Dell recently unveiled two new Copilot+ PCs powered by the Snapdragon X platform. These include the Dell XPS 13 with Snapdragon X Elite SoC and Dell Inspiron 14 Plus powered by the Snapdragon X Plus SoC. Focused on power, efficiency and bringing on-device AI and GenAI experiences to masses, the new PCs are Microsoft’s fresh attempt at finally making Windows on Arm dream a reality.
At the recent launch event in New Delhi, India, Ritesh Bendre, Global Content Manager at Counterpoint Research sat down with Atul Mehta, Senior Director and General Manager for India Consumer Channel at Dell Technologies to talk about new Copilot+ PCs. They discussed key features and AI experiences, target audience of the new PCs and much more.
The Interview
Key Takeaways from the Discussion
Dell’s definition of AI PCs:
AI PCs, compared to traditional PCs, feature an NPU (Neural Processing Unit) alongside CPU and GPU to offload AI related tasks.
These PCs also come with a dedicated Copilot button for quick access.
The Copilot+ PCs also offer up to 27 hours of battery life for the XPS 13, and up to 21 hours for the Inspiron 14 Plus.
Differentiating AI features of Dell Copilot+ PCs:
These PCs come with AI features like:
Live captions with translation in up to 44 different languages, all happening on-device.
Windows Studio Effects to add background blur to videos, auto-framing and more.
Cocreation tools for AI generated images and more.
Dell’s target audience:
The Copilot+ PCs are designed to cater to a wide range of users, including:
Professionals.
Gamers.
Content creators.
Students and programmers.
Atul also highlighted that some of key factors users consider when choosing a laptop include multitasking, workload and processing requirements.
Positioning and future of AI PCs as per Dell:
In terms of positioning, Dell acknowledges the premiumization trend in the PC industry, and these laptops are designed for premium users.
Talking about the future of AI PCs, Mr. Mehta highlighted that growth in AI use cases over time with also fuel the explosion of these AI devices.
Dell is excited to be one of the leading vendors in the AI PC market, and also across other categories such as all-in-one PCs, cloud solutions and data centers.
Analyst Takeaways:
Battery life is one of the key-selling points for these Windows on ARM, Copilot+ PCs, and with Dell claiming up to 27 hours of backup on XPS 13, we will be looking forward to test and see how it actually performs in everyday usage.
Compatibility with x86 apps has been a challenge for traditional Windows on ARM PCs, however, more compatible hygeine apps have been announced this time around where app developers have worked with Qualcomm and Microsoft to drive native development as well as run well on the Prism emulator. The stability and growth of apps will be something which will drive long-term success, for Arm based AI PCs.
While Apple has been successful in completely transitioning away from the Intel x86-based compute platforms to ARM based Apple Silicon, it remains to be seen how long does it take for the ARM based Copilot+ PCs to gain considerable share in the x86 dominated PC market.
Copilot+ PCs promises a step change in Windows PCs segment by bringing exceptional battery life and cutting-edge AI capabilities. It will be key to see how Microsoft, Qualcomm and OEMs together communicate these benefits to the average consumers and enterprises through channels strongly influenced by Intel.
AMD and Intel with their upcoming CoPilot+ PC centric platforms will make a strong play to defend their market share slip away from Qualcomm Snapdragon X based Windows on ARM AI PCs.
We forecast AI laptops will grow to over 70% of total PC shipments by 2027.
Global PC shipments grew 3.1% YoY in Q2 2024; AI PCs remain the most popular topic in PC industry.
Apple is waiting for the M4 MacBook launch to regain shipment momentum.
Arm-based AI laptops will be another driver for Arm PCs and will capture more than 25% market share in 2027.
We reiterate a ~3% YoY shipment growth forecast for 2024, driven by AI PC and replacement demand.
Seoul, Beijing, Buenos Aires, Hong Kong, London, New Delhi, San Diego, Taipei, Tokyo – July 16, 2024
Global PC shipments grew 3.1% YoY in Q2 2024 to reach 62.5 million PC device units, the market’s second consecutive quarterly YoY growth in shipments, according to Counterpoint’s preliminary data. The growth, which was driven by a moderate replacement cycle and a low base YoY, highlights the global PC market’s continued recovery momentum in 2024. AI PCs remain the most popular topic in the PC industry. Laptops with Qualcomm’s Snapdragon X platform have begun to trigger the first wave of the AI laptop boom. The upcoming AMD Ryzen AI 300 and Intel Lunar Lake in the second half of Q3 2024 will further drive the popularity of AI PCs.
Despite a lull in demand in China, Lenovo remained the leader among PC vendors, leaving the June quarter with a 23.6% market share and ~4% YoY shipment growth. HP and Dell were relatively flat in the quarter, supported by order recovery in North America. HP had a 21% market share in Q2 2024, while Dell had 16%. Apple remained quiet in the latest quarter with an 8.5% market share. The company is waiting for the launch of the next-generation M4-based MacBook to regain shipment momentum in late 2024. Overall, we see Q2 2024 brand rankings and shares largely unchanged.
AI laptops will expand the AI PC market in H2 2024, driven especially by surging Windows on Arm AI devices at the end of Q2 2024. These Qualcomm Snapdragon-based solutions feature better battery efficiency than the dominating x86 camp. In our view, MediaTek and NVIDIA’s partnership on next-generation Arm-based AI laptop solutions will also be another driver to Arm PC market share expansion in 2025-2026 at the expense of x86 vendors. We project Arm PCs to capture more than 25% market share in 2027.
Commenting on AI PC development, Senior Analyst William Li said, “H2 2024 will be the battlefield for AI PCs. Vendors are paving the way for the proliferation of AI PCs and the rise of a brand-new technology generation. However, due to the higher selling price of AI PCs, we believe enterprise (commercial) demand will run ahead of the mainstream consumer sector, although we may not see the most rapid adoption path in the first year of AI PCs. As more AI use cases and solutions are developed and applied to PCs, we will see the deployment accelerate in the future. Benefitting from AI PC demand and the replacement cycle after COVID-19, we reiterate our view of ~3% PC shipment growth in 2024.”
Background
Counterpoint Technology Market Research is a global research firm specializing in products in the TMT (technology, media, and telecom) industry. It services major technology and financial firms with a mix of monthly reports, customized projects, and detailed analyses of the mobile and technology markets. Its key analysts are seasoned experts in the high-tech industry.
