To The Moon

As Google’s Gemini $Alphabet(GOOGL)$ $Alphabet(GOOG)$ 3 model gains massive global traction, the market has begun to worry:
Could OpenAI fall behind in this round of competition?
And could its previously touted “trillion-dollar investment plan” fail to materialize?

Affected by these concerns, stocks closely tied to OpenAI tumbled. $Softbank Corp.(SOBKY)$ plunged 10% yesterday and is now 44% off its highs. $NVIDIA(NVDA)$ dropped more than 6% intraday, and $Advanced Micro Devices(AMD)$ fell nearly 9%.

Rising alongside Gemini 3 was Google’s TPU — its in-house AI accelerator chip, which is a type of customized ASIC (Application-Specific Integrated Circuit).

There are even reports that Meta plans to spend billions of dollars to purchase Google TPUs.

This triggered market anxiety:
“Could Google shake NVIDIA’s dominance in AI chips?”
Funds began selling NVIDIA and AMD, shifting instead toward “Google-linked” companies like $Broadcom(AVGO)$ , Google’s TPU partner.

1. GPU vs TPU: A Long-Standing Debate Rekindled

In simple terms:

• GPU (NVIDIA) → General-purpose computing. They can run Google’s models, OpenAI’s models, even Alibaba’s models, and can also be applied in robotics and many other fields.

• TPU (Google) → Specialized. Essentially a type of ASIC, fully customized. Google’s TPU is designed exclusively for its own models and cannot run others.

Previously, the market also worried that ASICs could erode NVIDIA’s market share. On September 5 this year, Broadcom reported that it secured its fourth major ASIC customer, with an order worth $10 billion, scheduled for delivery in FY2026 Q3. Insiders said this customer was OpenAI.

The impact on share prices:
Broadcom surged 9.4% that day, while NVIDIA fell 2.7%.

Beyond Google, other tech giants also have custom ASIC plans — $Meta Platforms, Inc.(META)$ co-developing with Broadcom, and Microsoft working with GUC (Global Unichip Corp).

These partnerships were all announced months ago, and ASICs have already been widely discussed. But the sudden popularity of Google’s new model has once again dealt a heavy blow to NVIDIA.

Although ASICs cannot replace GPUs, the explosive success of Google’s new model is reminding investors that OpenAI may not necessarily win the AI race. If its commercialization fails to meet expectations, its previously announced trillion-dollar investment plan may also fall apart.

Therefore, SoftBank’s sharp decline is not surprising.

2. How Did NVIDIA Respond?

Last night, NVIDIA commented:

“We’re happy for Google’s success — they’ve made tremendous progress in AI, and we will continue supplying them.”
“NVIDIA is a full generation ahead of the industry — the only platform that can run all AI models across all compute scenarios.”

NVIDIA added:

“Compared with ASICs designed for specific AI frameworks or functions, NVIDIA offers higher performance, stronger generality, and better interchangeability.”

Thus, Google’s TPU does not inflict major damage on NVIDIA, and competition in large AI models remains intense. It is still uncertain whether Google’s Gemini 3 will ultimately prevail.

3. Where Can Investors Look?

Returning to the investment landscape: NVIDIA will likely remain controversial. As Broadcom expands its ASIC customer base, market anxiety will continue.

Compared with the uncertainty surrounding NVIDIA and AMD, some semiconductor companies are positioned to win regardless.

① $Taiwan Semiconductor Manufacturing(TSM)$

Whether it’s Google’s TPU, NVIDIA’s GPU, or AMD’s chips, all rely on TSMC for manufacturing.
Market expectations suggest TSMC may build or expand 10 fabs globally next year, with supply-chain support being crucial. Institutions estimate TSMC’s capex could reach $50 billion next year (vs. $40–42B this year).

②. $Micron Technology(MU)$ / $SK Hynix, Inc.(HXSCF)$

Whether TPU or GPU, AI chips require high-bandwidth memory. Micron and SK Hynix lead this field.

③. Semiconductor equipment makers such as $ASML Holding NV(ASML)$

Tight memory supply and constrained TSMC capacity signal a new expansion cycle for the semiconductor industry.
As the monopoly supplier of advanced lithography, ASML is well positioned. Applied Materials and Lam Research also maintain absolute leadership in their respective segments.

[Allin]From a stock-performance perspective:
TSMC reversed from a 4% decline to closing in the green, ASML gained more than 1.5%, and Applied Materials hit a new stage high.
Investors have clearly voted with their feet on who they believe the future winners will be.

Worried about NVIDIA?
Perhaps semiconductor manufacturing and equipment companies are the better bet.

❓Welcome to discuss 👇

• Will Google’s TPU truly challenge NVIDIA?

• Or are TSMC, Micron, and ASML the real “sure winners”?

For SG users only, A tool to boost your purchasing power and trading ideas with a Cash Boost Account!

For whom haven't open CBA can know more from below:

🏦 Open a CBA today and enjoy privileges of up to SGD 20,000 in trading limit with 0 commission. Trade SG, HK, US stocks as well as ETFs unlimitedly!

Complete your first Cash Boost Account trade with a trade amount of ≥ SGD 5000* to get SGD 688 stock vouchers*! The trade can be executed using any payment type available under the Cash Boost Account: Cash, CPF, SRS, or CDP.

Click to access the activity

Other helpful links:

• 💰Join the TB Contra Telegram Group to Get $10 Trading Vouchers Now🎉

• How to open a CBA. How to link your CDP account. Other FAQs on CBA. Cash Boost Account Website.