Written by: Techub News Compilation

Introduction

In January 2023, during CES 2023, AMD Chairman and CEO Lisa Su delivered a keynote address on the main stage in Las Vegas, USA. This marked the first full return to in-person CES after the pandemic, and Lisa Su's appearance signified AMD's important voice as a semiconductor giant at a critical juncture in industry recovery and transformation. During the speech, Lisa Su announced several heavyweight new products covering mobile, desktop, data center, and adaptive computing fields. Furthermore, through live dialogues with partners and guests including Microsoft, HP, Lenovo, Intuitive Surgical, Magic Leap, and former NASA astronauts, she elaborated on how AMD technology empowers a wide range of scenarios from personal computing to space exploration.

Summary

• Launched the new Ryzen 7040 series mobile processors, integrating a dedicated AI engine "Ryzen AI" for the first time, ushering in a new era of AI PCs.
• Unveiled the Ryzen 7000X3D series desktop processors featuring 3D V-Cache technology, providing extreme caching performance for gaming and content creation.
• Previewed the Alveo V70, a data center-level AI inference accelerator, and the first APU-style data center processor Instinct MI300, aimed at addressing the explosive demand for AI and HPC with higher energy efficiency.
• Emphasized deep collaboration with ecosystem partners such as Microsoft, HP, and Lenovo to jointly optimize the user experience from chips to systems, especially in AI, hybrid work, and gaming.
• Showcased the irreplaceable role of adaptive computing (FPGA) in critical mission fields such as medical robotics and space exploration, highlighting the social value of technology.

AI for Everyone: From Ryzen AI to Windows Ecosystem

Lisa Su started by pointing out that AI is the most important macro trend in future technology. To fully unleash its potential, AI capability must be spread across the entire device spectrum from smart terminals, edge, to cloud. This requires diversified computing engines—CPU, GPU, and adaptive accelerators, and AMD is one of the few companies that can provide all these engines.

The first heavyweight product announced in the speech was the new Ryzen 7040 series mobile processor (codename "Phoenix"). It is based on a 4nm process and contains over 25 billion transistors. Its standout feature is the integration of a dedicated AI engine derived from the acquired Xilinx technology—the first application of the AMD XDNA™ architecture, named “Ryzen AI.” This makes the Ryzen 7040 the industry's first x86 mobile processor with a dedicated AI engine, capable of processing four AI streams simultaneously, delivering up to 12 trillion AI operations per second. Performance comparisons presented by Lisa Su showed that, under AI workloads, the Ryzen 7040 outperforms Apple's M2 by 20%, while x86 competitors have yet to integrate a dedicated AI engine.

Lisa Su emphasized that Ryzen AI will unlock a new user experience, such as more realistic video conferencing (enhanced audio and video), more efficient content creation, more real-time gaming performance, and innovative security monitoring methods. The first laptops featuring Ryzen 7040 will be launched in March 2023, with over 250 designs expected to be available throughout the year.

To further explain the ecological value of AI PCs, Lisa Su invited Microsoft Executive Vice President and Chief Product Officer Panos Panay to the stage. Panos Panay praised AMD as a longtime leader in computing technology and pointed out that “AI is the defining technology of our era,” marking a turning point where both cloud and edge are becoming smarter. He particularly cited Windows 11’s “Studio Effects” as an example, such as background blur, eye contact, automatic framing, and noise suppression—features requiring trillions of operations per second that can now run directly on the Ryzen AI engine, using almost no CPU and GPU resources, with power consumption of only a few hundred milliwatts, achieving a balance of high-performance AI and long battery life.

Looking ahead, Panos Panay predicted that AI will completely reshape how users operate on Windows. While large generative models (language, code, images) are powerful, they are computationally intensive and require an operating system that can blur the lines between cloud and edge, along with chips that are in the right places. He hinted that AMD and Microsoft's collaboration is aimed at this, with next-generation products promising immense inspiration.

