Intel announced its new Xeon 6 processors at the Computex tech conference in Taiwan on Tuesday 4th June 2024.
This announcement coincides with the recent launches of new artificial intelligence chips by rivals Nvidia and AMD on Sunday and Monday 2nd and 3rd June 2024 – as they compete for dominance in the rapidly growing industry.
Intel is making efforts to catch up with Nvidia and AMD, having been relatively absent from the AI surge that led tech giants such as Meta, Microsoft, and Google to purchase a significant number of Nvidia chips.
This comes half a year after Intel’s release of its 5th Gen Intel Xeon processors for data centre workloads and a couple of months following the announcement of the Gaudi 3 processor for AI model training and deployment.
Intel also disclosed that the Gaudi 2 and Gaudi 3 AI accelerators are priced lower than those of its competitors.
Furthermore, Intel shared architectural details of its forthcoming Lunar Lake processors, aimed at expanding the AI PC category. These processors, slated for release in the third quarter, are set to rival Nvidia’s and AMD’s offerings tailored for AI PCs.
While Nvidia and AMD focus on chip design, Intel stands out by both designing and manufacturing its chips. Nevertheless, Intel’s foundry business has faced challenges, with its operating loss widening to $7 billion in 2023 compared to the previous year.
Scientists have recently achieved a remarkable breakthrough by creating pure silicon, which could pave the way for quantum computing
The world’s purest silicon
Researchers have developed an ultra-pure form of silicon, known as silicon-28 (Si-28), which is fundamental for ‘silicon-spin qubits’ in quantum computers. This advancement addresses a major challenge in quantum computing: the ‘fragile quantum coherence.’
Quantum computers tend to accumulate errors quickly due to slight environmental changes, affecting their dependability.
Quantum bits, or qubits, are analogous to classical computer bits but are extremely sensitive to environmental interference.
Technical
Current quantum computers, even when cooled to near absolute zero, can only maintain error-free operation for a very short time.
This new technique generates qubits by embedding phosphorus atoms into crystals of pure, stable silicon. A concentrated silicon beam then directs onto a silicon chip, replacing impurities with pure silicon.
As a result, the impurity levels in silicon have been significantly reduced, from 4.5% to a mere 0.0002%.
David Jamieson, a project co-supervisor from the University of Melbourne, mentioned that the team achieved this level of purity using a standard piece of equipment – an ion implanter – that’s typically found in semiconductor fabrication laboratories.
Richard Curry, a professor at The University of Manchester where extensive research took place, believes that this advancement could accelerate the development of operational quantum computers. Processes that might have taken a decade to complete could now be accomplished in potentially half that time or less.
Potential impact
Practical quantum computers have the potential to revolutionize numerous fields
Energy Optimization: They can solve intricate problems related to energy.
Artificial Intelligence: Quantum computers may significantly boost AI capabilities.
Drug Discovery: They could expedite drug development and molecular simulations.
Communication: They can enhance encryption and communication protocols.
The creation of the world’s purest silicon represents a significant step forward in the development of large-scale quantum computers.
Intel’s new chip will go head-to-head with Nvidia and AMD
Intel unveiled new computer microchips on Thursday 14th December 2023, including Gaudi3, a chip for generative AI software.
Intel also announced Core Ultra chips, designed for Windows laptops and PCs, and new fifth-generation Xeon server chips. Intel’s server and PC processors include specialized AI parts called NPUs that can be used to run AI programs faster.
AI race
AI models, like OpenAI’s ChatGPT, run on Nvidia GPUs in the cloud. It’s one reason Nvidia stock has been up nearly 230% year to date while Intel shares have risen 68%. And it’s why companies like AMD and, now Intel, have announced chips that they hope will attract AI companies away from Nvidia’s dominant position in the market.
Gaudi3 will compete with Nvidia’s H100, the main choice among companies that build huge factories of the chips to power AI applications, and AMD’s forthcoming MI300X, when it starts shipping to customers in 2024.
CEO Gelsinger
‘We’ve been seeing the excitement with generative AI, the star of the show for 2023,’ Intel CEO Pat Gelsinger reportedly said at a launch event in New York where he announced Gaudi3 along other chips focused on AI applications.
Intel upping the anti with its Gaudi AI chip. The AI PC to become the new AI start of 2024 and beyond!
