Data Center Chip Giants Negotiate Political Moves, Tariffs, and Corporate Strategies

For followers of the semiconductor industry, August 2025 has been quite an interesting month. From the hands-on involvement of a sitting president, to an interesting method of tariff application, the usual suspects - NVIDIA, AMD, and Intel - all found themselves in the news for things strictly beyond their technology announcements.

Unprecedented Arrangements

In brief, AMD and NVIDIA both struck an unprecedented 15% revenue-sharing arrangement with the U.S. government for sales of approved AI chips into China, framed as the price of an export license in a high-tariff environment. This sits on top of broader tariff threats and EU/US trade adjustments that explicitly reference semiconductors.

This happened while President Trump very publicly called for the resignation of Intel CEO Lip Bu Tan, due to previous connections to China. Tan met with Trump, the tone flipped, and the administration began floating the idea of getting an equity stake in Intel. While this was going on, SoftBank committed $2 billion for just under 2% of Intel, both a lifeline and a signal that Intel is still a viable competitor.

Of course, nothing is ever as simple as it sounds, so let’s take a moment and drill down into what, exactly, has been going on in the semiconductor sector in August.

The Problem with Intel   

In what seemed like an odd piece of political theatre for the tech sector after days of public pressure in which President Trump said Intel’s Lip-Bu Tan “must resign”, Tan met with Trump at the White House. Intel then issued a conciliatory statement, and it seems that both sides of the disagreement have found common ground.

Intel, despite years of problems is clearly still an important engine in the sector, and pressure from the government is being applied to get the company back on track. Multiple reports indicate the administration is considering taking an equity stake, perhaps equaling 10%, in Intel, potentially by converting prior CHIPS Act grants. Commerce and Treasury officials have publicly acknowledged interest in equity as a way to protect taxpayer exposure.

That would be an historic intervention in the core of the U.S. semiconductor stack, far more hands on than efforts such as the CHIPS Act.

This signal of government support would come at a time when Intel posted steep losses in 2024 and continues to fight on three fronts:

•  Lagging process tech vs. TSMC for high-end logic.
•  A still-fragile foundry customer pipeline.  
•  Regaining product competitiveness in AI accelerators and server CPUs.

A SoftBank Landing?

In what could be construed as a “vote of confidence” SoftBank agreed to invest $2 billion for just under 2% of Intel, via a primary share issuance. Intel stock responded positively on the news; SoftBank gets optionality on a turnaround and political goodwill in Washington at limited cost.

Crucially, this is non-customer capital; it’s not a commitment to buy parts, it’s a company investment that signals industry support and provides capital to Intel.  Not a solution to the company’s problems, but an action that can provide investor confidence as a bridge while Intel addresses the three previously mentioned issues.

If the U.S. becomes a major Intel shareholder, we could expect government pressure to:

• (a) Land anchor wafers from the big U.S. fabless players such as NVIDIA, AMD, Qualcomm, Apple, and Broadcom.
• (b) Prioritize domestic 18A (Intel’s current leading -edge foundry process)/14A (Intel’s next generation process not expected to reach volume production until 2028) ramps. Lip-Bu Tan is reportedly considering halting the promotion of 18A to external clients and prioritizing 14A, potentially to better compete with TSMC and attract major players like Apple and Nvidia. This could involve ending external promotion of 18A, but Intel would still use it for its own products. Aggressive adoption of 14A would make Intel more competitive with TSMC fabs.
• (c) Condition support on governance/operational milestones. It would be interesting to see what goals/milestones were focused on.

That could realign parts of the U.S. fabless ecosystem toward Intel over time, but only if Intel hits process and yield targets. TSMC still holds the lead on next generation processes at this level.

At press time the only committed part of these agreements is SoftBank agreeing to invest the $2 billion for the stake in Intel.

Back to the Tariff Wars

So. AMD and NVIDIA both struck an unprecedented 15% revenue-sharing arrangement with the U.S. government for sales of approved AI chips into China. This has been framed as the price of an export license in a high-tariff environment. This sits on top of broader tariff threats and EU/US trade adjustments that explicitly reference semiconductors.

This 15% is not technically a tariff at the border; it’s a license fee tied to export approvals - but applied in this way is a fit for traditional profit and loss statements. After criticism that the approach monetizes national-security policy, the White House reiterated its support for the measure.

And with the current restrictions being placed on US manufacturers selling AI parts to China, reporting says NVIDIA is developing a Blackwell-based China chip, more capable than the current H20 but still structured to comply with U.S. export rules. Reuters reported that it would be  a single-die design (roughly half the compute of the dual-die B300), with HBM and NVLink, sampling as soon as next month. A second compliant workstation/inference product (RTX6000D) is also in development.

Chinese agencies have reportedly discouraged use of NVIDIA H20 in government work, favoring Huawei Ascend. However, there have been reports describing AI training using the Ascend to be “challenging”, forcing some AI firms to revert to NVIDIA for large-scale training while using Ascend for inference. This keeps China demand alive for compliant NVIDIA/AMD parts—hence the U.S. interest in revenue-sharing.

Meanwhile, AMD made its announcements at June’s “Advancing AI 2025” to set MI350 (CDNA 4) expectations and a yearly rollout rhythm that’s designed to erase NVIDIA’s time lead as much as fight on absolute perf/Watt. If MI350 systems ramp aligns with major cloud designs in 2026, AMD’s near-term objective is defending MI300X momentum while converting large customers to multi-vendor strategies (often pairing MI clusters with NVIDIA estates for redundancy and price leverage).

The 15% China license fee will shape how AMD prices MI-series export SKUs and whether Chinese hyperscalers still prefer them to the domestic alternative (Huawei Ascend), which continue to face software/toolchain challenges. If Chinese buyers balk or Beijing discourages purchases, the revenue-share may be moot; if they don’t, AMD has a path to keep seats warm in China while building MI350 demand elsewhere.

Beyond China export licenses, the U.S. and EU recently averted a larger trade war by settling near 15% on certain sectors, which included semiconductors, as opposed to the far more draconian rates floated earlier. Even if AMD isn’t the direct target (it fabs at TSMC), anything that nudges module/board assembly, test, or accessories into tariff buckets can creep into bill of material costs.

The Overall Impact

Fundamentally, we are still a little early to determine the impact of these tariff and production decisions on the US economy. 

We can look at the earnings numbers that NVIDIA drops at the end of August and wait on formal announcement of any China-specific part developments. For AMD, we need to watch how the tarrif impacts the cadence of new design rollouts and determine if there are any China-specific impacts on the adoption of the next generation AMD parts.

And as much as we would like to say that Intel will be having a significant impact on the sector, it is simply too early to tell. The political theatre has been interesting, and the Softbank investment a positive note, but success will only be measureable as the company moves to next generation production, hits its timeline goals, and starts to secure significant customer wins.

Keep pace with the fast-moving world of data centers and cloud computing by connecting with Data Center Frontier on LinkedIn, following us on X/Twitter and Facebook, as well as on BlueSky, and signing up for our weekly newsletters using the form below.