AI is an arms race, and the US wants $9 billion in Nvidia superchips to keep up


Nvidia GB10 chip

Nvidia

Follow ZDNET: Add us as a preferred source on Google.


ZDNET’s key takeaways

  • Nvidia’s GB10 chips power modern AI models.
  • The US wants $9 billion for AI superchips.
  • Congress still needs to approve the funding.

The sudden — even by tech standards — pivot to AI hasn’t just left businesses scrambling to play catch-up; even America’s spy agencies are struggling to keep up.

Also: AI isn’t getting smarter, it’s getting more power hungry – and expensive

This is why the government has given the thumbs-up to a secret $9 billion request for superchips that will allow the CIA and NSA to keep up with what big AI players like Anthropic and OpenAI are doing.

So what are these superchips?

The latest batch of AI models needs a massive amount of computing power — not to mention a huge supply of power and specialized cooling that comes as part of cutting-edge, modern-day data centers — to run, and the silicon to deliver this is Nvidia’s Grace Blackwell superchips, named after American mathematician David Blackwell and American computer scientist and Navy pioneer Grace Hopper.

Also: How to learn Claude Code for free with Anthropic’s AI courses – one took me just 20 minutes

These superchips, called the GB10, feature a 20-core Arm CPU made by MediaTek, codenamed Grace, with an Nvidia GPU based on the Blackwell architecture. Take this chip and add 128GB LPDDR5x — it’s the demand for memory for AI that’s skyrocketed the price of RAM and things like Raspberry Pi boards — and 4TB of storage in the form of an NVMe M.2 SSD, and you have a chip that offers 1 petaflop of FP4 AI performance for a power draw of only 140 watts. That’s just the chip.

This isn’t a lot when you consider that most modern gaming PCs can swallow up to 1,000 watts of power.

How scaling Blackwell architecture works.

How scaling Blackwell architecture works.

Nvidia

This one chip has the power to fine-tune AI models with 70 billion parameters. Just in terms of storage alone, a model like this needs some 140GB of space.

Want a GB10 system? Best Buy sells a rack version starting at about $5,000!

Also: What Google’s TurboQuant can and can’t do for AI’s spiraling cost

But the real power usage comes when you scale these up. The GB300 NVL72 is a rack of as many as 72 GPUs and 36 CPUs in a single, liquid-cooled unit. Now scale this up to data center proportions, and you can start to see why the power demand goes through the roof.

A rack can cost anywhere between $1.8 million and $4 million. And a data center can have as many as 100,000 racks.

But if you want to run big AI models such as Anthropic’s Claude, OpenAI’s GPT 5.5, or DeepSeek’s V4, this is what you need.

Why does the government need this much power?

AI is seen as both a next-generation tool and a national security threat, something that is again moving faster than governments can legislate for or put guardrails in place. Just the other day, a planned executive order that would have outlined a process where AI companies would “volunteer” their models for government testing for a period of up to 90 days ahead of public release was scrapped after pressure from industry leaders.

A Dell GB10 desktop AI computer

A Dell GB10 desktop AI computer

Dell

This order makes it clear that not only does the government want to leverage AI itself, but it also wants to be able to examine models used by the public.

Also: How Qualcomm’s new wearables chipset could spell the end of smartphone dominance

This would need serious hardware horsepower.

There’s also an element of playing catch-up from a lack of investment in computing hardware over the past years. Combine that with the current shortage of chips and other AI hardware, and all that means having to spend billions just to stay in the game.

The $9 billion, which still needs to be approved by Congress, would allow the government to acquire both the infrastructure and hardware it needs to stay relevant in the AI game.

Inside a GB10-based AI system.

Inside a GB10-based AI system.

Dell

But buying chips and expanding data centers takes time, so in the interim, some $800 million of the defense budget has reportedly been repurposed to acquire more cloud compute power. The intelligence services are also continuing to make use of an advanced AI model developed by Anthropic called Mythos, despite the company being labeled as a supply chain threat.

$9 billion is just the tip of the iceberg

And that $9 billion, while it sounds like a lot, really isn’t in the grand scheme of AI. Amazon Web Services is investing $50 billion to upgrade its government cloud computing services, a platform that the intelligence agencies use extensively.

Also: I quit ChatGPT for a free, private, and local AI called Ollama – here’s why

And the successor to Grace Blackwell silicon is in the pipeline — the Vera Rubin platform, named after an American astronomer. These chips combine a brand-new, custom-built Arm-based CPU called Vera and a high-performance GPU called Rubin, and are designed to offer up to 10 times more performance per watt compared to Grace Blackwell and make use of high-performance HBM4 memory.

AI is now very much a modern-day arms race, and governments wanting to even keep up are going to have to invest moonshot money.





Source link

Leave a Reply

Subscribe to Our Newsletter

Get our latest articles delivered straight to your inbox. No spam, we promise.

