Late last year, Valve announced the recent hardware trifecta of the new Steam Machine, Steam Frame headset, and Steam Controller. Many eager fans were looking for a Steam Deck 2 announcement, but it was notably absent. In an interview with IGN following the announcement, Pierre-Loup Griffais, Valve software engineer, clarified the reason why. Valve isn't interested in an incremental 20-30% boost. It's waiting for a major generational leap when it comes to their Steam Deck 2.

Rather than treating it like every other handheld on the market, Valve is instead treating the Steam Deck like a console similar to the PS5 or Xbox. This means the company won't release a sequel until the silicon allows for a demarcated jump in power without killing the 2-6 hour battery life that we currently expect. Until these hardware advancements have been made; the Steam Deck 2 just isn't on the cards, and right now that chip doesn't even exist yet.

Why is Valve waiting?

Many users would love a Steam Deck 2

To understand Valve's decision, it's important to look at the state of the current market. In terms of competition, there's quite a lot, and chips like the AMD Ryzen Z2 Extreme are hitting the market in a range of other handhelds. The problem is that while these chips are faster, they often achieve their best results at higher TDPs of around 25W to 30W. Valve's magic window for the Steam Deck is closer to 3W to 15W meaning that these chips are just not what Valve is going for. There's no use in offering higher performance if that means we're sacrificing battery life in turn.

Instead, Valve is looking for an APU that provides a massive RDNA4 or RDNA5 performance while still staying under that 15W ceiling. This means that readily available off-the-shelf laptop chips or mobile chips are currently just too power hungry for Valve standards.

One of the Steam Deck's greatest strengths is its fixed hardware target. Developers like those for Elden Ring or Cyberpunk 2077 can optimize for exactly one set of specifications, making it closer to console logic versus PC logic. This means there's a risk of fragmentation if Valve releases a Steam Deck 1.5 every single year. They would lose that Steam Deck verified badge reliability because consumers might become confused about whether this applies to their model of their Steam Deck or not. Instead, Valve wants to ensure that when they release the Steam Deck 2, it's a new baseline for the next four to five years of gaming, similar to a console life cycle.

They've also referenced it not really being fair to customers to release a slightly better version every 12 months. For those who want to invest, they may feel put off by the fact that a newer version will be coming out so soon that the investment may not be worth it.

What technology is Valve waiting for?

The Steam Deck 2 needs a range of advancements

What would a Steam Deck 2 chip actually look like, though? Well, right now Valve isn't looking for a bigger engine; it's looking for a more efficient one. The company wants to be able to move the needle without turning the Deck into a loud, hot, 45-minute battery disaster. Three specific technologies need to converge in order for this to happen.

The node shrink is the first. The current Steam Deck OLED uses a 6nm process. Moving to TSMC's N3, which is 3nm, or even the upcoming N2, which is 2nm, isn't just about fitting more transistors in; it's about the power curve. At the 15W TDP limit, a 3nm chip can deliver significantly more clock speed than a 6nm chip for the same energy cost. Smaller nodes also allow for the device to stay thin and the fans to stay quiet until 3 nm capacity is cheap enough for the $400-$600 handheld. The Goldilocks chip for the Deck 2 remains financially out of reach.

There also needs to be a major architectural leap. The Deck is currently powered by a custom Zen 2/RDNA 2 APU, while the industry has moved to Zen 4 and RDNA 3. This means Valve is likely looking to skip a generation to ensure a console-like leap. RDNA 5 is rumored to be a foundational clean sheet redesign of AMD's graphics architecture. For Valve, this could mean native hardware support for more advanced ray tracing and AI upscaling. This can all happen without draining the battery. In terms of CPU, Zen 6 will focus on better area efficiency, allowing Valve to potentially use more compact cores that handle background SteamOS tasks while leaving the heavy lifting for gaming.

Lastly, the memory bottleneck needs to be tackled. In an APU, the system RAM is shared with the GPU. The Steam Deck's biggest hidden bottleneck is often memory bandwidth, which is how fast data can move between the brain and the eyes. The next standard in low-power double data rate memory is expected to target speeds north of 10GB/s. This is absolutely crucial for 1080p or high-refresh-rate gaming on a handheld, meaning that Valve will need to opt for LPDDR6. There is also a heat problem, as high-speed memory usually generates significant amounts of heat. Valve is likely waiting for LPDDR6 to mature so they can get massive bandwidth gains without needing a dedicated heat pipe just for the RAM modules themselves.

Valve is still caring for their Steam Deck users

The software is picking up the slack

For the time being, Valve is attempting to make the old tech feel new by using Steam OS updates to squeeze more life out of the original hardware via better drivers and FSR (fidelity FX super resolution). These small quality of life changes mean that the software really and truly is picking up the slack, making many users of the Steam Deck not really notice a significant difference when comparing their Deck to alternatives on the market. While other alternatives rely on raw horsepower to brute force Windows, Valve relies on a lightweight Linux stack to make 15W feel like 25W, giving it a major competitive advantage.

However, eventually those who are still using a Steam Deck will absolutely demand a Steam Deck 2. Based on Valve's comments and AMD's roadmap, it's likely that the Steam Deck 2 will be coming in late 2026 or 2027. Right now, Valve is protecting the Steam Deck brand by refusing to iterate for the sake of sales. When the Steam Deck 2 finally arrives, it won't just be faster; it will change what we think is possible in a handheld overall.