During Evercore ISI TMT conference, Intel announced that the company would continue to lose market share, with a possible bounce back in the coming years. According to the latest report, Intel's CEO Pat Gelsinger announced that he expects the company to continue to lose its market share to AMD as the competition has "too much momentum" going for it. AMD's Ryzen and EPYC processors continue to deliver power and efficiency performance figures, which drives customers towards the company. On the other hand, Intel expects a competing product, especially in the data center business with Sapphire Rapids Xeon processors, set to arrive in 2023. Pat Gelsinger noted, "Competition just has too much momentum, and we haven't executed well enough. So we expect that bottoming. The business will be growing, but we do expect that there continues to be some share losses. We're not keeping up with the overall TAM growth until we get later into '25 and '26 when we start regaining share, material share gains."

The only down years that are supposed to show a toll of solid competition are 2022 and 2023. As far as creating a bounceback, Intel targets 2025 and 2026. "Now, obviously, in 2024, we think we're competitive. 2025, we think we're back to unquestioned leadership with our transistors and process technology," noted CEO Gelsinger. Additionally, he had a say about the emerging Arm CPUs competing for the same server market share as Intel and AMD do so, stating that "Well, when we deliver the Forest product line, we deliver power performance leadership versus all Arm alternatives, as well. So now you go to a cloud service provider, and you say, 'Well, why would I go through that butt ugly, heavy software lift to an ARM architecture versus continuing on the x86 family?"

The world of semiconductor IP licensing is complex by nature. If you use a company's IP, you must agree to its licensing terms. Today, it is precisely those terms that are being breached in the event of Arm Ltd. filing a lawsuit against one of its biggest customers, Qualcomm. When Qualcomm acquired Nuvia Inc., regarded as one of the best CPU design teams in the industry, it transferred Arm-Nuvia license agreements as its own. It continued the development of Arm IP under Qualcomm's name. This is a standard restriction, as Arm's licensing prohibits these sorts of IP transfers among companies to protect the IP.

As the UK-headquartered company reports: "Because Qualcomm attempted to transfer Nuvia licenses without Arm's consent, which is a standard restriction under Arm's license agreements, Nuvia's licenses terminated in March 2022. Before and after that date, Arm made multiple good faith efforts to seek a resolution. In contrast, Qualcomm has breached the terms of the Arm license agreement by continuing development under the terminated licenses. Arm was left with no choice other than to bring this claim against Qualcomm and Nuvia to protect our IP, our business, and to ensure customers are able to access valid Arm-based products."

Microsoft is reportedly planning to merge its ARM-powered Surface Pro X brand into the main Surface Pro line starting with the upcoming Surface Pro 9. This would see the Surface Pro 9 being offered with the upcoming Microsoft SQ3 processor which is derived from the Snapdragon 8cx Gen3. Microsoft has previously announced a desktop ARM developer kit codenamed "Project Volterra" featuring the Snapdragon 8cx Gen3 SoC and a neural processing unit that should offer similar performance. The Surface Pro 9 is also expected to gain 5G connectivity when it is announced alongside updated Surface Studio, and Surface Laptop products in the coming weeks.

Microsoft is announcing the general availability of the latest Azure Virtual Machines featuring the Ampere Altra Arm-based processor. The new virtual machines will be generally available on September 1, and customers can now launch them in 10 Azure regions and multiple availability zones around the world. In addition, the Arm-based virtual machines can be included in Kubernetes clusters managed using Azure Kubernetes Service (AKS). This ability has been in preview and will be generally available over the coming weeks in all the regions that offer the new virtual machines.

Earlier this year, we launched the preview of the new general-purpose Dpsv5 and Dplsv5 and memory optimized Epsv5 Azure Virtual Machine series, built on the Ampere Altra processor. These new virtual machines have been engineered to efficiently run scale-out, cloud-native workloads. Since then, hundreds of customers have tested and experienced firsthand the excellent price-performance that the Arm architecture can provide for web and application servers, open-source databases, microservices, Java and.NET applications, gaming, media servers, and more. Starting today, all Azure customers can deploy these new virtual machines using the Azure portal, SDKs, API, PowerShell, and the command-line interface (CLI).

