The air in New York at AMD’s 2025 Analyst Day last Tuesday was electric — and being there on-site made it even clearer that the real story wasn’t the record-breaking numbers. It was the palpable shift in the industry narrative.While the tech world remains fixated on Nvidia’s staggering valuation and Intel’s ongoing turnaround saga, AMD, under the steady hand of CEO Lisa Su, has been quietly executing a multi-year masterclass in strategy.The results speak for themselves: a record revenue of $9.2 billion, soaring 36% year-over-year (YOY) and obliterating consensus estimates. It wasn’t just a good quarter; it was a dominant performance, with non-GAAP net income skyrocketing 152% sequentially to $2 billion.AMD is no longer just the scrappy underdog; it’s a quiet giant steadily outmaneuvering rivals by capitalizing on their core weaknesses, like Intel’s market vulnerabilities and Nvidia’s growing list of self-inflicted problems.A fascinating undercurrent at the analyst event was the quiet chatter among attendees about customer frustration. There’s a growing sentiment among IT customers that they feel “cheated” by major OEMs that continue to default to non-AMD solutions.For years, Intel-based servers and Nvidia-based AI platforms were the “safe” choice. Now, IT managers are realizing they are paying a premium for solutions that are often less economically viable and, critically, not power-efficient enough.Recent candid remarks from Microsoft CEO Satya Nadella validated this sentiment. He stated that Microsoft’s primary bottleneck to AI expansion is no longer a chip shortage, but rather a power shortage. Nadella revealed that Microsoft has a stockpile of advanced GPUs it cannot deploy because it lacks the data center “warm shells” with enough available electricity to plug them in.This dynamic is where Nvidia’s “performance at all costs” mantra hits a wall of physical reality. As I’ve argued before, Nvidia’s proprietary reign invites disruption. If your most powerful GPUs are too power-hungry for your biggest customer to deploy, you have a massive market vulnerability. That reality has created a golden opening for a competitor focused on efficiency and real-world deployment, and AMD is walking right through it.Nvidia’s success has placed the company on a dizzying pedestal, but that height comes with its own kind of gravity. To keep supporting its massive valuation, Nvidia must constantly push its highest-margin, highest-hype products to the forefront. In that environment, it’s easy for blind spots to form — and for the company’s focus to drift away from what customers increasingly care about: cost and power.AMD, by contrast, is essentially flying under the radar. Free from the burden of a $5 trillion market cap, its leadership can remain laser-focused on a long-term plan. This strategy, as outlined at the analyst day, is a multi-pronged assault: compute technology leadership, data center leadership, pervasive AI, open software platforms, and custom silicon.This disciplined execution is a hallmark of AMD’s leadership team, many of whom, including Lisa Su and CTO Mark Papermaster, share roots (as I do) at IBM. This IBM heritage runs deep — shaping a culture of engineering rigor, long-term planning, and a customer-centric focus on solving complex enterprise problems.As a result, many companies are finding AMD a more reliable partner. They are perceived as collaborators, not dictators, who are willing to build custom silicon solutions rather than forcing a one-size-fits-all proprietary product down a customer’s throat.AMD’s Q3 results show it is masterfully executing a pincer move, attacking both its rivals on their home turf.On one front, AMD is systematically carving up Intel’s heartland:On the second front, AMD is outflanking Nvidia in the AI data center:This growth is so robust that AMD delivered a record quarter despite an $800 million Q2 charge tied to U.S. export restrictions on its MI308 China-bound chip. That kind of performance reflects a business that is both resilient and diversified, not reliant on any single product or market.AMD’s leadership isn’t resting on these results. The company is investing aggressively — $40 billion in organic R&D, much of it focused on AI, and another $60 billion earmarked for acquisitions. Together, these initiatives support an ambitious five-year plan to capture more than 50% of the server revenue market, over 40% of the client market, and more than 70% of the embedded market.Hitting those milestones would put AMD on a path toward a projected 35% CAGR. At the center of that growth story is the Helios Rack, a full AI building block built around AMD’s next-gen CPUs, GPUs, UALink, and AI NICs, all tuned for ROCm. Notably, partners — not AMD — will sell these systems, positioning them as a more open and efficient counterweight to Nvidia’s vertically integrated stack.The company’s guidance for Q4 2025 projects revenue of approximately $9.6 billion with an even stronger non-GAAP gross margin of 54.5%. This forecast, which excludes any revenue from the restricted China chips, shows that AMD’s growth is now fully self-sustaining.AMD’s Q3 2025 financial report is a powerful testament to the value of quiet execution over market hype. The company is perfectly poised to slipstream Nvidia in the AI race by offering a more efficient, open-source, and customer-friendly platform, all while simultaneously exploiting Intel’s market weaknesses to capture record share in the client and server markets.This dual-front war is succeeding, and the record-breaking $9.2 billion revenue — achieved despite significant geopolitical headwinds — proves AMD’s strategy is not only working but accelerating.While Nvidia grapples with the pressures of its own valuation and power-hungry hardware its customers can’t even deploy, AMD is focused, disciplined, and clearly building the foundation to become the most trusted and versatile high-performance computing leader of the next decade.As the 2025 gift season arrives, the Geekom A9 Max is poised to claim the title of “best-performing micro-PC on the market.” Priced at $999 for the 32GB DDR5/2 TB SSD bundle, it delivers powerhouse hardware in a diminutive footprint.At the heart of the system lies AMD’s Ryzen AI 9 HX 370 (12-core/24-thread, 4 nm Zen 5), paired with Radeon 890M graphics and up to an 80 TOPS NPU for on-device AI workloads. Other reviews note that it outperforms many competitors at comparable prices in both CPU and GPU tasks. With support for up to 128GB of RAM and dual SSDs (one 2280, one 2230), the A9 Max offers unheard-of upgrade headroom in many mini PCs.With its compact dimensions of 135 × 132 × 47 mm and all-metal build, the A9 Max blends into dorm rooms, family living areas, or talented teens’ streaming setups with ease. Its robust IO lineup — dual USB-C (USB4), dual HDMI 2.1, dual 2.5 GbE, Wi-Fi 7, and Bluetooth 5.4 — means it’s ready for gaming, content creation, or AI applications right out of the box.This holiday season, for the tech-savvy child or teen who juggles gaming, schoolwork, and streaming, the A9 Max hits the sweet spot. Its compact chassis frees up desk space for both homework and play. Its high-performance handles AAA games and video editing alike. For a parent evaluating value, it’s a one-stop machine with minimal need for upgrades — the 32GB/2TB configuration is already generous.Thanks to its portability, the A9 Max also fits into a dorm, shared living space, or family gaming lounge without requiring a full-sized tower. With built-in support for multiple monitors and serious cooling, it’s ready for long gaming sessions or creative bursts without overheating or throttling.If you’re shopping for a micro-PC that doesn’t compromise on performance, the Geekom A9 Max is hard to top. At $999, it delivers full-sized power in a mini package — perfect for both gaming and creative pursuits. For a holiday gift this year, it stands out as an innovative blend of portability, efficiency, and raw capability that the next-gen user truly needs, making the Geekom A9 Max Mini PC my Product of the Week.

Read full article at source