‘We think the AI PC will be the star of the show for the upcoming year,’ Gelsinger added. And that’s where Intel’s new Core Ultra processors, also announced on Thursday, will come into play.
Nvidia’s revenue grew 206% from year 2022 during the quarter ending 29th October 2023, according to data from Nvidia.
Net income, at $9.24 billion, or $3.71 per share, was up from $680 million, in the same quarter of 2022.
The company’s data centre revenue came in at: $14.51 billion, up a massive 279% and above consensus of $12.97 billion. Half of the data centre revenue came from cloud infrastructure providers such as Amazon, and the other from consumer internet and large companies, Nvidia said. Healthy uptake also came from clouds that specialized in renting out GPUs to clients.
Earnings: $4.02 per share, adjusted, vs. $3.37 per share expected
Revenue: $18.12 billion, vs. $16.18 billion expected
The gaming segment contributed $2.86 billion, up 81% and higher than the $2.68 billion general consensus. Nvidia’s future guidance suggested $20 billion in revenue for Q4, implying a nearly 231% revenue growth.
Year on year Nvidia share price movement.
Year on year Nvidia share price movement – Nov 2022 – Nov 2023
During the quarter, Nvidia announced the GH200 GPU, which has more memory than the current H100 and an additional Arm processor onboard. The H100 is expensive and in demand. Nvidia said Australia-based Iris Energy, an owner of bitcoin mining data centers, was buying 248 H100s for $10 million, which works out around $40,000 each.
Nvidia share price moved down 1% in after-hours trading on Tuesday 21st November 2023 after the reporting fiscal Q3 results that surpassed predictions. But the company called for a negative impact in the next quarter because of export restrictions affecting sales to organizations in China and other countries.
‘We expect that our sales to these destinations decline significantly in the Q4 2024, though we believe the decline will be more than offset by strong growth in other regions’, Nvidia reported.
Nvidia has recently announced its latest high-end chip, the GH200 Grace Hopper Superchip, which is designed for training AI models at giant scale.
The GH200 is a breakthrough accelerated CPU that combines the NVIDIA Grace™ and Hopper™ architectures using NVIDIA® NVLink®-C2C to deliver a CPU+GPU coherent memory model for AI and HPC applications. The superchip delivers up to 10X higher performance for applications running terabytes of data, enabling scientists and researchers to reach unprecedented solutions for the world’s most complex problems.
The technical bit
The GH200 features 72 cores of Grace CPU outfitted with 480 GB of ECC LPDDR5X memory, as well as the GH100 compute GPU that is paired with 141 GB of HBM3E memory that comes in six 24 GB stacks and uses a 6,144-bit memory interface.
The GH200 also has a new 900 gigabytes per second (GB/s) coherent interface, which is 7X faster than PCIe Gen5, and supercharges accelerated computing and generative AI with HBM3 and HBM3e GPU memory. The GH200 can run all NVIDIA software stacks and platforms, including NVIDIA AI Enterprise, HPC SDK, and Omniverse™.
Nvidia unveils its newest GH200 high-end AI Superchip.
The GH200 is available as part of the NVIDIA DGX GH200, a massive memory supercomputer that fully connects 256 GH200 Superchips into a singular GPU. The DGX GH200 offers 144 terabytes (TB) of shared memory with linear scalability for giant AI models.
The DGX GH200 is a turnkey data centre-class solution that includes integrated software and white-glove services from NVIDIA, from design to deployment, to speed the ROI of AI.
The DGX GH200 is the only AI supercomputer that offers a massive, shared memory space of 144TB across 256 NVIDIA Grace Hopper Superchips, providing developers with nearly 500X more memory to build giant models.
Investors gobbled up UK microchip designer Arm Holdings at its U.S. debut on the Nasdaq on 14th September 2023, sending its market value soaring to $60 billion (£48.3 billion).
The shares ended the day worth more than $63 each, after climbing by almost 25% from the high end start of $51 per share set by Arm.
The sale was the biggest initial public offering of the year, raising $4.87 billion for owner Softbank Group.
Despite some concerns surrounding the company’s exposure to risks in China and a potential AI slowdown – the shares soared.
British tech
A star of the British technology industry, Arm designs microchips for devices including smartphones and game consoles. It estimates that some 70% of the world’s population uses products that rely on its chips, including nearly all of the world’s smartphones. And with AI nestling in on the horizon, the future potential for Arm is massive.
