EP688 | 🦞
- This year's Computex marked a shift away from abstract visions and gimmicks toward highly practical tech integrated into daily workflows and entertainment.
- NVIDIA's consumer showcase highlighted deep collaboration with Adobe, leveraging Unified Memory to slash video editing and rendering times.
- The NVIDIA-MediaTek AI laptop chip delivers RTX 5070-class gaming performance, with massive memory pools key for local Agentic AI execution.
- Consumer brands like ASUS ROG went all out, with a lightweight AR/VR glasses prototype demonstrating real-world usability that eclipses heavy headsets.
- Despite massive corrections in index futures, the host views the leverage washout as healthy, backed by strong industry fundamentals seen at Computex.
Escaping the Hype: Computex Welcomes a New Era of Practical Tech
Attending this year's Computex felt fundamentally different from the past two or three years. Previously, the industry was flooded with lofty, abstract AI promises, robotic dogs with no clear utility, and gimmicky AI gadgets that left everyone with severe hype fatigue and fears of an AI bubble. This year, however, the fluff was completely gone, replaced by highly practical, mature products that make you want to open your wallet immediately.
Technology adoption often behaves like a gacha game where upgrading a card to a certain threshold suddenly triggers a qualitative leap. We have reached that tipping point. Instead of preaching a dystopian future where human creators are obsolete, this year's showcase placed humans back at the center, focusing on human-AI collaboration that makes daily tasks faster and more seamless.
Agentic AI Drives Hardware Architecture, Server Tech Evolves Amid Delays
At NVIDIA's Enterprise private showcase (held at TICC and Grand Hilai), although some highly anticipated LPU (liquid-cooled packaging) products faced further development delays and were kept under wraps due to complex engineering challenges, the server architectures remained highly impressive. Among them, the CPU server racks optimized for Agentic AI Workflows stole the show.
While some bears argue that CPU demand will decline as software models get optimized and stop hitting performance walls, history shows that lowering technology costs always unlocks larger user bases and brand-new applications. In an agentic workflow, CPUs act as dispatchers and coordinators, meaning high-end CPU configurations will remain a crucial secular trend to watch. Additionally, massive Power Racks with supercapacitors were shown, driving massive ASP (average selling price) expansion.
This is a real reasoning move, not just a bullish slogan. The host first states the bear case clearly — if models get better and stop "bouncing off walls," maybe CPU demand falls — then shifts the frame from per-task efficiency to total-system demand. That matters because tech debates often confuse "uses less per inference" with "needs less overall," when lower costs can expand the user base enough to more than offset the savings. The transferable lesson is to separate unit intensity from market-wide volume before concluding a component gets structurally disintermediated.
NVIDIA and Adobe's Deep Synergy: Unified Memory Unleashes Local Creativity
On the consumer side, NVIDIA's deep partnership with Adobe stood out. Despite Wall Street pricing Adobe as a prime casualty of the generative AI wave, NVIDIA demonstrated extensive, under-the-radar software-hardware optimization. By leveraging Unified Memory, RTX laptops running Premiere or Photoshop achieved rendering and shot-detection speeds three to five times faster than standard laptops.
This technology is not meant to replace human creators, but to supercharge their workflows. Creators who rely entirely on fully automated AI risk producing generic "AI Slop" and speeding up their own creative obsolescence. However, backed by local compute and vast memory pools, they can maintain their unique voice while achieving incredible efficiency. Unified Memory enables instantaneous rendering of massive, detailed cityscapes, providing a major performance boost for local 3A gaming and professional creative suites.
Sign up to keep reading
8 more sections await — finish reading on Podket.
Sign up freeFree to read. No credit card required.
Tickers are shown only because the company was mentioned in this episode, for your reference. Not investment advice, not a recommendation to buy or sell.
Disclaimer: The above is an AI-generated summary of a third-party programme, may contain errors or omissions, and is provided for personal, non-commercial reference only. Podket accepts no responsibility for its content and makes no representation or warranty as to its accuracy, completeness, quality, timeliness or reliability, and expressly disclaims any liability for any loss or damage arising from all or part of it. The content may reflect the personal opinions and views of the original programme's author and does not represent the position of Podket. Nothing in it constitutes any solicitation, offer, opinion or recommendation by Podket for any investment, nor legal, tax, accounting or investment advice or services regarding the returns or suitability of any security or investment. Investors must make their own investment decisions in light of their own investment objectives and financial circumstances.