GenAI has become a top priority for companies and Qualcomm is pushing the envelope by developing solutions for on-device AI PCs. At Collision Toronto 2024, Emily Herbert, Research Analyst at Counterpoint Research, had a quick chat with Qualcomm’s Durga Malladi, SVP and GM of technology planning and edge solutions. He spoke about the new Snapdragon X Elite SoC line that allows on-device GenAI features to run directly on PCs with companies like Lenovo and HP that have already launched AI PCs. Some applications include live transcription, new studio effects using the web camera and Windows Studio Effects, and ‘paint’ to bring drawn images to life.
The interview:
Key takeaways from the discussion:
How Qualcomm defines AI PCs:
• According to Qualcomm, AI applications are critical to define AI PCs as they provide a personalized experience for the user with the privacy and security of keeping data stored in the device itself.
Benefits of on-device AI:
• Privacy and security are major benefits for running on-device AI as data stays on the device as opposed to cloud processing.
• With on-device AI, users can access these features without an internet connection, including scenarios like on transatlantic flights or in rural areas where the device may be used in the enterprise space with unreliable connectivity.
Qualcomm’s efforts in helping developers push on-device GenAI:
• Qualcomm AI Hub is a great tool to help developers access pre-trained models, write their applications, and test them remotely on real devices.
• Developers can now take the use cases to their desecration.
• People who did not have access to this technology across various fields like medicine, logistics, agriculture, and various other sectors can now use AI in ways that tech developers may not have been able to think of as the technology becomes democratized into the wider enterprise and consumer space.
Analyst takeaways:
• On-device GenAI has seen an excellent push within the enterprise space and now we are seeing real-life applications for the average consumer. The initial integration is expected to be slow as people learn about this technology and begin integrating it into day-to-day life. Increased knowledge and advertising in the market can help with wider consumer penetration.
• Now is a good time for on-device GenAI as hardware is reaching a point where it can support these higher processing applications and still reach the consumer price band, granted it is still within that ‘ultra-premium’ price point.
• The ultimate immersion of these AI technologies across all consumer segments will become easier and more cost-effective when the devices can be manufactured for and sold to more average users.
The HP Envy x360 14 is powered by an Intel Core Ultra 7 155U processor with a dedicated NPU for AI tasks.
The laptop also has a dedicated Copilot button for quick access to Microsoft’s AI assistant.
The 2-in-1 laptop features a 2.8K OLED display with a 120Hz refresh rate and IMAX certification.
AI is turning into a ubiquitous phenomenon in the world of technology, especially smartphones and PCs. Intel kick-started the AI PC era with its new Meteor Lake processors in early 2024. HP, along with other PC makers, is one of the first to launch AI PC laptops powered by Intel’s Core Ultra processors. But what exactly is an AI PC and why should one care? What sets them apart from previous PC generations?
Along with the CPU and GPU, AI PCs also have a dedicated Neural Processing Unit (NPU). This NPU helps in offloading some AI tasks from the CPU and GPU, helping save battery life. Since we are still in the early stages of this technology, the definition of AI PC will keep changing as things progress. Microsoft, at Build 2024, announced its Copilot+-powered AI PCs. According to Microsoft, a PC needs at least 40 trillion operations per second (TOPS) of AI performance to qualify as an AI PC.
Based on this, we may have entry-level GenAI basic PCs to the most advanced ones with top-notch configurations.
The HP Envy X360 14 is a 2-in-1 laptop powered by an Intel Core 7 Ultra 155U processor. However, it has a Meteor Lake architecture, which is not as advanced as Intel’s upcoming Lunar Lake architecture, that can deliver up to 120 TOPS AI performance. The HP Envy with Meteor Lake processor qualifies as a basic GenAI PC and, includes AI features that enhance the overall user experience from camera to audio, battery life to display, and much more.
Commenting on the growth momentum of AI PCs, Senior Analyst William Li said, “Microsoft’s announcement at the Build 2024 developer conference reaffirmed AI PC momentum in 2024, especially in the second half. Qualcomm’s first wave of AI laptops will enter the market in the upcoming month, while Intel and AMD are preparing for their next-generation AI solutions for PCs. Intel entered the AI PC market at the beginning of the year and is already working with PC partners to deliver a variety of laptops with the Meteor Lake platform, including the HP Envy x360 14. We believe Intel will continue to dominate the PC market in terms of shipments while Windows on Arm will also be in the spotlight in H2 2024. Overall, we believe three out of four laptops sold in 2027 will be AI laptops with advanced GenAI capabilities.”
We have been testing the HP Envy x360 14 for about three weeks to explore its AI features and see how the 2-in-1 laptop, priced at around $1,550, fits in the hybrid work scenario. Below is our detailed analysis.
WATCH: HP Envy x360 14 Review
2-in-1 form factor with a look and feel that evokes envy
At the launch event, HP mentioned that the Envy x360 14 is designed for creative professionals and freelancers. The laptop features a sleek and slim form factor, weighing just 1.39kgs, making it easier to carry around. The 2-in-1 design with a 360-degree hinge also makes it easier for different use cases. It can be used in the tent mode for watching content or can be folded all the way to be used as a tablet.
On the connectivity front, you get a USB Type-A 3.1 port on the left, an HDMI out, and two USB Type-C Thunderbolt 4 ports. On the right, you get a USB Type-A 3.1 port and a 3.5mm headphone jack. Unfortunately, the laptop misses out on an SD or microSD card slot.
The chiclet keyboard keys have enough travel to offer a good typing experience. The keyboard is backlit, making it easier to work and type in the dark. The keyboard also has a dedicated emoji button among the function keys at the top. The keyboard has a dedicated Microsoft Copilot button, located beside the arrow keys, indicating that the PC also has a dedicated NPU.
Using the dedicated Copilot button, you can quickly invoke the assistant to ask questions, to generate an image or email for you. You can also ask for recipes, travel ideas and itineraries, ask to change wallpaper, turn volume/brightness levels high and low, and much more. You can even use your voice to ask these questions, but unlike Siri on Macs which gets invoked by voice, here you will have to press the mic icon before you ask a question.
Great display with touchscreen functionality and stylus support
The Envy x360 14 features a 14-inch 2.8K resolution (2880 x 1800 pixels) OLED touchscreen display. With 500 nits of peak HDR brightness, the screen supports an adaptive refresh rate of 48Hz to 120Hz, which can automatically be adjusted based on the on-screen content.