Arm stock chart 14th September 2023
Arm said it expects the total market for its chip designs to be worth about $250 billion by 2025, including new growth areas such as data centres and cars.
Legacy
Many of Arm’s royalties come from products released decades ago. About half of the company’s royalty revenue of $1.68 billion in 2022, came from products released between 1990 and 2012.
Bright Future
The future looks bright for Arm but the company is trading at more than 25 times its most recent full year of revenue, and at more than 100 times profit.
And that could be where things get tricky for Arm in the not too distant future. Projections for future profits will be interesting, esecially if it’s to keep up with Nvidia for instance.
Chip design firm Arm on 5th September 2023 submitted an updated filing for its upcoming initial public offering on the New York Stock Exchange, setting a price range between $47 and $51. Only 9.4% of Arm’s shares will be freely traded on the NYSE.
Arm was previously listed in London and New York, before SoftBank acquired it for $32 billion in 2016.
Chip design firm Arm on Tuesday is looking to acquire as much as $4.87 billion in its upcoming initial public offering on the New York Stock Exchange, according to the new filing.
The deal could value the company at as much as $52 billion
As a British company, Arm qualifies as a foreign private issuer in the U.S. and its shares will count as American depositary shares, or ADS’s. It is reported that the company will list some 95.5 million ADS’s at a price range of between $47 and $51. At the upper end of that range it is estimated that Arm will likely raise up to $4.87 billion. At the lower end, the IPO would fetch $4.49 billion of fresh capital for Arm. It could do even better.
Institutional funds
When the company floats in New York, it will look to enjoy a very deep pool of professional institutional funds. Arm seeks to ramp up its investments in research and development, particularly as it pursues growth in the artificial intelligence (AI) space with some of its newer chips. The company recently released new chips specifically targeted at AI and machine learning use cases.
Arm seeks up to $52 billion valuation in U.S. IPO
Upper end
At the upper end of the pricing range, Arm would also touch a total valuation of $52 billion or more. Only 9.4% of Arm’s shares will be freely traded on the New York Stock Exchange, with SoftBank expected to own roughly 90.6% of the company’s outstanding shares after the completion of the IPO.
Arm’s listing is set to be the biggest technology IPO of the year. Investors are hoping that the listing could breathe new life into an IPO market that has been ‘slack’ since 2022.
250 billion chips globally
Arm says its energy-efficient processor designs and software platforms are integrated into more than 250 billion chips globally, into products ranging from sensors and smartphones to supercomputers.
The company estimates it enjoys approximately 48.9% share of the market for semiconductor design. Other players, such as Intel and AMD, have raced to catch up on designing their own chip architectures, but have struggled so far.
U.K. misses out… again
The U.K. government had originally hoped Arm would list on the London Stock Exchange, but the company instead dealt a major blow to Britain’s ambitions to become the leading global tech hub by opting for New York. The U.S. financial center has a deep institutional investor base and analysts who have a close understanding of the technology sector.
BIG interest
Chip design firm Arm said in a Tuesday filing that Apple, Google parent Alphabet, Nvidia and other technology companies are interested in buying up to $735 million in its shares as it seeks to go public on Nasdaq.
The investments might not happen, but the fact that these companies are considering them underlines the importance of Arm, whose designs are used for processors in data center servers, consumer devices and industrial products.
Arm chip – some 250 billion chips globally
Chip makers Intel, Samsung and TSMC are interested in investing alongside the three trillion-dollar technology companies, along with AMD and MediaTek, which make chip designs based on Arm architectures. Cadence Design Systems and Synopsys, which make electronic design automation software for processor development, have also expressed interest, according to a revised prospectus for Arm’s shares sale. This IPO could easily be the biggest of the 2023!
As part of the deal, Arm could wind up with a $52 billion market capitalization and almost $5 billion in new cash.
This is likely to be the biggest IPO of 2023
It is estimated that there will be about 19 billion devices using the Arm processor in the world by the end of 2023.
Arm target
The market share of Arm across different technology markets worldwide, which was 90% for mobile application processors, 34% for embedded computing, and 5% for data center and cloud in 2019.
Arm has a target of increasing its market share to more than 90%, 50%, and 25% respectively by 2028.
