Nvidia's Record Quarter Masks a Dual Challenge: A Cancelled Chip and a Top Investor's Warning on Customer Concentration
27.05.2026 - 20:41:36 | boerse-global.de
Nvidia just posted quarterly numbers that would make any semiconductor company envious. Revenue hit $81.62 billion in the first fiscal quarter of 2027, an 85.2 percent jump from a year ago. The data center division alone generated $75.25 billion, climbing 92 percent. Yet the stock can't catch a break — early US trading saw shares drop 1.8 percent to $211.04, while Frankfurt-listed stock fell 1.7 percent to €181.32. Over the past week, the decline has stretched past 5 percent.
The disconnect between financial performance and market sentiment stems from two distinct pressures. One is a warning from famed investor Michael Burry, who on May 25 laid out a stark case against the stock. His central criticism: Nvidia's three largest customers account for 64 percent of the company's outstanding receivables. That concentration, he argues, has created a "sell-the-news" pattern, where even blowout earnings and analyst price target hikes fail to sustain upward momentum. Rothschild & Co Redburn recently raised its target to $300 with a Buy rating, noting 14 consecutive quarters of sequential revenue growth, but the shares barely budged.
The second headwind is internal. Nvidia is quietly overhauling its product roadmap, and one casualty appears to be the much-anticipated Rubin CPX inferencing chip. Originally slated to pack 128 GB of GDDR7 memory and arrive in the second half of the year, the chip has disappeared from the company's roadmaps — notably absent from the GTC 2026 keynote. Supply chain signals confirm the shift: no visible orders for the required memory or specialized substrates. Industry observers now treat the project as effectively killed, even if Nvidia stops short of an outright cancellation.
Should investors sell immediately? Or is it worth buying Nvidia?
In its place, Nvidia is doubling down on a licensing deal with Groq 3 LPX worth $20 billion, pivoting toward lower latency, real-time inference, and specialized AI infrastructure rather than a pure GPU refresh cycle. The company is also pushing aggressively into agentic AI. At a Rescale event, it unveiled PhysicsNeMo, a platform designed to embed AI agents into industrial simulation workflows. Early results from partners such as McLaren Automotive show promise: a threefold productivity boost, a 30-times improvement in cost efficiency, and a 60 percent reduction in development cycles. The platform also allows four times as many design candidates to be evaluated in the same timeframe.
On the hardware front, Nvidia is not retreating entirely. Its first in-house CPU, Vera, posted strong Linux benchmarks, using 88 Olympus ARM cores and achieving 1.2 TB/s of memory bandwidth. That places it 55 percent ahead of Intel's Xeon 6980P and 10 percent ahead of AMD's EPYC 9575F, with significantly better energy efficiency than traditional x86 processors. Analysts estimate the addressable market for the Vera platform at $200 billion, and Nvidia could book $20 billion in CPU-related revenue by the end of fiscal 2027. Meanwhile, it expanded its collaboration with Marvell Technology through a $2 billion investment to develop custom XPUs and network solutions tied to NVLink Fusion.
Financially, the company has the firepower to fund this transformation. Management guided for roughly $91 billion in second-quarter revenue, representing 95 percent year-over-year growth and 22 percent sequential expansion. Outstanding purchase commitments stand at $119 billion, reflecting deep supplier ties. Shareholder returns remain generous: an $80 billion share buyback program has been authorized, and the quarterly dividend was lifted to $0.25 per share. The next-generation GPU architecture is due to ship in the third quarter of 2026, with initial orders already exceeding all prior product cycles.
Burry's concentration warning and the roadmap shift are not unrelated. As Nvidia reduces its reliance on any single chip generation and moves toward licencing, custom design, and platform-based AI services, it is also betting that a broader revenue base can dilute the customer dependency that has made the stock so sensitive to news flow. Whether that bet pays off will depend on how quickly the Vera CPU, the Groq deal, and the agentic AI push translate into earnings — and whether the market is willing to look past the concentration risk in the meantime.
Ad
Nvidia Stock: New Analysis - 27 May
Fresh Nvidia information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