The OLED screen is IMAX-certified. It offers great viewing angles, good contrast, vibrant color reproduction and stands out particularly when watching immersive HDR content on Netflix. The touchscreen responsiveness is smooth. We did not experience any issues when scrolling through web pages, performing pinch-to-zoom gestures or playing games like Asphalt 9: Legends.
The screen also supports the HP stylus MPP2.0 Tilt Pen. The stylus is easier to hold and is just like a regular pencil in appearance. There is a small cover at the end of the stylus, sliding which reveals the Type-C charging port.
In terms of usage, the stylus has good input precision, something that creative designers will find handy. We tried sketching using MS Paint, and it worked well. The screen supports palm rejection, but the optimization could have been better.
HP Envy x360 (14-fc0100TU) specifications
Under the hood, the HP Envy x360 14 is powered by Intel’s Core Ultra 7 155U processor. Built on the Intel 4 Node (7nm fabrication), this is our first time testing a Meteor Lake SoC. The processor has 12 cores and 14 threads:
• Two performance cores (P-cores) with a base frequency of 1.7GHz and a max turbo frequency of 4.8GHz.
• Eight efficiency cores (E-cores) with a base frequency of 1.2GHz and a max turbo frequency of 3.8GHz.
• Two low-power efficiency cores (LPE-cores) with a base frequency of 700MHz and a max turbo frequency of 2.1GHz.
The processor has a base power of 15W, which can go up to 57W on turbo mode. On the graphics side, the laptop comes with Intel Graphics with four X-cores and dynamic frequency of up to 1.95GHz. The integrated graphics card supports ray tracing, and a maximum integrated resolution of 3840×2400 pixels with a 120Hz refresh rate.
One factor that differentiates Intel’s Meteor Lake disaggregated SoC architecture from its predecessors is the NPU. Intel has included the Intel AI Boost NPU with a max frequency of up to 1.4GHz. It supports AI software frameworks like OpenVINO, which we will talk about in detail in a bit.
Other specifications include 32GB of LPDDR5 RAM (onboard) and 1TB of PCIe Gen 4 NVMe SSD. Wireless connectivity options include Wi-Fi 6E and Bluetooth 5.3. The laptop comes with a 3-cell, 59Wh battery and a 65W Type-C charging adapter, which can charge the laptop from empty to 50% in about 30 minutes, and a full charge in roughly one hour 30 minutes.
The processor is also Intel Evo-certified, meaning it can offer a battery life of over nine hours, charges over a Type-C port, wakes up from sleep in less than a second, and includes Thunderbolt 4 and Wi-Fi 6E connectivity options.
Dedicated NPU, unlocking new AI PC experiences
We have been hearing and talking a lot about AI PCs, but what are some of their use cases, and how are they different from traditional PCs? Well, the HP Envy x360 14 is loaded with AI features, which we experienced while using the laptop.
Microsoft’s Windows 11 OS, which the laptop runs on, has Windows Studio Effects. This feature allows the user to access various effects on the laptop’s 5MP IR camera such as auto-framing, eye contact, and background effects. The workload of these AI effects is entirely on the NPU (rather than on CPU/GPU), which also helps in conserving battery life. The auto-framing uses AI to detect your face in the frame and keep you in focus.
The eye contact feature ensures that your eyes are looking into the camera, even if you are looking elsewhere, say into your phone. Lastly, the background effects separate the subject from the background while adding blur effects.
The AI also filters unwanted background noise when on voice or video calls. This will be highly beneficial for people who work in hybrid environments, be it offices, café, airports and more. The speakers, tuned by Poly Studio, are good enough for casual media consumption and video calls. The AI does all this work in the background to optimize the laptop’s battery life.
HP has also included a Presence Detection feature which uses the IR sensor and AI to detect when you are around (max distance four feet) and looking into the laptop. Once you look away from the screen for a few seconds, the screen brightness is reduced. It increases again after you start looking at the screen. When you walk away from the screen (for over two minutes), the laptop automatically gets locked. The AI automatically unlocks the laptop when you get back and Windows Hello authenticates your face. However, one thing to note is that to awaken the laptop a few minutes after it goes into sleep mode, you will have to press the power button.
Testing Intel’s OpenVINO AI models
To spread AI everywhere, Intel recently announced over 500 AI models optimized for Core Ultra processors, unlocking new experiences. We tried a couple of these OpenVINO AI effects on the popular open-source audio editing app, Audacity, and the open-source photo editing app, GIMP.
Audacity OpenVINO AI Effects
On the Audacity app, there are four key AI features – music separation, music suppression, music generation, and transcription. All these AI features work on-device and do not need an internet connection. Music separation is great for musicians as it allows editors to separate the vocals, drums and other instruments, and then remix the song. We tried it with a couple of songs, like Hotel California by Eagles, and it worked pretty well.
Music suppression is good for podcasters as it greatly helps to reduce the background noise, like fan noise, AC hum, and wind noise. Using music suppression, we recorded an audio clip with background noise, and the results were impressive.
With the music generation GenAI feature, you can generate original music based on a prompt. While it worked well, the prompt needs to be detailed to get the desired results. You get an option to use NPU as well, and you can see most of the load taken by the NPU, while keeping the CPU and GPU free. In our test, it took two minutes and 11 seconds to generate a track using the NPU.
Lastly, we also tested the audio transcription feature which uses the Whisper model to transcribe audio to text. For a 28-minute podcast audio file, it took around three minutes and 40 seconds to transcribe with decent output.
GIMP OpenVINO AI Effects
We used GIMP for two tasks – text-to-image generation using Stable Diffusion 1.4. The output was decent, and it took around 35 seconds to generate an image using the text.
It also supports semantic segmentation fully using the NPU, which took about seven seconds to separate the background from the subject. The edge detection was not that good, but since these key tasks are handled by the NPU, it unburdens the processor.
(Do note, that installation is not a straightforward process. There is no direct installer, and you will need to download resources from GitHub, install Python, and install the requisites using Command Prompt.)
Filmora AI Effects
Intel and Wondershare have partnered to optimize the processor for the popular video editing app, Filmora. The AI effects allow you to separate the subject (foreground) from the video and add background blur. You can also replace the background while keeping the person in the foreground.
There are other AI effects too, like noise cancellation, video effects like flares, neon lights, and more. All these AI effects make use of the NPU with Intel AI Boost, while also sharing resources with the CPU and GPU. In our test, adding some of these effects to a one-minute video and rendering took about six minutes, which is not bad.
