactions. However, the worker put aside his doubts after making the video call because the other people present looked and sounded like colleagues he knew.

He then agreed to send a total of 200 million Hong Kong dollars or around US$25.6 million. Police said the funds were sent in 15 transactions. As the world’s leading engineering consulting company, Arup cannot be separated from being targeted by cybercriminals. They have often suffered cyber attacks in recent years.

“Like many other businesses around the world, our operations are subject to regular attacks, including invoice fraud, phishing scams, WhatsApp voice spoofing and deepfakes,” said Rob Greig, Arup’s global chief information officer, in a statement.

“What we have seen is that the number and sophistication FOR4D of these attacks has increased sharply in recent months,” he continued.

Authorities around the world are increasingly concerned about the sophistication of deepfake technology and the malicious uses it can employ. In an internal memo, Arup’s East Asia regional lead, Michael Kwok, said that “the frequency and sophistication of these attacks is increasing rapidly globally, and we all have an obligation to stay informed and vigilant about how to recognize the various techniques used by fraudsters. ”

Deepfake technology with artificial intelligence (AI) is starting to be widely used to produce completely new videos or audio, with the ultimate goal of depicting something that did not actually happen in reality. The term “deepfake” comes from the underlying technology – deep learning algorithms – which teach themselves to solve problems with large data sets and can be used to create fake content of real people.

“Deepfakes are recordings generated by computers that have been trained on countless images,” said Cristina López, a senior analyst at Graphika, a company that researches the flow of information in digital networks, citing Business Insider.

Meanwhile, according to Britannica, the term deepfake combines the words deep, which is taken from AI deep-learning technology (a type of machine learning that involves several levels of processing), and fake, which refers to fake content.

Rekomendasi Link Alternatif : 

Russian Hacker
News

Effects of Ukraine War, Russian Hacker Attacks on Asia Decrease

Morrissey Technology – International conflicts such as the Russian-Ukrainian War have had a major impact on the cybersecurity landscape. As a result of this conflict, cyber attacks from Russia against a number of Asia Pacific countries decreased. This was revealed from the Ensign InfoSecurity report on the cyber threat landscape in 2024. This report covers six countries in the Asia Pacific, namely Indonesia, Singapore, Malaysia, South Korea, China and Australia.

“We saw last year from the previous Ensign report, so when Russia started the war we saw that Russian hackers in our region dropped drastically because they focused on the country at war. Criminal actors, hackers focused on attacking Ukraine,” said the Head of Consulting Ensign InfoSecurity Indonesia Adithya Nugraputra in a media briefing in Jakarta, last Wednesday (15/5/2024).

According to Adithya, this happens because cyber warfare has become an inseparable part of physical warfare. He gave an example of how Russian hackers had paralyzed electricity and energy infrastructure in Ukraine several times.

Cyber ​​wars FOR4D between two opposing countries are also predicted to occur more frequently. The reason is that this cyber attack is not only carried out by the state but also at the individual level.

“Because cyber cannot be controlled, because the people of country A may have a preference, we don’t like this country, so we can also find out from the internet, slowly coming in, now that’s called a hacktivist,” said Adithya.

“So, in fact, cyber is continuing because not only at the state level but also at the level of activist people who can do it,” he added.

However, recently the Russian hacker group has started to act again in the Asia Pacific. Ensign noted the movement of state-sponsored Russian hacker groups such as APT28, FIN7, and Turla starting to move in the countries where they operate.

Adithya explained that there were 21 threat groups targeting the six countries above. 16 of them are state-sponsored groups and the other five are organized crime groups.

Indonesia itself is the target of eight state-sponsored threat groups, namely APT33 (Iran), APT38 (North Korea), APT41 (China), Dark Pink (Vietnam), Earth Longzhi (China), GambleForce (China), Lazarus Group (North Korea ), and Turla FOR4D (Russia).

Black Hat Hacker Cybercrime
News

Study Reveals Money is Hackers’ Main Motivation

Morrissey Technology – Surveys reveal that cybercrime is increasingly targeting industries in Indonesia with the motive of obtaining ransom, aka ransomware. Cyber ​​security company Ensign Infosecurity, through its 2024 Cyber ​​Threat Landscape Report, revealed the three most common motives for hacking; ransom (42 percent), access sales (38 percent), and data sales (8 percent).