NVIDIA designed the Grace CPU, a processor in the classical sense, to replace the Intel Xeon or AMD EPYC processors it was having to cram into its pre-built HPC compute servers for serial-processing roles, and mainly because those half-a-dozen GPU HPC processors need to be interconnected by a CPU. The company studied the CPU-level limitations and bottlenecks not just with I/O, but also the machine-architecture, and realized its compute servers need a CPU purpose-built for the role, with an architecture that's heavily optimized for NVIDIA's APIs. This, the NVIDIA Grace CPU was born.

This is NVIDIA's first outing with a CPU with a processing footprint rivaling server processors from Intel and AMD. Built on the TSMC N4 (4 nm EUV) silicon fabrication process, it is a monolithic chip that's deployed standalone with an H100 HPC processor on a single board that NVIDIA calls a "Superchip." A board with a Grace and an H100, makes up a "Grace Hopper" Superchip. A board with two Grace CPUs makes a Grace CPU Superchip. Each Grace CPU contains a 900 GB/s switching fabric, a coherent interface, which has seven times the bandwidth of PCI-Express 5.0 x16. This is key to connecting the companion H100 processor, or neighboring Superchips on the node, with coherent memory access.

Qualcomm is a company well known for designing processors going inside a vast majority of smartphones. However, the San Diego company has been making attempts to break out of its vision to focus on smartphones and establish new markets where it could show its potential for efficient processor design. According to Bloomberg's insights, Qualcomm is planning to re-enter the server market and try again to compete in the now very diverse space. In 2014, Qualcomm announced that the company is developing an Arm ISA-based CPU that will target servers and be an excellent alternative for cloud service providers looking at efficient designs called Centriq. Later on, in November of 2017, the company announced the first CPU Centriq 2400, which had 48 custom Falkor cores, six-channel DDR4 memory, and 60 MB of L3 cache.

What happened later is that the changing management of the company slowly abandoned the project, and the Arm CPU market was a bit of a dead-end for many projects. However, in recent years, many companies began designing Arm processors, and now the market is ready for a player like Qualcomm to re-enter this space. With the acquisition of Nuvia Inc., which developed crazy fast CPU IPs under the leadership of industry veterans, these designs could soon see the light of the day. It is reported that Qualcomm is in talks with Amazon's AWS cloud division, which has agreed to take a look at Qualcomm's offerings.

GIGABYTE Technology, an industry leader in high-performance servers and workstations, today became the first-to-market with a dual-socket motherboard, MP72-HB0, that supports up to 256 Arm cores, making it ideal for cloud native workloads. Also, the launch includes two more servers, G242-P35 and G242-P36, to offer up to 120 TB of NVMe (Gen4) storage capacity paired with Ampere Altra or Ampere Altra Max processors. These GPU-centric servers and motherboard will quickly find a home with hyperscaler and cloud workloads. Altra Max processors have predictable high-performance by having a high core count CPU with one thread per core, 128 threads on a monolithic 128-core chip. Multi-socket support and a wealth of PCIe/CCIX lanes make the platform highly scalable. At the same time, there is industry-leading power efficiency/core, which is highly sought after by our customers.

New product alert! In January last year, we launched the $4 Raspberry Pi Pico, our first product built on silicon designed here at Raspberry Pi. At its heart is the RP2040 microcontroller, built on TSMC's 40 nm low-power process, and incorporating two 133 MHz Arm Cortex-M0+ cores, 264kB of on-chip SRAM, and our unique programmable I/O subsystem. Since launch, we've sold nearly two million Pico boards, and RP2040 has found its way into a huge number of third-party products. We always believed that RP2040 was a great fit for commercial and industrial applications, but the global semiconductor shortage has vastly accelerated adoption. With millions of units on hand today, and pipeline in place for tens of millions more, design engineers who have been let down by their current suppliers have a perfect excuse to experiment.

Fast cores, large memory, and flexible interfacing make RP2040 a natural building block for Internet of Things (IoT) applications. But Pico itself has one obvious missing feature for IoT: a method for connecting to the network. Now, this is about to change. Today, we're launching three new members of the Pico family. Raspberry Pi Pico W is priced at $6, and brings 802.11n wireless networking to the Pico platform, while retaining complete pin compatibility with its older sibling. Pico H ($5) and Pico WH ($7) add pre-populated headers, and our new 3-pin debug connector, to Pico and Pico W respectively. Pico H and Pico W are available today; Pico WH will follow in August.

Hewlett Packard Enterprise (NYSE: HPE) today announced that it is the first major server provider to deliver a new line of cloud-native compute solutions using processors from Ampere. The new HPE solutions provide service providers and enterprises embracing cloud-native development with an agile, extensible, and trusted compute foundation to drive innovation.