LM Studio
Lastly, we downloaded LM Studio, an app that lets you download small language models (SLMs) and large language models (LLMs), such as Meta’s Llama3, Microsoft’s Phi-3 and Google’s Gemini, to run these AI chatbot models locally. We asked a bunch of questions like the best places for burgers, or top tourist places in different cities, and it offered good results. However, the token generation time taken for the final output was slow as the workload was taken by the GPU. There were no options to shift the workload to CPU or NPU.
Good performance, even better battery life
The Envy x360 is powered by the very competent Intel Core Ultra 7 155U processor and offers reliable performance for a wide range of day-to-day tasks. This includes web browsing, audio/video calls on platforms like Teams and Zoom, checking and responding to emails on Outlook app, and using apps like Word and PowerPoint.
Sure, the 155U is not as powerful as the Core Ultra 7 155H, but it does offer a good balance between performance and efficiency. We were pleasantly surprised by the steady frame rates when playing casual games like Left4Dead2 (hitting even 130fps at times), Rocket League (hitting around 40fps), and Counter-Strike 2 (hitting about 30fps). This is despite not having a discrete graphics card, or the latest Intel Arc graphics solution (as on 155H).
HP has also included the OMEN Gaming Hub, a customization tool for all your games to ensure better all-round gaming performance.
In terms of battery life, we got a little over seven hours of backup with usage that included writing blogs, web browsing and a few audio calls on Teams. With slightly heavy usage, including editing a couple of video reels and watching videos, we still managed to get around five hours of backup. With casual gaming, the laptop still managed close to two hours of backup. (Do note, throughout our heavy usage like gaming, video editing and testing AI tasks, the battery performance settings were set to high mode.)
Final thoughts
The HP Envy x360 14 is a compelling option for users who are looking for a powerful and versatile 2-in-1 laptop. With Intel’s latest Core Ultra 7 processor at the helm, along with a dedicated NPU, it offers innovative AI features thoughtfully designed for everyday use cases. From enhanced video conferencing experiences to handy features like presence detection and a dedicated Copilot button, NPU unlocks a new level of user experience.
All this paired with a great OLED display, good keyboard and long battery life positions the Envy x360 14 as a well-rounded AI PC for professionals, creators and casual users alike.
Qualcomm announced the Snapdragon X Elite SoC at the Snapdragon Summit in 2023, followed by the Snapdragon X Plus a couple of months ago. At Microsoft Build 2024, Qualcomm took the stage along with Microsoft to announce the first set of CoPilot+ PCs. At Computex 2024 in Taiwan earlier this week, Qualcomm shed more details on design wins, performances and partnerships. William Li, Senior Analyst at Counterpoint Research, sat down with Qualcomm’s Alex Katouzian, Group General Manager for Mobile, Compute and XR, to discuss Windows on Arm, the future of computing and more.
The interview
Key takeaways from the discussion
Snapdragon X platform co-designed with Microsoft:
• Qualcomm worked closely with Microsoft to optimize the Snapdragon X platform for Windows 11 CoPilot+ features.
• Snapdragon X can deliver high performance with low power consumption and is superior to traditional x86 solutions. Thus, it also helps extend the battery life on laptops.
Focus on AI to enable new experiences:
• AI is becoming pervasive across all Qualcomm products, smartphones, IoT devices, wearables, cars and XR.
• Qualcomm aims to offer intelligent edge solutions powered by AI across Snapdragon platforms.
• Qualcomm is leveraging its experience in mobile AI to bring similar capabilities to AI PCs.
• The Snapdragon X platform comes with a powerful NPU that enables advanced AI features, running both small language models (SLMs) and large language models (LLMs) on-device.
Beyond laptops:
• The Snapdragon X platform is designed for various form factors.
• It can power laptops, desktops, mini-PCs, tablets and even enterprise devices.
Snapdragon Seamless for multi-device connectivity:
• The Snapdragon platform comes with a low-power Bluetooth protocol called Snapdragon Seamless.
• It allows all devices – smartphones, laptop, tablets and TWS, to seamlessly interact with each other, and even enables wireless data sharing.
• The best part is that Snapdragon Seamless works across platforms and operating systems. This means one can easily connect an Android smartphone with a Windows PC and unlock seamless multi-device experiences.
Analyst takeaways:
• Snapdragon X is changing the way we interact with our PCs. Qualcomm claims to offer unparalleled computing performance, lower power consumption, and longer battery life which will further push the limits of PCs. The whole PC market has its eyes on Qualcomm’s Snapdragon X.
• Snapdragon leads not just in terms of computing performance but also performance per watt. It claims to offer 2.6x better performance per watt compared to the Apple M3 and a further 5.4x superior performance per watt compared to the Intel Core Ultra 7. Regarding thermal performance, it claims to offer a tremendous 5x better performance than the Intel Core Ultra 7 when running at lower thermals.
• Consumers use different types of mobile devices on a daily basis – Snapdragon can help improve the experience. Snapdragon Seamless allows users to easily move files between devices. Snapdragon Connect offers lightning-fast 5G and Wi-Fi 7, while Snapdragon Sound provides immersive and lossless audio. There is also Snapdragon Secure which enables chip-to-cloud security.
Computex Taipei 2024, Asia’s biggest tech event, has officially kicked off and Counterpoint’s analysts are on the ground to bring you the latest updates.
Day 1 was set in motion with keynote addresses from industry leaders such as NVIDIA CEO Jensen Huang, . Here are the standout announcements:
NVIDIA unveiled its latest AI advancements, promising faster, smarter tech.
AMD showcased new processors that redefine performance and efficiency.
Qualcomm revolutionizes the AI PC with Snapdragon X Elite.
Arm powers the AI future with 100 billion devices.
NVIDIA’s Ecosystem and AI Innovations
Comprehensive AI Ecosystem: NVIDIA’s keynote on the first day of Computex Taipei 2024 extended beyond GTC’s focus on AI innovations by highlighting the broader ecosystem that supports AI technology advancements. Taiwan’s AI Server Original Design Manufacturers (ODMs) play a crucial role as they serve as production partners and potential clients for NVIDIA’s advanced AI servers. The keynote showcased the company’s comprehensive ecosystem, which includes a thriving developer community, a vast array of optimized applications and a diverse portfolio of chips, systems, networking solutions and proprietary technologies.