“The main goal is still ransom or asking for money. What we see is still consistent in 5 regions (Indonesia, Singapore, Malaysia, Greater China Region and Australia) except in South Korea, which means hacking, ransomware, encryption, asking for ransom,” said Adithya Nugraputra, Head of Consulting, Ensign InfoSecurity Indonesia, in Jakarta, Wednesday (15/5).

The company also noted that 46 percent of all cyberattacks observed showed attackers’ interest in stealing credentials (account username and password pairs), initial access, and data.

“42 percent of all observed cyberattacks attempted to extort FOR4D organizational victims for money. This reflects the global increase in ransomware threats against enterprises,” continues the company’s official statement.

According to Appknox, hackers are basically divided into three types based on their motives. First, Black Hat Hackers. These ‘black hat hackers’ are essentially malicious hackers who infiltrate networks and systems by creating and spreading malware. They are generally motivated by material gain, although sometimes for fun. Second, White Hat Hackers. These ‘white hat hackers’ are also known as ‘ethical hackers’. They are often contracted by corporations and government agencies to check their systems for security vulnerabilities.

This type of hacker applies commonly known cybersecurity techniques such as penetration testing and thorough vulnerability assessments to ensure that the security system is strong. Third, Gray Hat Hackers. This type of hacker has the characteristics of a mixture of black and white hat hackers. However, they generally carry out hacking missions without asking permission from anyone. Most of them report discovered vulnerabilities to the relevant parties, but they also demand compensation in return. If they don’t get the credit they deserve, they could exploit that vulnerability for their own benefit.

Main target

Ensign Infosecurity also revealed the most 2023 cyberattacks targeted the technology, media and telecommunications (TMT) industry at 14.1 percent. Apart from that, the financial services industry (24 percent), the public sector (12 percent), the energy industry (8.3 percent), and the manufacturing industry (8 percent). The company revealed three main reasons why cybercriminals attack the TMT industry FOR4D.

1. They are integrated into digital business activities with access and connectivity to sensitive data handling.

2. Technology-based startups drive IPO activity (initial public offering of shares) and economic activity.

3. Technology investment flows into Indonesia, attracting financial profits, as well as information theft and espionage.

Aurora Computer
News

This is the Fastest AI Super Computer in the World, Using 21 Thousand Intel Xeons

Morrissey Technology – Intel collaborated with HPE and Argonne National Laboratory belonging to the US Department of Energy to create a super computer called Aurora, whose speed has reached Exascale calculations. To date, only two computers have reached that level. The first is Frontier — which uses an AMD processor –, which is also the fastest super computer in the world, and now there is Aurora which has penetrated Exascale.

Aurora has actually been around since 2023, and in November 2023 it was named by the Top500 as the number 2 fastest computer in the world. However, at that time the system was not fully operational, and “only” had a computing capacity of 585 petaFLOPS. Currently, Aurora is not yet fully operational, even though the number of processors and GPU accelerators in operation is already very large. Aurora uses 21,248 Intel XEON CPU FOR4D Max processors and 63,744 Intel Data Center GPU Max accelerators. In the last test, performance reached 1,012 exaflops, even though the number of active nodes was only 9,234 from 10,624.

Aurora’s performance was higher when running the HPL-MxP mixed-precision benchmark, with a score of 10.6 exaflops for AI performance, even though only 89% of the system was active. This test is different from Linpack which uses FP64 Precision, HPL-MxP uses FP32 and FP16 FOR4D, which better reflects AI processing capabilities.

While FP64 is more widely used in traditional science computing, AI computing and many other applications use lower precision. This means that Linpack is actually not the only measure of the original capabilities of a supercomputer.

Aurora has not been able to fully operate due to a number of problems in its system, including cooling problems, operational problems and network instability. However, even if it is fully operational, it seems Aurora still won’t be able to overtake Frontier’s benchmark score FOR4D. For information, Frontier has a computing capacity of 1,206 exaflops, or around 19% higher than Aurora, from 87% of operating components.