Redefining Work and Entertainment: Hybrid Work and High-Performance Gaming

The pandemic has changed work modes, making hybrid work the norm, but it has also exposed issues with insufficient technology readiness. Lisa Su cited data indicating that over 80% of employees prefer the flexibility of working from home, but over half feel connectivity issues hinder career development. She believes that deep collaboration in hardware, software, and system design is needed to make hybrid work seamless for both users and IT departments.

HP President and CEO Enrique Lores took the stage as a guest, asserting that we are currently experiencing “the biggest transformation in work styles ever seen” and predicting that in five years, 50% of the American workforce will be freelancers. This shift brings opportunities for disruption and innovation. Enrique Lores used the HP Dragonfly Pro laptop launched that morning as an example to illustrate the value of deep collaboration with AMD. This product targets freelancers, achieving a 40% higher PassMark score than Apple M1 through the integration of Ryzen processors, along with long battery life, dedicated support services, and HP Presence smart audio and video technology. He disclosed that a series of new products based on AMD Ryzen 7000 technology will be launched in Spring 2023, and expressed high expectations for future innovative collaborations.

Gaming is in AMD's DNA. Lisa Su announced the launch of the Ryzen 7045HX series mobile processors designed for performance-hungry gamers and creators, which is the first mobile CPU based on a “chiplet” design, featuring up to 16 Zen 4 cores with a boost frequency of 5.4 GHz. Meanwhile, AMD also released the Radeon RX 7600M XT mobile GPU based on RDNA 3 architecture. In terms of performance, the new processors deliver an average frame rate in 1080p gaming that is 24% higher than the previous generation, and over 50% faster in creative applications compared to competitors.

Lenovo Group Executive Vice President and President Matt Zielinski took the stage to introduce the collaborative outcomes in the gaming field. He announced that Lenovo is about to release the new Legion Pro series gaming laptops powered by the Ryzen 7045 HX series processors, making these Lenovo's most powerful AMD gaming laptops ever. Matt Zielinski specifically mentioned that AI is key to enhancing gaming experiences, as the new Ryzen 7000 series works in conjunction with the AI engine within Lenovo systems to dynamically adjust the CPU for up to a 5% performance boost. In 2023, Lenovo will launch over 50 models of laptops, desktops, and workstations powered by AMD.

On the desktop front, Lisa Su brought long-awaited news for gamers: the expansion of 3D V-Cache technology to the Ryzen 7000 series. The new Ryzen 7 7800X3D, Ryzen 9 7900X3D, and flagship Ryzen 9 7950X3D processors feature stacked cache, significantly enhancing gaming performance. Among them, the 16-core Ryzen 9 7950X3D boasts 144MB of cache, continuously delivering higher frame rates in 1080p gaming than competitors, earning it the title of the ultimate processor for gaming and creation. These new products will be available in February 2023. Lisa Su also announced that the popular game "Star Wars Jedi: Survivor" will be bundled with select Ryzen 7000 series CPUs.

Empowering Life and Exploration: The Profound Impact of Adaptive Computing

Through the acquisition of Xilinx, AMD has gained leadership in the adaptive computing (FPGA) space. Lisa Su pointed out that the hardware of such chips can be repeatedly modified and optimized for specific tasks and even add new functions at runtime, which is crucial in fields like automotive, industrial, aerospace, and medical.

Bob DeSantis, Executive Vice President and Chief Product Officer of Intuitive Surgical, took the stage to introduce the da Vinci Surgical Robot System, which has served over 11 million patients. Bob DeSantis explained that AMD’s adaptive computing technology is essential for achieving precise motion control, visualization processing and enhancement, as well as safety mechanisms for the robotic arms. Their multi-generation scalable computing platform assists Intuitive Surgical in integrating the same components across different products and programming to adapt to specific needs. He particularly mentioned that their Ion platform, utilizing robot-assisted bronchoscopy technology, is capable of deep lung biopsy, which is expected to significantly reduce the diagnostic time for lung cancer from months or even years, enabling earlier treatment.