Available in Q3 2022, the new HPE ProLiant RL300 Gen11 server is the first in a series of HPE ProLiant RL Gen11 servers that deliver next-generation compute performance with higher power efficiency using Ampere Altra and Ampere Altra Max cloud-native processors.

This time last year, I wrote about how digital experiences had never been more important, from personal to business devices - they helped us stay connected and entertained at a time when we needed it most. Compute continues to define our experiences in the modern world, and now these experiences are becoming even more visual.

Smartphones are at the center of our connected lives. From gaming to productivity, through video calling, social media or virtual environments, it is the device that provides us the connection to everyone and everything, in real time. For developers, making these immersive real-time 3D experiences even more compelling and engaging requires more performance. Arm sets the standard for performance and efficient compute, and our latest suite of compute solutions for consumer devices will continue to raise the threshold of what's possible in the mobile market, shaping the visual experiences of tomorrow.

MediaTek today announced the Dimensity 9000+, an enhancement to the company's top-of-the-line 5G smartphone chipset. This new high-end offering delivers a boost in performance over the Dimensity 9000 to make the next generation of flagship smartphones even more powerful and efficient.

The new Dimensity 9000+ system-on-chip (SoC) integrates Arm's v9 CPU architecture with a 4 nm octa-core process, combining one ultra-Cortex-X2 core operating at up to 3.20 GHz (compared to 3.05 GHz with the Dimensity 9000) with three super Cortex-A710 cores and four efficiency Cortex-A510 cores. The advanced CPU architecture and Arm Mali-G710 MC10 graphics processor built into the new chipset provide more than a 5% boost in CPU performance and more than 10% improvement in GPU performance.

With the launch of Apple Silicon for Mac computers, Apple has established itself as a great user of the Arm instruction set. Starting with M1, the company released an entirely new family of products running Apple Silicon. Today, thanks to the research of Strategy Analytics company, we have information that Apple is capturing as much as 90% of the revenue share present in the Arm PC market. The Arm PC market is a tiny subset of the entire PC market, mainly equipped with one-off Windows-on-Arm devices, Chromebook PCs, and Apple Macs. With the naturally low prices of the remaining Arm PCs, Apple Arm PCs offer a relatively high price point and a much more incredible selection of products.

On the global scale, Arm PCs now account for 9% of the total PC market share, where x86 vendors are dominating the field. "Apple's M-series family of processors set the benchmark and gave Apple a 2-3-year lead over the rest of the Arm-based PC processor vendors. Qualcomm captured just 3% revenue share in the Arm-based notebook PC processor market in 2021 and lags Apple in CPU performance," said Sravan Kundojjala, Director of Handset Component Technologies service at Strategy Analytics. This points to a particular case of Apple's better product and feeding the demand with higher-performing processors. Qualcomm's acquisition of Nuvia should yield different results in the coming years, as the new IP is yet to appear in Qualcomm SoCs.

Apple M1 chips are a part of the Apple Silicon family that represents a new transition to Arm-based cores with new power and performance targets for Apple devices. A portion of building a processor is designing its security enclave, and today we have evidence that M1 processors got a new vulnerability. The PACMAN is a hardware attack that can bypass Pointer Authentication (PAC) on M1 processors. Security researchers took an existing concept of Spectre and its application in the x86 realm and now applied it to the Arm-based Apple silicon. PACMAN exploits a current software bug to perform pointer authentication bypass, which may lead to arbitrary code execution.

The vulnerability is a hardware/software co-design that exploits microarchitectural construction to execute arbitrary codes. PACMAN creates a PAC Oracle to check if a specific pointer matches its authentication. It must never crash if an incorrect guess is supplied and the attack brute-forces all the possible PAC values using the PAC Oracle. To suppress crashes, PAC Oracles are delivered speculatively. And to learn if the PAC value was correct, researchers used uArch side channeling. In the CPU resides translation lookaside buffers (TLBs), where PACMAN tries to load the pointer speculatively and verify success using the prime+probe technique. TLBs are filled with minimal addresses required to supply a particular TLB section. If any address is evicted from the TLB, it is likely a load success, and the bug can take over with a falsely authenticated memory address.

