TrendForce Corporation’s Post

💥 Rivaling Qualcomm’s next-generation AI PCs? Intel’s release date for its next-generation processor, Lunar Lake, has been moved up! ⚠️ Yet, this shift has sparked speculation about potential disruptions to the industry's status quo, with reports suggesting that brands are quietly voicing their discontent regarding how the memory has been bundled with the processor. 💡 Explore more behind their potential discontent and details about Intel’s projected release of Lunar Lake here: https://buff.ly/3KaozXy 🔗 #Intel #AIPC #LunarLake

What is "AI PC"? AI PC is the new buzz word in the world of technology. Every other company is racing to build an AI PC. We all know that an AI PC must have a CPU, NPU, & GPU, but there is still no unified definition of "AI PC." For the first time, Intel Corporation & Microsoft teamed up last week to define "AI PC" during Intel's AI summit in Taiwan. According to both: 1. The #Copilot key will be a must if you want to market your PC as an AI PC to catch the hype. 2. minimum of 40 TOPS (Power unit) in terms of NPU (Neural Processing Unit) performance 3. Most #AI PC will be from Intel, but over time, other options, i.e AMD, Qualcomm, will also be available. But it's Intel that's going all in the AI PC game. 4. Microsoft wants to run AI on devices instead of #cloud servers. Source: XDA https://lnkd.in/gyB5qE4R #Ammar #futuretech #aihype

Exciting times at Intel! We’re proud to introduce the Intel Core Ultra 200s Series, marking a new era in desktop computing. With up to 24 performance and efficiency cores and turbo speeds reaching 6GHz, this groundbreaking processor is designed to handle the most demanding tasks, whether you're a creator, gamer, or professional. But what really sets it apart? The integrated AI engine that revolutionizes how your PC works, delivering smarter performance and enhanced multitasking. Plus, with Wi-Fi 7 and Thunderbolt 5, you get faster, more seamless connectivity. The Core Ultra 200s Series is built to power the future, offering the ultimate balance between top performance and cutting-edge AI capabilities. This launch is just the beginning, and we can’t wait for you to experience everything this processor can do! Discover more here: https://lnkd.in/g8KySgzC #IAmIntel #IntelCoreUltra #AI #DesktopInnovation #Performance #NextGenComputing #NewEra

#AIFD4 has been full of surprises, and this morning kicked off with a doozy! Modesty from Intel Corporation on the CPU's role in inference. I argue that they actually understated their case...and make a few more observations on what I consider the most interesting innovation corner of the industry in 2024, AI silicon delivery. Check out my thoughts here:

20 years ago would any of you have expected that Taiwan would become the epicenter of announcing new CPUs, proclaiming the dawn of AI PC? And that an AMD CEO would give the opening keynote at Computex instead of an Intel CEO? And that the market cap of AMD is almost double that of Intel? And that nVidia would steal all the limelight by giving a pre-keynote keynote to proclaim the next phase of AI at National Taiwan University attended by all the corporate luminaries in Taiwan? Jensen showed off his mastery of Taiwanese and Mandarin beside English. All because of TSMC who was led by an CEO who was born in China but grew up in HK and Mainland China and became a talent in the US, plus 2 semiconductor CEOs who were born in Taiwan but grew up in the US and became talents in the US. Universities in Taiwan and the United States played a big part, whether it's public or private. Massachusetts Institute of Technology Stanford University Oregon State University National Taiwan University

High Performance Computing is getting higher performing. ExtremeTech has an article today discussing the current two supercomputers with exascale performance: AMD's Frontier and Intel Corporation's Aurora (both from Hewlett Packard Enterprise have reached this accomplishment, Frontier in 2022 and Aurora this year. The statistics are impressive (and confusing since there are a number of different statistics being discussed plus the fact that Aurora is only running at partial functionality during testing) with Frontier showing 1.2 exaflops and Aurora showing 1.012 exaflops currently. Of course that is only part of the story. Intel claims that Aurora is an #AI monster with 10.61 "AI exaflops" making it the system with "the world's best performance for AI". Fortunately (I guess) AMD is ready for the challenge: "However, AMD isn't just sitting around counting its flops. It's also planning to deploy its own next-generation supercomputer this year called El Capitan, which is expected to offer 2 exaflops of performance." And, of course, NVIDIA is playing the game too: "Nvidia has also joined the party and announced today that nine supercomputers around the world are now running its Grace Hopper CPU+GPU combo." All of which leads to the fact that it is a great time to be in the business of needing supercomputers.

As the world continues to explore the massive power of generative AI, semiconductor manufacturer AMD has positioned itself as the leader in the silicon chips that fuel this demand. At the company's #AdvancingAI event, it announced a succession of new products spanning desktop and data centre CPUs, GPU accelerators, and even DPUs and AI NICs. Every item is part of its strategy to own the end-to-end AI processing market, delivering more grunt, more bandwidth, but yet in a smaller footprint with lower energy requirements and a faster RoI and higher TCO than ever before. In fact, not only will the AMD hardware train AI models and infer faster, they can do much more per core, giving customers up to 60% more value out of their per-core licensing (e.g., database servers) than the competition. There is a lot of news, a lot of substance, but one thing's for sure: the competitor who had recognition simply by being "inside" your computer has a lot of catching up to do. #AMD #AdvancingAI #AI #generativeAI #CPU #GPU #DPU #NPU #Ryzen

Intel Corporation takes its AI game to the next level: 48 TOPS of AI performance, 80% more gaming performance and 60% better battery life vs previous gen! Intel #LunarLake is here! We are in Taiwan; Jayesh Shinde from Digit got a deep-down on the AI chops and performance boosts that Intel's Lunar Lake chips are about to bring to the fiercely contested AI chip battleground. What's baked into the Lunar Lake Compute Tile you ask? > Xe2 GPU: Boasting over 80% more gaming performance and more than five times the AI throughput compared to the previous generation, the Xe2 GPU delivers up to 67 tera-operations per second (TOPS). Its new microarchitecture for display and media engines ensures a significantly improved visual experience compared to the older-gen Xe GPU. > NPU 4: Offering three times more TOPS over its predecessor, the NPU 4 accelerates AI computations with impressively low power, reaching up to 48 TOPS. > IPU: Enhances camera experiences while reducing power consumption. Intel is going all guns blazing and how! Who will stand tall after the dust settles?

According to Intel executives at the AI Summit in Taipei, Microsoft's Copilot AI service will soon run locally on PCs instead of in the cloud. The company also confirmed a 40 TOPS requirement for NPUs on next-gen AI PCs. While Microsoft has been quiet about its plans for AI PCs, the company allowed Intel to announce its new definition of an AI PC, which includes an NPU, CPU, GPU, Microsoft's Copilot, and a physical Copilot key on the keyboard. The first wave of AI PCs meeting these requirements is already shipping, but Intel divulged future AI PC requirements in response to questions about potential memory criteria. #microsoft #intel #techgiants #ai #generativeai #copilot #npu #cpu #processor #gpu #hardware #computerhardware #aipc #technologynews

However, just as Intel had established dominance in the data center market, a new computing trend emerged—artificial intelligence (AI). AI tasks were a challenge for Intel's main chips, the central processing units (CPUs). CPUs were designed to be versatile, general-purpose workhorses, serving as the "brains" of computers and data centers. However, they weren't well-suited for AI's specific demands.