Annual AI Accelerator Updates: NVIDIA CEO Jensen Huang emphasized the company’s commitment to continuous improvement in AI accelerators. The annual upgrades focus on reducing costs and energy requirements while adding new tools and software models. This approach ensures that NVIDIA remains at the forefront of high-performance AI computing. The keynote featured a virtual avatar and voice clone of CEO Jensen Huang, demonstrating the power of NVIDIA’s standalone GPUs in delivering high Trillion Operations Per Second (TOPS) performance, crucial for real-time rendering of highly realistic virtual avatars and voice cloning.
NVIDIA Inference Microservice (NIM): At the event, NVIDIA introduced the NVIDIA Inference Microservice (NIM), a pre-trained AI model integrating cloud-native technologies like CUDA, QDNN, TensorRT and Triton. NIM is designed to support corporate data centers and Cloud Service Providers, adding significant value to NVIDIA’s offerings. This new tool exemplifies NVIDIA’s focus on providing comprehensive solutions that cater to the evolving demands of AI technology.
AMD’s Breakthroughs in AI and High-performance Computing
Meeting NVIDIA’s Annual Release Cycle: AMD is keeping pace with NVIDIA’s annual release cycle with the introduction of the MI325X, which is set to launch in Q4 2024. This new release underscores AMD’s commitment to providing high-performance AI computing solutions that can compete head-to-head with NVIDIA’s offerings.
Highlighting its Presence in PCs and Laptops with Strix Point: AMD highlighted its presence in PCs and laptops with the new Strix Point processors, part of the third-generation Ryzen series. These processors, designed for ultra-thin and premium notebooks, combine Zen 5 CPU, RDNA 3.5 graphics and XDNA 2 NPU, delivering 50 TOPS of compute for AI experiences at low power.
Third-generation Ryzen AI Processors and Copilot+ PC: The third-generation Ryzen AI series processors significantly boost AI and computing performance. AMD’s latest NPU achieves 50 TOPS, with 5x the compute capacity and 2x the energy efficiency of its predecessor. The Copilot PC, powered by AMD’s new processors, integrates on-device AI for seamless functionality, enabling advanced features with robust performance and efficiency.
Qualcomm and the AI Revolution in PCs
The PC is Reborn: Qualcomm’s focus at Computex Taipei 2024 centered around AI PCs, partnerships and power efficiency. The messaging highlighted its commitment to advancing both processing power and energy efficiency, which is crucial for next-generation AI applications. The Snapdragon X Elite platform aims to transform how we use the PC by offering advanced AI capabilities. Qualcomm is committed to supporting the ecosystem’s shift to AI PCs, ensuring seamless user experiences.
Strategic Partnerships: Qualcomm‘s partnerships with major players like Microsoft, Dell, Lenovo and HP signify its strong entry into the PC industry. These collaborations are pivotal for Qualcomm’s strategy to establish a foothold in the competitive PC market, leveraging the strengths of its partners to drive innovation and adoption of AI-enabled PCs.
100 Billion AI Devices by the End of 2025 at Arm’s Reach
Accelerated Innovation Timeline: Arm CEO Rene Haas announced that by the end of 2025, over 100 billion Arm devices will be ready for AI applications. This milestone highlights Arm’s significant impact on the AI landscape and its commitment to facilitating developer creativity. The aggressive innovation timeline includes leveraging the latest advancements in technology to ensure rapid development and scalability of AI applications.
New CSS for Clients and Cortex X925: Arm introduced the new CSS for clients and the Cortex X925, promising major advancements in AI and security for next-gen PCs. These innovations, previously exclusive to data centers, are now poised to enhance the performance and efficiency of consumer devices.
Microsoft, along with partners, is set to completely redefine the PC experience by rearchitecting Windows to integrate generative AI-powered native CoPilot agents at every layer.
The stars are aligned for Microsoft to bring a step change to the PC experience with advancements in semiconductors driven by its partnership with Qualcomm.
Microsoft has announced CoPilot+, building upon its cloud-based CoPilot and CoPilot Pro to bring on-device native GenAI experiences powered by models (SLM, LLMs, LMs) running in the background.
The PC is one electronics category that is still called PC, unlike devices like phones, watches and TVs which have transitioned to smartphones, smartwatches and smart TVs. This has been possible because of years of research and development in software, AI and smart applications, making these devices disrupt the market. On the other hand, the PC has been relegated to the position of being a traditional productivity device with limited intelligence or “smartness”.
However, this year, the PC is going through a pivotal moment in becoming more intelligent. Microsoft, along with partners, is set to completely redefine the PC experience by rearchitecting Windows to integrate generative AI-powered native CoPilot agents at every layer, all of which are fully optimized with the silicon underneath. These generative AI-based applications will warrant advanced silicon in the form of compute as well as memory which is not only high performance, driving more than 40 TOPS computations powered by an NPU alongside CPU, GPU and advanced memory (min 16GB DRAM, 256GB SSD), but also highly power efficient.
Stars are aligned for Microsoft with right partners – Qualcomm and OpenAI
This is where the adoption of Arm-based silicon has been solving the power efficiency problems in a PC, as we have seen with the M-series Apple Silicon. However, Qualcomm has leapfrogged Apple, Intel and AMD to bring its leadership in low-power compute with Arm-based designs, along with years of experience and IP with AI Engine and its acquisition of Nuvia, to rearchitect a compute silicon for the next-generation AI PC or what Microsoft calls the CoPilot+ PC.
Source: Microsoft
The stars are aligned for Microsoft to bring a step change to the PC experience with advancements in semiconductors driven by its partnership with Qualcomm, which will help architect a completely differentiated, intelligent and efficient PC experience integrated with Microsoft’s own AI expertise and partnership with OpenAI to have a formidable answer to Apple’s Macs.
Qualcomm’s Snapdragon X Elite platforms, which were purposefully built for these next-generation AI PCs, are the only chipsets to power more than 20 designs across the top seven global PC OEMs at launch. Intel and AMD are behind and will follow up with more x86-based CoPilot+ designs later this year. But this is a significant win and entry into the PC space for Qualcomm, completely redefining the competitive landscape with a solid offering.