Arm is a British semiconductor and software design company that is known for its Arm processors, which are widely used in smartphones, tablets, laptops, and other devices. Arm was founded in 1990 as a joint venture between Acorn Computers, Apple Computer, and VLSI Technology. The company was originally called Advanced RISC Machines, but later changed its name to Arm Ltd in 1998.
In 1985, the first Arm silicon chip was created by Acorn engineers Sophie Wilson and Steve Furber, who designed a 32-bit processor with a simple and elegant instruction set.
In 1990, Arm was spun off from Acorn as a separate company, with Apple as a major investor. Arm’s first product was the ARM6 processor, which was used in Apple’s Newton personal digital assistant.
Impression of the Apple Newton PDA device
In 1993, Arm introduced the ARM7 processor, which became one of the most successful embedded processors in history. It was used in devices such as the Nokia 6110 mobile phone, the Nintendo Game Boy Advance, and the Lego Mindstorms robotics kit.
In 1994, Arm launched the ARM9 processor family, which offered higher performance and lower power consumption than previous generations. The ARM9 was used in devices such as the Sony PlayStation Portable, the Palm Treo smartphone, and the Amazon Kindle e-reader.
In 1997, Arm introduced the ARM10 processor family, which featured a superscalar architecture and a floating-point unit. The ARM10 was used in devices such as the Apple iPod, the Samsung Galaxy S smartphone, and the Raspberry Pi computer.
In 1998, Arm changed its name from Advanced RISC Machines to Arm Ltd, reflecting its global expansion and recognition.
In 1999, Arm launched the ARM11 processor family, which featured a vector floating-point unit and a TrustZone security extension. The ARM11 was used in devices such as the iPhone 3G, the Nintendo DS, and the Raspberry Pi Zero.
In 2000, Arm became a public company, listing on the London Stock Exchange and the Nasdaq. The company raised £213 million in its initial public offering.
In 2001, Arm introduced the Cortex processor family, which offered a range of performance, power, and cost options for different applications. The Cortex processors are used in devices such as the Samsung Galaxy S10, the Apple Watch, and the Tesla Model 3.
In 2005, Arm acquired Artisan Components, a provider of physical intellectual property (IP) for chip design. This enabled Arm to offer a complete solution for system-on-chip (SoC) development.
In 2006, Arm announced the Mali graphics processing unit (GPU) family, which complemented its CPU offerings with high-performance graphics capabilities. The Mali GPUs are used in devices such as the Huawei Mate 20 Pro, the Oculus Quest, and the Samsung Smart TV.
Artistic image of ARM chip
In 2009, Arm partnered with IBM, Samsung, Texas Instruments, and others to form the Linaro consortium, which aimed to improve the Linux software ecosystem for Arm-based devices.
In 2010, Arm unveiled the Cortex-A15 processor, which was the first Arm processor to support virtualization and big.LITTLE technology. The Cortex-A15 was used in devices such as the Google Nexus 10, the LG G3, and the Nintendo Switch.
In 2011, Arm announced the Cortex-M0+ processor, which was the world’s most energy-efficient microcontroller. The Cortex-M0+ was used in devices such as the Arduino Nano 33 IoT, the Fitbit Flex 2, and the Nest Thermostat.
In 2012, Arm launched the Cortex-A53 and Cortex-A57 processors, which were the first Arm processors to support the 64-bit ARMv8 architecture. The Cortex-A53 and Cortex-A57 were used in devices such as the iPhone 6s, the Samsung Galaxy S6 Edge+, and the Microsoft Surface Pro X.
In 2013, Arm acquired Geomerics, a developer of real-time lighting technology for video games. This enhanced Arm’s graphics portfolio with dynamic illumination and global illumination effects.
In 2014, Arm introduced the Cortex-A72 processor, which delivered a 50% performance improvement over the previous generation. The Cortex-A72 was used in devices such as the Huawei P9, the Xiaomi Mi 5s Plus, and the Amazon Fire HD 10.
In 2015, Arm announced the Cortex-A35 processor, which was the most efficient Arm processor for smartphones and tablets. The Cortex-A35 was used in devices such as the Nokia 2.1, the Samsung Galaxy J2 Core, and the Lenovo Tab M7.