NVIDIA today announced that Taiwan's leading computer makers are set to release the first wave of systems powered by the NVIDIA Grace CPU Superchip and Grace Hopper Superchip for a wide range of workloads spanning digital twins, AI, high performance computing, cloud graphics and gaming. Dozens of server models from ASUS, Foxconn Industrial Internet, GIGABYTE, QCT, Supermicro and Wiwynn are expected starting in the first half of 2023. The Grace-powered systems will join x86 and other Arm-based servers to offer customers a broad range of choice for achieving high performance and efficiency in their data centers.

"A new type of data center is emerging—AI factories that process and refine mountains of data to produce intelligence—and NVIDIA is working closely with our Taiwan partners to build the systems that enable this transformation," said Ian Buck, vice president of Hyperscale and HPC at NVIDIA. "These new systems from our partners, powered by our Grace Superchips, will bring the power of accelerated computing to new markets and industries globally."

MediaTek today announced the Filogic 880 and Filogic 380 Wi-Fi 7 platform solutions for high-bandwidth applications in the operator, retail, enterprise and consumer electronics markets. This pair of chips will be among the first Wi-Fi 7 solutions to hit the market, allowing device makers to deliver cutting-edge products with the latest connectivity technology. Filogic 880 is a complete platform that combines a Wi-Fi 7 access point with a new advanced host processor solution to provide the industry's best router and gateway solution for operator, retail and enterprise markets. It offers a scalable architecture that can support up to penta-band 4x4 with a maximum speed of 36 Gbps. Filogic 380 is designed to bring Wi-Fi 7 connectivity to all client devices, including smartphone, tablet, TVs, notebooks, set-top boxes and OTT streaming devices. The chip's dual concurrent 2x2 capability will be optimized "out-of-the-box" for these devices as MediaTek also supplies the corresponding platform solutions. This helps to streamline the design process, maximize performance and accelerate time-to-market.

"Our wireless connectivity solutions are designed to deliver the fastest performance using the most advanced technologies, and represent MediaTek's commitment to drive Wi-Fi 7 adoption in a large number of new markets," said Alan Hsu, corporate vice president and general manager of the Intelligent Connectivity business at MediaTek. "With Filogic 880 and Filogic 380, our customers can deliver fast, reliable and always-on connected experiences to meet the industry's growing connectivity demands." MediaTek's Filogic 880 combines a Wi-Fi 7 access point with a powerful application processor and network processing unit (NPU) to support maximum Wi-Fi, Ethernet and packet processing performance. The chip offers a wide range of interfaces and peripherals, making it easy to customize designs for various end products and applications.

GIGABYTE Technology, an industry leader in high-performance servers and workstations, today announced a new supercharged, scalable server, G492-PD0, that supports an Ampere Altra Max or Altra processor with NVIDIA HGX A100 Tensor Core GPUs for the highest performance in cloud infrastructure, HPC, AI, and more. Leveraging Ampere's Altra Max CPU with a high core count, up to 128 Armv8.2 cores per socket with Arm's M1 core, the G492-PD0 delivers high performance efficiently and with minimized total cost of ownership.

GIGABYTE developed the G492-PD0 in response to a demand for high-performing platform choices beyond x86, namely the Arm-based processor from Ampere. This new G492 server was tailored to handle the performance of NVIDIA's baseboard accelerator without compromising or throttling CPU or GPU performance. This server joins the existing line of GIGABYTE G492 servers that support the NVIDIA HGX A100 8-GPU baseboard on the AMD EPYC platform (G492-ZL2, G492-ZD2, G492-ZD0) and Intel Xeon Scalable (G492-ID0).

Nintendo could release its next-generation handheld game console, a successor to the crazy-popular Switch, by 2024. This could be powered by an NVIDIA-sourced SoC. NVIDIA recently put up a job listing for a "Game Console Developer Tools Engineer," looking for talent designing the software development applications for the next-generation console. Nintendo is a long-standing customer of NVIDIA chips for its handheld consoles. The next-gen SoC could implement an NVIDIA-designed Arm CPU, and a highly efficient GPU based on the NVIDIA graphics architecture of the time.

Russia has been dealt yet another blow to its technological capabilities, as countries around the world engage in an economic war against the nation following its invasion of Ukraine back in February. The UK government has recently added two major chip-design Russian companies, Baikal Electronics and MCST (Moscow Center of SPARC Technologies) to its sanctioned entity list, meaning that these companies are no longer able to acquire ARM licensing for its homegrown chips. The move is just another in a series of coordinated (and uncoordinated) sanctions.