Source: Qualcomm
Unlocking powerful native intelligent experiences and more
Microsoft has announced CoPilot+, building upon its cloud-based CoPilot and CoPilot Pro to bring on-device native GenAI experiences powered by models (SLM, LLMs, LMs) running in the background. The data doesn’t leave the PC, boosting privacy and security. The CoPilot agents are built on OpenAI’s GPT4.
Some of the experiences that Microsoft has showcased are compelling on how GenAI can drive new use cases, automation and multiple assistants to boost productivity, creativity and security. Every Microsoft CoPilot+ PC is secured with the Pluton Security Processor out of the box and deeply integrated with personalized privacy controls. Some of the native CoPilot+ experiences Microsoft has showcased include:
Recall: This is a flagship CoPilot+ feature that leverages the “personal semantic index”, giving a PC “photographic memory” by remembering every instance of action, content and settings across any application, website or document stored on the device. This is done by more than 40 AI models always running in the background and taking snapshots of information to build a timeline and retrieve information quickly and intuitively when prompted.
This is a flagship GenAI feature showcasing the power of AI-powered semantic search and retrieval of information which is very personalized, private and intuitive, helping users save time and energy.
Source: Microsoft
Cocreate: This is another native GenAI-based CoPilot+ feature baked into specific applications such as Paint, where any ink strokes or text inputs can help generate images in near-real time, helping one brainstorm or express ideas, art and creativity. This is a glimpse of AI-powered photo editing and styling capabilities which when scaled with more context, intelligence and creativity can help generate great graphics and memories within seconds.
Source: Microsoft
Live captions: This is another meaningful GenAI feature of CoPilot+. It drives real-time live translations as the audio traverses through the PC across all the apps consistently. More than 40 languages can be translated into English subtitles and vice versa in a live or pre-recorded format and without needing the internet, with everything on the device being powered by the magical DSPs inside the NPU.
Growing native partner experiences
Several announcements have been made on native ARM64 experiences, apart from native Microsoft apps such as Chrome, Spotify, Zoom, WhatsApp, Blender, Affinity Suite and Slack. Many of them, such as Adobe Suite, CapCut, Cephable, Camo, DaVinci Resolve, djay Pro LiquidText and Luminar Neo, are leveraging the powerful 40 TOPS+ NPU to build multiple native GenAI experiences within their applications.
Microsoft has also announced Prism, a more powerful emulator for running native or x86 applications more efficiently and seamlessly. This should be the key moment for the Windows-on-Arm ecosystem to compete effectively with the incumbent Windows-on-x86’s dominance.
New Microsoft Surface portfolio includes CoPilot+ PCs
Microsoft rearchitected the new Surface and Surface Pro lineup powered by the Snapdragon X Elite platform as CoPilot+ PCs. A good refresh and many “first” specs.
Displays: 12.8” to 15” with the Surface Pro sporting an OLED HDR display for the first time, 120 Hz refresh rate, Dolby Vision IQ
Processors: Qualcomm Snapdragon X Elite and X Elite Plus
Connectivity: Wi-Fi7/5G, two USB4 ports supporting three external 4K displays
Accessories: New Surface Pro keyboard can be used in both attached and detached mode with CoPilot key and includes the new Surface Slim Pen and quiet haptic touchpad.
Source: Microsoft
AI PC outlook and opportunities
CoPilot+ fundamentally changes how users interact with PCs. It makes their experience more intelligent, contextual and meaningful. This positions Microsoft and its ecosystem uniquely in the market, particularly when compared to Apple.
As CoPilot+ expands to enterprise PCs, the GenAI models will not only boost productivity in core applications like Microsoft Office, Azure AI and Dynamics CRM but also within solutions from partners such as Adobe, Cognizant, IBM, ServiceNow and Amdocs.
By integrating CoPilot+, enterprises can expect improved workflows, smarter task management and enhanced collaboration, ultimately driving innovation and efficiency.
According to our AI PC Tracker, GenAI-capable laptop PCs will account for more than three in four laptop PCs shipped in 2027.
As a result, the semiconductor content value for the PC segment will also grow almost 2x by 2027.
The Microsoft announcements show how users will use PCs that are more intelligent, contextual and meaningful, taking creativity, productivity and usage to a different level.
Ahead of Apple’s WWDC event in June, this is a very significant differentiation for Microsoft and its ecosystem versus Apple. At this point, we believe Qualcomm should be the main growth driver for Arm PCs along with GenAI capabilities in the short-to-mid term.
AI PCs will revive the global laptop PC market’s replacement demand in the coming years.
We estimate that AI laptop PCs with the ability to run advanced GenAI applications will account for every three out of four laptop PCs sold in 2027.
Microsoft will play a pivotal role in driving the AI laptop PC ecosystem, while Apple will be a dark horse.
Intel and AMD will look to take the GenAI-capable compute mainstream next year to compete with more efficient Arm-based Qualcomm and Apple solutions in this AI race.
Seoul, Beijing, Buenos Aires, Fort Collins, Hong Kong, London, New Delhi, San Diego, Taipei – April 30, 2024
Personal computers (PCs) have been used as the major productivity device for several decades. But now we are entering a new era of PCs based on artificial intelligence (AI), thanks to the boom witnessed in generative AI (GenAI). We believe the inventory correction and demand weakness in the global PC market have already normalized, with the impacts from COVID-19 largely being factored in. All this has created a comparatively healthy backdrop for reshaping the PC industry. Counterpoint estimates that almost half a billion AI laptop PCs will be sold during the 2023-2027 period, with AI PCs reviving the replacement demand.
Counterpoint separates GenAI laptop PCs into three categories – AI basic laptop, AI-advanced laptop and AI-capable laptop – based on different levels of computational performance, corresponding use cases and the efficiency of computational performance. We believe AI basic laptops, which are already in the market, can perform basic AI tasks but not completely GenAI tasks and, starting this year, will be supplanted by more AI-advanced and AI-capable models with enough TOPS (tera operations per second) powered by NPU (neural processing unit) or GPU (graphics processing unit) to perform the advanced GenAI tasks really well.
Commenting on the GenAI adoption in PCs, Senior Analyst William Lisaid, “Counterpoint forecasts that the penetration of AI-advanced laptops, which will form a majority of the AI laptop PC segment, will pick up in the next two years as chip vendors scale the GenAI capability to mainstream tiers. With the proliferation of GenAI use cases at the edge or in the cloud or in a hybrid way in the next few years, GenAI will become a de facto and must-have capability in the PC segment. Building the right kind of tools and ecosystem will be crucial for this adoption. Players such as Qualcomm have already accelerated with partners such as Microsoft, Hugging Face and a set of advanced AI Stack toolsets, developer evangelism.”