In 2016, Arm was acquired by SoftBank Group for £24.3 billion, becoming a subsidiary of the Japanese conglomerate. The deal was motivated by SoftBank’s vision of investing in technologies that would drive the future of artificial intelligence (AI), internet of things (IoT), and smart cities.
In 2017, Arm launched Project Trillium, a suite of machine learning (ML) solutions that included an ML processor , an object detection processor , and an open-source software framework. The Project Trillium products aimed to enable low-power and high-performance ML applications on edge devices.
In 2018, Arm unveiled the Cortex-A76 processor , which offered a 35% performance boost over its predecessor. The Cortex-A76 was used in devices such as the OnePlus 7T, the Huawei MateBook D14, and the Acer Chromebook Spin 13.
In 2019, Arm announced the Cortex-A77 processor , which improved on its predecessor with a higher clock speed, a larger cache, and better branch prediction . The Cortex-A77 was used in devices such as the Samsung Galaxy S20, the Asus ROG Phone II, and the Lenovo Yoga C940.
In 2020, Arm introduced the Cortex-X1 processor , which was its most powerful CPU design to date. The Cortex-X1 was designed to deliver peak performance for premium device , such as flagship smartphones, laptops and gaming consoles. The Cortex-X1 was used in devices such as the Samsung Galaxy S21 Ultra, the Xiaomi Mi 11, and the Google Pixel 6.
In 2021, Arm launched the Cortex-A78C processor , which was optimized for high-performance computing (HPC) applications. The Cortex-A78C featured up to eight CPU cores , a larger L3 cache, and support for ECC memory. The Cortex-A78C was used in devices such as the Samsung Galaxy Book Pro, the HP Elite Folio , and the Acer Chromebook Spin 513.
Microchip
In 2022, Arm unveiled the Cortex-A710 processor, which was its first big core to support the Armv9 architecture. The Cortex-A710 offered a 30% energy efficiency improvement over its predecessor, as well as enhanced security and ML features. The Cortex-A710 was used in devices such as the OnePlus 10 Pro, the Huawei MatePad Pro 2, and the Microsoft Surface Laptop Studio.
In 2023, Arm announced the Immortalis GPU family , which was its next-generation graphics solution that included hardware-based ray-tracing and variable rate shading capabilities . The Immortalis GPUs aimed to deliver realistic and immersive graphics for gaming, VR and AR applications on mobile devices . The Immortalis GPUs were used in devices such as the Samsung Galaxy S22 Ultra , the Sony Xperia 1 IV, and the Oculus Quest 3.
Powerful world presence
Arm is a leading semiconductor and software design company that has revolutionized the computing industry with its innovative and efficient processor architectures. Arm’s processors power billions of devices across various domains, such as mobile, IoT, AI, HPC, and gaming. Arm has been at the forefront of technological advancements for over three decades, delivering performance, energy efficiency, and security to its customers and partners.
Arm is a subsidiary of SoftBank Group and has a massive global presence.
British microchip designing giant Arm has announced that it has filed paperwork to sell its shares in the U.S.
The Cambridge-based company, which designs chips for devices from smartphones to game consoles, plans to list on New York’s Nasdaq in September. The highly anticipated IPO in the U.S. comes after UK Prime Minister, failed to convince Arm to float in London or pursue a dual UK-U.S. listing.
Arm’s decision to list in New York rather than London has fuelled fears that the City is losing its competitiveness to Wall Street, where valuations are typically higher. SoftBank-owned chip designer Arm on 21st August 2023 disclosed a modest 1% fall in annual revenue as it made public the paperwork for a U.S. listing that is expected to be the year’s biggest initial public offering. The company is reportedly looking for a valuation of between $60bn (£47bn) to $70bn.
Arm was bought in 2016 by Japanese conglomerate Softbank in a deal worth £23.4bn. Prior to the takeover, it was listed in both London and New York for 18 years.
Companies that use ARM processors in their products
Some of the companies that use ARM processors include Apple, Qualcomm, Samsung, Broadcom, and Fujitsu. ARM technology is used in a wide range of devices, from smartphones to game consoles to supercomputers.
ARM
Arm is a British semiconductor and software design company that is known for its Arm processors, which are widely used in smartphones, tablets, laptops, and other devices. Arm was founded in 1990 as a joint venture between Acorn Computers, Apple Computer, and VLSI Technology. The company was originally called Advanced RISC Machines, but later changed its name to Arm Ltd in 1998.