Recent Reviews


Modern displays are amazing when it comes to detail, brightness, color, and all the ingredients that make for an impressive picture—except motion clarity.

CRT screens are still the king of motion clarity, but plasma flat-panel screens hold a respectable second place, and in many ways I still miss my old 720p 51-inch plasma TV and the crisp motion I gave up by switching to a 4K LCD.

Plasma solved motion the “right” way

Plasma displays didn’t just show an image—they flashed it.

While they operate on different principles, CRTs and plasma TVs have a few things in common. First, the phosphors used by CRTs and plasma displays are the same. Second, because these phosphors fade quickly, they need to be continuously refreshed.

In a CRT, the electron beam scanning from the top to the bottom of the screen achieves this, and in a plasma, a high-speed electric pulse does the same. Because of this rapid pulse-and-fade, these screen technologies have crisp perceptual motion, since our brains tend to interpret moving images that don’t pulse as “smearing” across our retinas.

The pulsing nature of plasma technology isn’t the only reason for its better motion reproduction. These screens also have very low latency and very fast pixel response times. Combined, it’s not quite as good as CRT motion handling, but it’s significantly better than LCD and OLED technology, even today.

Modern TVs rely on sample-and-hold—and that’s the problem

Stand and deliver blurry images

Blur Busters UFO Test

Modern LCD and OLED televisions are “sample and hold” technologies. They can hold each frame of video perfectly for the entire duration of that frame without deviating in brightness and then instantly snap to the next frame without any dipping to black in-between.

On paper, this sounds like a good thing, but your eyes don’t stay still when tracking motion. As they follow a moving object, the image being held on screen effectively drags across your retina, creating the perception of blur. Even if the panel itself is perfectly sharp.

You might not even realize how blurry motion is on modern displays if all you’ve ever seen with the naked eye is an LCD or plasma. However, if you see a CRT or plasma in person, the difference is quite striking.

The sample and hold issue means that no matter how much you increase the refresh rate, that type of blur persists. It’s why my 85Hz CRT monitor is clearly less blurry in motion than my 240Hz LCD monitor. It’s especially apparent when you’re playing 2D games that scroll the entire screen, with LCDs or OLEDs smearing the image in a way that gives me a bit of a headache if I’m being honest.

Playing Diablo 2 on a CRT. Credit: Sydney Louw Butler/Shutterstock.com

It creates this weird situation where a modern TV can be incredibly sharp in a freeze frame but somehow look softer than a lower-resolution display that isn’t sample and hold as soon as you press play.

Motion interpolation is a workaround, not a solution

It’s an abomination, that’s what it is

One of the “fixes” that TV makers came up with to reduce unwanted motion blur is a technology known as frame interpolation, or more commonly “motion smoothing.” Here an algorithm creates fake frames that guess at what the middle step of motion would look like if it were captured. This creates a high frame-rate video output, which we see as smoother and more crisp.

While this doesn’t take away sample-and-hold blur, it does improve motion clarity. Unfortunately, it also destroys the intended frame rate that shows and movies were meant to be seen at. It’s also useless for video games, because it introduces an enormous amount of input lag. NVIDIA’s DLSS technology is also frame interpolation, but it works for games because of several mitigations NVIDIA put into the technology. These measures don’t exist on TVs.

While some people think motion smoothing isn’t all bad, TV makers are no longer activating it by default as much anymore, and my advice is to always turn it off because the trade-offs are just not worth it.

Screenshot 2025-07-01 at 9.21.03 AM

7/10

Brand

TCL

Display Size

85-inches

The 2025 model TCL QM6K Google TV delivers a stunningly clear and bright picture with a new Mini-LED panel, improved local dimming zones, Dolby Vision IQ, and a neat new Halo Control system for improved visuals. Get this TV and elevate your living room. 


Black frame insertion tries to recreate plasma—but comes with trade-offs

Who turned out the lights?

The other trick sample-and-hold screens have to mimic what CRTs and plasma TVs do naturally is called BFI, or Black Frame Insertion. As the name suggests, the display inserts a full black frame between every original frame. This provides an instant and dramatic increase in motion clarity. However, it also has a big impact on brightness. As much as half of the light is now gone, so the image is much dimmer. Pushing overall brightness to compensate makes things hotter and more energy-hungry.

Some BFI implementations cause visible flicker, for which I personally have no tolerance at all, but the biggest problem here is that BFI doesn’t have the smooth pulsing roll off of the phosphors used in CRTs and plasma.


The future might circle back—but we’re not there yet

That might be changing, however, because a new generation of LCDs can leverage the power of multi-zone backlight technology to strobe the backlight across the screen in a way that mimics a CRT scanline.

NVIDIA’s G-SYNC Pulsar has received rave reviews from the biggest motion blur haters, and I sincerely hope that a similar technology becomes standard in TVs going ahead, so we can go back to enjoying the crisp motion we used to have without all the compromises.



Source link