Li added, “While the overall laptop PC market will register ~3% CAGR during 2023-2027, the AI laptop PC segment is bound to grow at 59% CAGR. The advanced semiconductor content from compute to memory will drive a positive ASP (average selling price) growth associated with newer GenAI capabilities, translating into more value for consumers than before. In addition, we estimate that AI laptop PCs with the ability to run advanced GenAI applications will account for every three out of four laptop PCs sold in 2027.”
The chip vendors are playing a major role in driving GenAI in PCs. Highlighting the trend, Associate Director Brady Wangsaid, “The first wave of AI PCs will gradually come up with three major CPU platforms – Intel Meteor Lake, AMD Hawk Point and Qualcomm Snapdragon X Elite series. These vendors are also preparing for next-generation solutions for AI laptop PCs rolling out later this year, which will accelerate the adoption of AI PCs at multiple price points. We believe Intel and AMD will also take GenAI-capable compute mainstream next year to compete with relatively efficient Arm-based Qualcomm and Apple solutions in this AI race.”
Wang added, “At the same time, we expect an upward trend in AI-capable laptops, which leverage high-level to premium-level GPUs to enable even higher-end GenAI applications. This will range from current large language models (LLMs) processed at the edge to advanced large models and even autonomous gaming graphics generating (AI Gfx) models. Players like NVIDIA will look to extend their capabilities and leadership in this segment.”
Generative AI supported by both Cloud and Edge
The adoption of GenAI in computing, which was initially solely dependent on the cloud, warrants higher computational capability and performance, from training the model to inferencing millions of potential requests per second. However, this is not feasible as well as viable from capacity, cost and energy perspectives. As a result, GenAI capability at the edge or in the PC becomes paramount.
Source: Counterpoint Research
Further, besides better work efficiency, data privacy and security are also important, which can be facilitated by GenAI applications at the edge. This also becomes pivotal to AI at the edge to preserve data sovereignty and environmental efficiency.
Commenting on GenAI use cases, Associate Director Mohit Agrawal said, “Initial GenAI adoption in the PC will be driven by Microsoft through its Copilot AI deeply integrated across Microsoft properties and further in upcoming Windows 12 along with app developers and partners such as OpenAI, Adobe and Hugging Face, which will catalyze the GenAI and overall AI experiences, initially around productivity and content creation in the PC. It remains to be seen how AI-powered Windows 12 charts upgrade path and AI capabilities for the existing installed base of devices which might not support necessary hardware capabilities to run GenAI applications natively.”
Agrawal added, “Apple could be the dark horse when it comes to adding GenAI capability in the Macs. The company can use its end-to-end vertical approach to leverage its self-designed Arm-based M series of advanced powerful processors, heavily optimized MacOS, newly designed LLM and powerful GenAI application ecosystem.”
Background
Counterpoint Technology Market Research is a global research firm specializing in products in the TMT (technology, media and telecom) industry. It services major technology and financial firms with a mix of monthly reports, customized projects, and detailed analyses of the mobile and technology markets. Its key analysts are seasoned experts in the high-tech industry.
Qualcomm has launched the Snapdragon X Plus platform, setting a high benchmark for the first AI PCs to be launched in 2024.
The Snapdragon X Plus is the second chip under the Snapdragon X series, which was announced in October 2023 as the company’s flagship ARM-based processor for Windows PCs.
The biggest talking point of the Snapdragon X series is on-device AI capabilities. At 45 NPU TOPS, the two chips are among the most sophisticated AI PC platforms in the market.
Qualcomm Technologies announced the launch of the Snapdragon X Plus platform on April 24. The Snapdragon X Plus is the second chip under the Snapdragon X series, which was announced in October 2023 as the company’s flagship ARM-based processor for Windows PCs. Further, the Snapdragon X Plus is a cut-down version of the Snapdragon X Elite, but both chips boast impressive features, setting high benchmarks for the first AI PCs to be launched in 2024.
CPU: Best in Class in Raw Computing Power, Superior Power Efficiency
Platform
Features
Snapdragon X Elite
Qualcomm Oryon 12 cores, up to 3.8GHz, 42MB cache capacity
Snapdragon X Plus
Qualcomm Oryon 10 cores, up to 3.4GHz, 42MB cache capacity
Thanks to the Oryon CPU cores, which represent Qualcomm’s first attempt to design its own ARM processor at 64-bit, the Snapdragon X series offers ‘best-in-class’ CPU performance. Qualcomm believes the Snapdragon X Elite and Plus offer 41% and 28% faster CPU performance (multi-threaded) than the competition, respectively. Besides, both can achieve peak PC performance at 48% and 39% less power than the competition. These come down to Qualcomm’s heritage in ARM processor designs which champion users on the move, where optimizing resource efficiency is prioritized. The immediate user experience takeaway will be faster app launch speeds and web browsing, and longer battery life when using productivity and video apps.
Source: Qualcomm
WATCH: “AI PC” Era Beckons with Snapdragon X Elite: Deep Dive with Qualcomm’s Kedar Kondap
GPU: Top Graphics for Immersive Entertainment
Platform
Features
Snapdragon X Elite
Adreno GPU, up to 4.6 FLOPS
Snapdragon X Plus
Adreno GPU, up to 3.8 FLOPS
The Snapdragon X series comes with integrated GPUs capable of running most games consistently at 30-40fps and topping at 60fps. The highest-payload AAA games should be perfectly playable on the GPU, satisfying the needs of most casual gamers. However, we believe Qualcomm will be wary of venturing into the gaming laptop race due to different market dynamics, and different user behavior expectations, such as a much heavier and bulkier form factor, which, more often than not, is plugged into a power source.
Source: Counterpoint Research
Building on the productivity theme, the Snapdragon X series GPUs can support up to three external 4K UHD monitors running at 60Hz. As the boundaries of the office blur and home entertainment requirements become a higher priority, these are important GPU features for the next-generation mobile workstations.
AI: Built-for AI Workloads on the Edge
Platform
Features
Snapdragon X Elite
Hexagon NPU, 45 TOPS
LPDDR5x, with RAM bandwidth of 135GB/s
Capability to run 13B+ parameter models on-device
Generates 30 tokens/sec for 7B LLMs
Snapdragon X Plus
The biggest talking point of the Snapdragon X series is on-device AI capabilities. At 45 NPU TOPS, the two chips are among the most sophisticated AI PC platforms in the market. According to Counterpoint Research’s definition of AI PC, the Snapdragon X series currently is the only platform available in the ‘AI-Advanced Laptop’ segment.