According to a note provided by the UK government, the aim of the sanctions is to curb Russia's access to devices that could support its war efforts in Ukraine while propping-up the country's degrading technological infrastructure. This effectively shuts the door for Russia to solicit manufacturing from foundries outside its own territory. And even if Russia elected to manufacture its ARM-based designs in-country, thus skirting sanction-respecting foundries, the country's chip manufacturing is currently thought to only be capable of manufacturing chips at a 90 nm process - which the Western world has moved on from since 2006 already.

According to the report from Reuters, SoftBank has managed to regain control of the Arm China branch office that went rogue under the chairman and CEO Allen Wu's leadership. Arm China is SoftBank's venture to operate Arm Ltd. business in the Chinese region. That means that Arm can use all the licensing and development done on the mainland with SoftBank's supervision and conduct business. However, that idea was tough to pull off when now ex-chairman/CEO Allen Wu decided not to give up his leadership role for almost two years, despite being fired in 2020.

Not everything is terrible, as the SoftBank operation managed to make some progress in getting back the control of the Arm China venture. The company reports that the Arm China board has voted to replace Allen Wu unanimously and appoint Dr. Renchen Liu alongside Eric Chen as two co-CEOs. Dr. Liu is a vice dean at the Research Institute of Tsinghua University in Shenzhen, and the agency in Shenzhen has registered him as the company leader and general manager. Eric Chen is a managing partner at the SoftBank Vision Fund, helping Dr. Liu with business operations. Later after this decision, Allen Wu posted a letter signed by 430 employees that stated that there were law enforcement errors in his replacement process and that he would continue to lead the company. It is a matter of time before Chinese authorities take this action a step further and see more details.

One of the largest cloud providers in China, Alibaba, has today announced a preview for a new instance powered by Yitian 710 processor. The new processor is a collection of Alibaba's efforts to develop a home-grown design capable of powering cloud instances and the infrastructure needed for it and its clients. Without much further ado, the Yitian 710 is based on Armv9 ISA and features 128 cores. Ramping up to 3.2 GHz, these cores are paired with eight-channel DDR5 memory to enable sufficient data transfer. In addition, the CPU supports 96 PCIe 5.0 lanes for IO with storage and accelerators. These are most likely custom designs, and we don't know if they are using a blueprint based on Arm's Neoverse. The CPU is manufactured at TSMC's facilities on 5 nm node and features 60 billion transistors.

Alibaba offers these processors as a part of their Elastic Compute Service (ECS) instance called g8m, where users can select 1/2/4/8/16/32/64/128 vCPUs, where each vCPU is equal to one CPU core physically. Alibaba is running this as a trial option and notes that users should not run production code on these instances, as they will disappear after two months. Only 100 instances are available for now, and they are based in Alibaba's Hangzhou zone in China. The company notes that instances based on Yitian 710 processors offer 100 percent higher efficiency than existing AMD/Intel solutions; however, they don't have any useful data to back it up. The Chinese cloud giant is likely trying to test and see if the home-grown hardware can satisfy the needs of its clients so that they can continue the path to self-sustainability.

It appears that 2022 might be the year WiFi 7, or 802.11be as it's also known, launches, despite the fact that the specification isn't set to be ratified until 2024. Broadcom has just unveiled its first set of WiFi 7 radios, as well as accompanying SoC and it looks like the new hardware will bring a big jump in performance under the right circumstances. Broadcom's first WiFi 7 SoC will be known as the BCM4916, which sports a quad core ARMv8 processor. It appears to be a custom Broadcom design that's capable of delivering up to 24 DMIPS of performance, but beyond mentioning that it has 1 MB L2 cache and 64 kB L1 cache, Broadcom doesn't go into any specifics.

The Arm based processor is backed up by what Broadcom calls a Dual Issue Runner Network Processor, or DI-XRDP, which is a packet processor that shuffles the data around between the wired and wireless network interfaces. There's also a crypto coprocessor and Broadcom has added support for Arm TrustZone and Secure Boot, with the prior having support for a trusted execution environment. What stands out with regards to the BCM4916 compared to most current router SoCs is the wide arrange of high-speed connectivity options. For starters, it has no less than four PCIe interfaces of unspecified generation, as well as a native 10 Gbps network interface and as well as the option for two more. In addition to this, it has a further option for a 2.5 Gbps Ethernet interface, as well as a typical four port Gigabit switch built in. A pair of USB 3.2 ports are also part of the chip, but these are likely to be 5 Gbps ports. The SoC also has support for DDR3 and DDR4 memory.