In the field of medical imaging, AMD introduced the new Vitis medical imaging library to help equipment manufacturers achieve high-resolution imaging at over 1000 frames per second. The speech also showcased through video how companies like Clarius leverage AMD's adaptive technology to bring portable, AI-driven ultrasound devices to resource-limited remote areas, changing the accessibility of healthcare.

Peggy Johnson, CEO of augmented reality company Magic Leap, explained the differences between AR and VR, and talked about how customized processors developed in collaboration with AMD empower their enterprise-grade AR headset Magic Leap 2. This device features industry-leading optical stacks and dynamic light control and has received 60601 certification, allowing it to be used in operating rooms. Partner SentiAR's technology enables doctors to see a 3D real-time map of the patient's heart and catheter positions during surgery through Magic Leap 2, which Peggy Johnson described as a "game-changing" application.

Lastly, former NASA astronaut Dr. Kathy Coleman shared insights from an explorer's perspective regarding the key role of adaptive computing in space missions. Using Mars Perseverance Rover as an example, she illustrated how adaptive technology is used for navigation on the Martian surface, object recognition, terrain avoidance, and soil and rock sample analysis. In the recent Artemis 1 lunar mission, adaptive technology was utilized for engine controllers, displays, and command systems. Kathy Coleman emphasized that space missions are lengthy, and the ability to modify instrument functions through reprogramming after launch and to take advantage of new innovations is immensely valuable. She encouraged the industry to tell their stories to spark the next generation's interest in STEM, as the future moon and Mars explorers might be among today's youth.

A Sustainable Future: Energy Efficiency Revolution in Data Centers and Exascale Computing

The closing part of the speech returned to the cornerstone that drives the global digital world—data centers. Lisa Su pointed out that while meeting the unlimited computing power demands of high-performance computing and AI, it must be done in the most sustainable manner. She first reviewed the fourth-generation EPYC (Milan) processors released last November, whose energy efficiency can be up to 2.6 times higher than competitors. She calculated that a typical rack using competitors' servers would require 15 units to achieve the same computing power that can be achieved with just 5 units of the fourth-generation EPYC. If EPYC were adopted in approximately 15 million servers deployed worldwide last year, it could save 52 billion kilowatt-hours of electricity and avoid 26 million tons of CO2 emissions, equivalent to the carbon that would require more trees than the area of all national forests in California to absorb.

In response to the explosive demand for AI inference in data centers, AMD previewed its first dedicated inference accelerator, Alveo V70. It is based on the same XDNA architecture as Ryzen AI but scaled up to server level, providing 400 trillion AI operations per second. In video analytics applications, the V70, at a power consumption of 75 watts, can offer 70% more street coverage (smart city), 72% more bed coverage (patient monitoring), and 80% more checkout lane coverage (smart retail) than competitors. This accelerator is now available for pre-order and will be launched in Spring 2023.

Finally, Lisa Su unveiled what is hailed as "the most complex chip ever manufactured by AMD"—Instinct MI300. This is the industry’s first integrated processor (APU) designed for data centers combining “CPU+GPU+Memory,” utilizing 3D stacking technology to integrate multiple 5nm GPU chiplets and 6nm CPU chiplets, equipped with 128GB HBM3 memory and more than 146 billion transistors. The MI300 is specifically designed for next-generation exascale supercomputing and AI, delivering 8 times the performance on AI workloads compared to the MI250X, which currently powers the world’s fastest supercomputer, with a 5 times boost in energy efficiency.

Lisa Su used the popular ChatGPT, which gained traction in March 2025, as an example to illustrate its significance: training such large language models typically requires thousands of GPUs, takes months, and consumes millions of dollars worth of electricity. The MI300 is expected to reduce training times from months to weeks and drastically lower energy costs. More importantly, it will support larger models, laying the groundwork for future more powerful and advanced AI services. The MI300 has been sampled to customers and is expected to be available in the second half of 2023.

In conclusion, Lisa Su reiterated that the semiconductor and technology industries are in the best of times, and AMD will work with partners and the entire ecosystem to tackle the major challenges facing the world.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。