Qualcomm says it is working closely with major app developers (see below), focusing on two areas – delivering a top Windows-on-ARM user experience and scaling up AI features and capabilities. Questions from the industry on AI use cases are growing too loud to ignore, and Qualcomm suggests that a gradual process is underway with app developers to move key processing workloads from the CPU to the NPU. This will open up the opportunity to adopt multi-format models in the future, which can help the device deliver a more personalized user experience through inferencing at the edge.
Source: Qualcomm
Qualcomm demonstrated AI capabilities including real-time translation, image generation (Stable Diffusion) and image editing (GIMP). We felt the most impressive demo was Stable Diffusion, generating 10 images from text prompts at less than a second per frame. Nevertheless, these are well-known AI use cases that are unlikely to move the needle materially for the market.
The upcoming Windows 11 update, expected to be announced in late May, could be the turning point in the AI PC race, as the long-rumored AI Explorer feature will solidify the use case of AI on PCs. In short, AI Explorer (likely activated via Copilot) can help the user track all activities, such as apps, text, webpages, documents, photos and audio, via a searchable memory, using natural language. The Snapdragon X series will be among the first candidates to receive these AI updates.
BONUS VIDEO: On-Device AI on PC: Generate Text and Images Within Seconds
Play Video
Product positioning
The Snapdragon X series is aimed at the top end of the laptop space, with the likes of Apple’s M series and Intel’s Core Ultra being considered as key competitors. The Snapdragon X series has a decent edge in absolute performance but a much larger lead in power consumption. We think the key positioning message is a mobile generalist workhorse, capable of heavy-duty creative tasks and future-ready for hybrid and multi-modal AI scenarios.
In addition to the obvious benefits of ARM-based processors, like power consumption, Qualcomm will be keen to highlight the growing gap between traditional x86 CPUs and SoCs as all-inclusive platforms, with the latter offering graphics, audio, connectivity, sensor and security features that are bundled with the Snapdragon X series platform. This opens the opportunity for more heterogenous systems and seamless user experiences in the future, potentially also extending to closer integration between Snapdragon’s PC and mobile platforms.
We think the first batch of Snapdragon X series laptops will be positioned as consumer devices, as we expect x86 to continue to dominate the ‘for business’ laptop segment. As a matter of fact, Microsoft is expected to launch soon both the Surface Pro 10 and Surface Laptop 6 with Intel Core Ultra 5 and 7 and Snapdragon X Elite processors. The Intel models are officially dubbed ‘Surface Pro 10 for business’ and ‘Surface Laptop 6 for business’. We think the first Snapdragon X series laptops will be priced at close to $1,500, but in the medium term, mid-to-high-end AI laptop prices should drop toward the $1,000 level.
What next?
There is a palpable sense of excitement as the AI PC era edges near. May and June 2024 are expected to be busy as players put the final touches on their first AI PC offerings. The Windows 11 update on May 20, the Computex event on June 3 and Apple’s WWDC on June 10 will set the stage for the first innings of the AI PC race. The first AI PCs should go on sale by mid-2024, but shipments should take off only by 2025. By the end of the decade, almost all PCs will be AI PCs. However, between now and then, a lot of work, such as defining new features, creating fundamentally different consumer and business use cases, and more ecosystem support, will need to be undertaken.
In the coming days we will publish a full report on Counterpoint Research’s learning during recent events hosted by Qualcomm, including further analysis on the key selling point, positioning and competitiveness of the Snapdragon X series, and expected market implications in the fledging AI PC sector. We also preview the potential key AI features and use case enhancements that are likely to arrive on the first Qualcomm Snapdragon X series PCs, set to arrive in stores by mid-2024.
In order to access
Counterpoint Technology Market Research Limited (Company or We hereafter) Web sites, you may be asked to complete a registration form. You are required to provide contact information which is used to enhance the user experience and determine whether you are a paid subscriber or not.
Personal Information
When you register on we ask you for personal information. We use this information to provide you with the best advice and highest-quality service as well as with offers that we think are relevant to you. We may also contact you regarding a Web site problem or other customer service-related issues. We do not sell, share or rent personal information about you collected on Company Web sites.
How to unsubscribe and Termination
You may request to terminate your account or unsubscribe to any email subscriptions or mailing lists at any time.
In accessing and using this Website, User agrees to comply with all applicable laws and agrees not to take any action that would compromise the security or viability of this Website. The Company may terminate User’s access to this Website at any time for any reason. The terms hereunder regarding Accuracy of Information and Third Party Rights shall survive termination.
Website Content and Copyright
This Website is the property of Counterpoint and is protected by international copyright law and conventions. We grant users the right to access and use the Website, so long as such use is for internal information purposes, and User does not alter, copy, disseminate, redistribute or republish any content or feature of this Website. User acknowledges that access to and use of this Website is subject to these TERMS OF USE and any expanded access or use must be approved in writing by the Company.
– Passwords are for user’s individual use
– Passwords may not be shared with others
– Users may not store documents in shared folders.
– Users may not redistribute documents to non-users unless otherwise stated in their contract terms.
Changes or Updates to the Website
The Company reserves the right to change, update or discontinue any aspect of this Website at any time without notice. Your continued use of the Website after any such change constitutes your agreement to these TERMS OF USE, as modified.
Accuracy of Information:
While the information contained on this Website has been obtained from sources believed to be reliable, We disclaims all warranties as to the accuracy, completeness or adequacy of such information. User assumes sole responsibility for the use it makes of this Website to achieve his/her intended results.
Third Party Links:
This Website may contain links to other third party websites, which are provided as additional resources for the convenience of Users. We do not endorse, sponsor or accept any responsibility for these third party websites, User agrees to direct any concerns relating to these third party websites to the relevant website administrator.
Cookies and Tracking
We may monitor how you use our Web sites. It is used solely for purposes of enabling us to provide you with a personalized Web site experience.
This data may also be used in the aggregate, to identify appropriate product offerings and subscription plans. Cookies may be set in order to identify you and determine your access privileges. Cookies are simply identifiers. You have the ability to delete cookie files from your hard disk drive.