When NVIDIA announced its Grace CPU Superchip, the company officially showed its efforts of creating an HPC-oriented processor to compete with Intel and AMD. The Grace CPU Superchip combines two Grace CPU modules that use the NVLink-C2C technology to deliver 144 Arm v9 cores and 1 TB/s of memory bandwidth. Each core is Arm Neoverse N2 Perseus design, configured to achieve the highest throughput and bandwidth. As far as performance is concerned, the only detail NVIDIA provides on its website is the estimated SPECrate 2017_int_base score of over 740. Thanks to the colleges over at Tom's Hardware, we have another performance figure to look at.

NVIDIA has made a slide about comparison with Intel's Ice Lake server processors. One Grace CPU Superchip was compared to two Xeon Platinum 8360Y Ice Lake CPUs configured in a dual-socket server node. The Grace CPU Superchip outperformed the Ice Lake configuration by two times and provided 2.3 times the efficiency in WRF simulation. This HPC application is CPU-bound, allowing the new Grace CPU to show off. This is all thanks to the Arm v9 Neoverse N2 cores pairing efficiently with outstanding performance. NVIDIA made a graph showcasing all HPC applications running on Arm today, with many more to come, which you can see below. Remember that NVIDIA provides this information, so we have to wait for the 2023 launch to see it in action.

Today, Microsoft launches Azure Virtual Machines (VM) based on the Ampere Altra Cloud Native Processor. This marks an important milestone as developers can now take advantage of these modern high-performance VMs for their existing and greenfield applications. The Ampere Altra processor family leads in performance across a range of broadly deployed cloud workloads and is now making available the Arm architecture on Azure.

Industry leading performance and the most sustainable solution
Cloud users who have pushed the limits of legacy x86 architectures now have a high-performance compute alternative that scales up in a linear fashion and delivers predictable performance even at full utilization. For example, Ampere Altra VMs outperform equivalently sized Intel and AMD instances from the same generation by 39% and 47%, respectively.* In addition to being the high-performance choice, Ampere Altra processors are extremely power efficient, directly reducing users' overall carbon footprint.

Rather surprisingly, SK Hynix has said that it's interested in being part of a consortium to purchase Arm, during its annual shareholders meeting. Although several companies have voiced their interest in Arm, such as Qualcomm and even Intel, SK Hynix hasn't even been on the radar until now. However, the interesting part here is the bit about a consortium and Park Jung-ho, vice chairman and CEO of SK Hynix is quoted saying "I don't believe Arm is a company that could be bought by one company," by the Yonhap News Agency.

Although SK Hynix has only recently started working on the idea of purchasing Arm, most likely due to NVIDIA's deal to purchase Arm from SoftBank falling through, it looks like SK Hynix have a very different plan to that of NVIDIA. Park continued "It doesn't have to be buying a majority of its shares to be able to control the company," suggesting that even if SoftBank was to list Arm on the stock market as a standalone company, SK Hynix could try to take control of a majority of its shares. However, the mention of the consortium is something other industry players have proposed and it seems like the most sensible way to move forward, without causing negative effects for the Arm licensees in the long term.

NVIDIA has today announced its Grace CPU Superchip, a monstrous design focused on heavy HPC and AI processing workloads. Previously, team green has teased an in-house developed CPU that is supposed to go into servers and create an entirely new segment for the company. Today, we got a more detailed look at the plan with the Grace CPU Superchip. The Superchip package represents a package of two Grace processors, each containing 72 cores. These cores are based on Arm v9 in structure set architecture iteration and two CPUs total for 144 cores in the Superchip module. These cores are surrounded by a now unknown amount of LPDDR5x with ECC memory, running at 1 TB/s total bandwidth.

NVIDIA Grace CPU Superchip uses the NVLink-C2C cache coherent interconnect, which delivers 900 GB/s bandwidth, seven times more than the PCIe 5.0 protocol. The company targets two-fold performance per Watt improvement over today's CPUs and wants to bring efficiency and performance together. We have some preliminary benchmark information provided by NVIDIA. In the SPECrate2017_int_base integer benchmark, the Grace CPU Superchip scores over 740 points, which is just the simulation for now. This means that the performance target is not finalized yet, teasing a higher number in the future. The company expects to ship the Grace CPU Superchip in the first half of 2023, with an already supported ecosystem of software, including NVIDIA RTX, HPC, NVIDIA AI, and NVIDIA Omniverse software stacks and platforms.