Nvidia’s $100 Billion AI Power Play, Apple’s Surprise Intel Investment, and a Steam Deck Discount You’ll Actually Like

The tech world never sleeps — and this week, it’s been buzzing with billion-dollar deals, surprising partnerships, and a dash of gaming hardware drama. From Nvidia’s jaw-dropping AI investment to Intel’s ongoing fight for survival, and even a discount on the beloved Steam Deck, here’s your full roundup of what’s happening in the world of chips, GPUs, and gaming.

Nvidia’s $100 Billion Bet on OpenAI

Let’s start with the headline that rocked Silicon Valley. Nvidia, the undisputed leader in AI hardware, is reportedly preparing to invest up to $100 billion in OpenAI — the company behind ChatGPT and one of the biggest forces shaping the future of artificial intelligence.

According to early reports, the plan includes a direct $10 billion investment, giving Nvidia roughly a 2% stake in OpenAI. The rest will be tied up in a five-year GPU leasing deal, designed to supply OpenAI with around 4–5 million Nvidia GPUs for its next-generation AI data centers.

To put that into perspective, the projected 10 gigawatts of power required for this buildout is more than what many small nations consume — a staggering indicator of how energy-hungry large-scale AI operations have become.

The move has raised eyebrows across tech and policy circles, with some analysts wondering if the deal crosses into antitrust territory. Others point out the somewhat circular logic: Nvidia sells GPUs to OpenAI, OpenAI pays Nvidia for GPUs, and Nvidia profits twice. Yet, if history tells us anything, Nvidia has no plans to slow down its dominance in the AI arms race — even if regulators are starting to take notice.

US Government Eyes New Chip Rules — and Steep Tariffs

As if Nvidia didn’t have enough on its plate, Washington is reportedly drafting a new rule that would force companies like Nvidia to manufacture one chip domestically for every chip they import. The alternative? Up to 100% import tariffs.

The proposal, while well-intentioned in its bid to strengthen U.S. semiconductor independence, seems detached from current realities. Most of the world’s cutting-edge chip production still takes place in Taiwan and South Korea, with the U.S. lagging behind in advanced fab capabilities. Critics argue the new policy could strain supply chains further — especially given how different chips can be in complexity and purpose, from an Apple A19 SoC to a Nvidia B300 AI accelerator.

In short: a classic case of political ambition meeting technical reality.

Apple’s Quiet $5 Billion Stake in Intel

Meanwhile, Intel, America’s only leading-edge semiconductor manufacturer, has been battling for relevance. After years of missed opportunities, the company has turned to some unlikely allies — including Nvidia, the U.S. government, and now, Apple.

This week, reports revealed that Apple plans to invest $5 billion in Intel, giving it about a 4% ownership stake. The logic is strategic: if Intel’s domestic fabs survive and thrive, Apple could secure a U.S.-based chip manufacturing partner for its future products.

The move follows earlier rumors that both the U.S. government and Nvidia were considering similar investments. Together, these cash infusions are seen as attempts to stabilize Intel’s ship — a ship that’s been listing for years due to slow innovation, heavy buybacks, and a failure to ride the early AI wave.

Can Intel truly stage a comeback? Possibly, but not without producing real, profitable silicon. The company’s upcoming 18A and 14A process nodes might finally attract major fab clients like Nvidia or Apple — something Intel desperately needs to stay afloat.

Intel’s Frame Generation Surprise

Moving from boardrooms to codebases, an interesting tidbit surfaced on Reddit this week: Intel’s latest graphics driver references something called “Multi-Frame Generation via XeSS” — complete with an unreleased logo.

For the uninitiated, frame generation is a feature that uses AI to synthesize extra frames between rendered ones, boosting perceived frame rates. It’s controversial — Nvidia’s implementation often gets flak for inflating benchmarks — but it can make a real difference in smoother gameplay for single-player titles.

Intel’s version could be more open, though. The driver notes suggest compatibility across Intel, AMD, and Nvidia GPUs, aligning with Intel’s goal to make its XeSS technology more cross-platform than Nvidia’s DLSS. If this pans out, it might debut alongside the upcoming ARC B770 and B750 GPUs, expected later this year.

Intel Still Hiring for High-End Gaming

In other Intel news, the company recently listed a job opening for an SoC performance engineer focused specifically on gaming optimization and validation. The role’s location? Israel, where Intel’s GPU team is based.

Interestingly, the listing appeared shortly after rumors about Nvidia partnering with Intel to co-develop x86-based RTX SoCs — a potential game-changer if true. It also reinforces the idea that Intel hasn’t abandoned discrete GPUs yet, despite the challenges facing its ARC lineup.

Intel’s CPU Prices Just Went Up — Again

On the CPU front, reports from DigitTimes suggest that Intel is quietly raising prices by around 10% on its 13th-gen Raptor Lake processors. That means budget chips once priced near $150 may now sell closer to $170–$180.

The move comes amid weak sales for Intel’s Core Ultra lineup and ongoing concerns about chip degradation in some 13th and 14th-gen models. Whether this price hike helps Intel’s cash flow or drives more users to AMD remains to be seen — but it’s not the best look for a company trying to regain consumer confidence.

AMD Levels Up for Local AI

AMD didn’t make many headlines this week, but its developers quietly delivered something big for open-source AI enthusiasts. The company released a preview of ROCm 6.4.4, bringing native PyTorch support to Windows and Linux systems using RDNA3 and RDNA4 GPUs.

In plain English, this means you can now run local large-language models and image generators on many AMD graphics cards — without resorting to cloud-based solutions. The update supports cards like the RX 9000 and RX 7000 series, as well as certain Strix Halo APUs, and includes a step-by-step setup guide via the GPUOpen website.

It’s a small but important step toward breaking Nvidia’s monopoly in AI computing — and a big win for anyone who wants more control over their data.

Palit Gets Festive with 3D-Printable GPU Shrouds

On the lighter side, GPU manufacturer Palit decided to celebrate Oktoberfest with a quirky marketing twist. The company released a set of 3D-printable GPU shroud files, encouraging PC enthusiasts to create custom covers for their graphics cards — even ones shaped like beer mugs.

While the concept toes the line between creative and bizarre, it’s a fun example of how GPU brands are embracing DIY culture. Just… maybe skip the liquid cooling if your custom shroud involves actual beer.

Steam Deck Discount: The Real Deal

Finally, some good news for gamers. Valve’s original Steam Deck is currently 20% off, dropping the price to just $320 (down from $400). This deal runs until October 6th, and while it’s the older LCD model, it still delivers solid performance with its 1280x800 display and 256GB of storage.

At that price, the Steam Deck remains one of the most versatile portable gaming PCs available — and a perfect entry point for anyone wanting to play their Steam library on the go.

Wrapping Up

From Nvidia’s billion-dollar AI gamble to Intel’s desperate fight to stay relevant and AMD’s quiet software wins, this week’s tech news perfectly sums up the state of the industry: powerful, unpredictable, and a little bit chaotic.

And if all that sounds too heavy, just remember — the Steam Deck’s on sale, and that’s one bit of tech news you can actually enjoy without worrying about tariffs, power grids, or corporate mergers.

Click for more News