Amazon.com (AMZN), Marvell Technologies (MRVL), and Alphabet Inc (GOOG) are advancing custom chip strategies, but NVIDIA (NVDA) leads in AI GPUs and the CUDA ecosystem, outperforming AMZN (AWS) and GOOG (TPUs) in high-end AI tasks

Despite growing competition from AMZN, MRVL, and GOOG, NVIDIA remains dominant in AI, particularly in training large models, autonomous driving, and data centers, and continues to lead AI-driven data center demand and enterprise applications

Selling NVDA at $180 may be reasonable, while holding is ideal for those confident in its continued AI and data center dominance, and buying rising stars like AMZN or MRVL could be worthwhile long-term despite their smaller scale

The chip war is long-term, with rivals emerging, but NVDA maintains a strong edge, so whether to buy, sell, or hold at $180 depends on its future in AI and data centers, with holding ideal if NVIDIA leads and diversifying into rivals a move if competition grows。。。

Tag :@Huat99  @Snowwhite  

Amazon, Marvell, Google Challenge NVIDIA: Is $180 a Buy or Sell?

@Tiger_comments
Recently, $NVIDIA(NVDA)$ hasn’t had an easy time. In addition to last month’s impact from $Alphabet(GOOG)$’s TPU, $Amazon.com(AMZN)$ yesterday announced a new in-house AI chip, which it claims is more cost-effective than Nvidia’s. $Marvell Technology(MRVL)$ jumped nearly 10% after reporting announcing a $3.25B acquisition of AI-chip startup Celestial AI. This move pushes Marvell another step closer toward becoming a viable Nvidia alternative. 1. “Three-way battle”: Nvidia vs. Google TPU vs. Amazon AWS (Trainium) Nvidia has established a nearly unshakable de-facto standard with its GPU hardware + CUDA ecosystem. Google, through its self-developed TPU chips, has achieved extreme vertical integration for AI workloads. TPUs primarily serve Google’s own large models (such as Gemini) and remain a key strategic asset in maintaining their AI leadership. AWS focuses on cost-performance through in-house chips, aiming to provide customers with more economical AI compute options on the cloud. 2. How institutions view Amazon’s AI chips — Is Amazon undervalued? Citi target price: $320 Based on 33x 2027E GAAP EPS ($9.68) EV/EBITDA = 13x, below the 10-year median of 18x → Citi sees the market undervaluing AMZN Trainium ecosystem: more powerful, faster, larger scale. Over 1 million Trainium chips deployed. Trainium2 is AWS’s fastest-ramping AI chip ever, producing chips 4× faster than the previous generation. Morgan Stanley Rating: Overweight/Price target: $315 (~35% upside) Views re:Invent 2025 as delivering major updates across GPUs, Trainium, and agent platforms. Expects AWS to maintain low- to mid-20% growth in 2026–2027. Competition intensifies: Marvell challenges Nvidia & Broadcom Marvell will acquire Celestial AI for up to $5.5B, aiming to absorb its leading photonic interconnect platform and strengthen its position in AI data-center connectivity. Amazon also provided warrant support to the deal. Between Broadcom’s ASIC order wins and now Marvell’s acquisition, the market is questioning: Can Nvidia still maintain its revenue trajectory next year? Worse yet: OpenAI turbulence adds another layer of uncertainty OpenAI’s CEO reportedly issued an internal “red alert” yesterday: Google Gemini 3 poses a serious threat to ChatGPT Cash-flow problems may mean OpenAI faces another funding crunch Nvidia itself may be fine, but the ecosystem it relies on faces three major risks: $150B+ debt from ORCL / CRWV / NBIS Long-term CUDA ecosystem pressure from Google & Amazon OpenAI cash-flow shortage, jeopardizing demand Back in September, Nvidia announced a $100B investment in OpenAI, and both parties announced a $500B collaboration. In October, SoftBank injected $22.5B more. But yesterday Nvidia’s CFO admitted: The highly publicized “$100B investment” is still only at the MOU stage, not finalized. Meanwhile, Anthropic is rumored to be preparing for an IPO as early as 2026. Some in the market believe NVDA may fall to the $155 support level. So how do you view the current three-way chip war? Are chip rivals steadily siphoning Nvidia’s revenue? Should you sell at 180 or hold? Buy the rising stars? Leave your comments to win tiger coins~
Amazon, Marvell, Google Challenge NVIDIA: Is $180 a Buy or Sell?

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment

  • Top
  • Latest
empty
No comments yet