NVIDIA, AMD and Intel aim for maximum power at CES 2025

There was no doubt that NVIDIA’s RTX 5000 GPUs would be one of the biggest stories at CES 2025, and I figured Intel and AMD would arrive with new hardware of their own. But I didn’t expect that each of these companies would, in their own way, push the pedal all the way when it came to power for their chip designs. After all, we’ve spent the last few years covering AI PC processors that have been focused on efficiency more than raw performance.

While NVIDIA RTX 5000 GPUs seem to deliver the performance jump we’ve come to expect over its 2022-era cards, AMD is also redefining what’s possible for mobile workstations with its Ryzen AI Max chipswhich combine powerful graphics with a lot of built-in memory. Intel isn’t standing still either – it’s finally moving Arrow Lake into the high performance and gaming arena with its own Core Ultra 200HX chipswhich can reach up to 24 cores and 5.5 GHz speeds.

I’m not just talking about power in the pure performance sense. NVIDIA’s $1,999 RTX 5090 requires a 1,000-watt power supply to function and uses up to 575 watts. Meanwhile, Ryzen AI Max chips can consume up to 120 watts. Intel’s Core Ultra 200HX chips go up to 120 watts. Clearly, none of this hardware is for anyone worried about their energy bills or the potential battery life of a laptop.

So what do you get for all that power consumption? AMD says the RTX 5090 will deliver roughly twice the performance of its previous flagship, the $1,499 RTX 4090. In 4K Cyberpunk 2077 demo with full ray tracing, the 4090 ran around 108 fps, while the 5090 hit 240 fps. This frame count is a bit controversial, however, as the RTX 5090’s DLSS 4 AI upscaling generates three frames for every natively rendered frame. The end result may look smoother to most people, but some gamers may question the integrity of the so-called fake frames.

These are the same AI-generated frames that allow NVIDIA to claim that the $549 RTX 5070 can be as powerful as the 4090. That may be true when it comes to sheer framerate, but it certainly won’t be for rasterized performance without DLSS 4.

AMD’s Ryzen AI Max chips don’t aim for the same kind of graphical heights as NVIDIA’s new GPUs, but they still stand out for the sheer amount of hardware they pack. The top Ryzen AI Max+ 395 features 16 CPU Zen 5 cores, 50 TOPS AI performance and 40 RDNA 3.5 GPU compute units. According to AMD, it should match Apple’s 14-core M4 Pro chip (and even faster in the Vray benchmark) and is 2.6 times faster in 3D rendering than Intel’s Core Ultra 9 288V.

c interview with AMD CVP and Product CTO Joe Macrihe told Engadget that the success of Apple Silicon is a major reason for Ryzen AI Max’s existence. “What Apple has shown is that consumers don’t care what’s inside the box,” he said. Macri later noted, “I’ve always known, as we’ve been building APUs and I’ve been pushing for this big APU forever, that I can build a system that’s smaller, faster, and I can give much higher performance in same time power.”

AMD too briefly reviewed his RDNA 4 graphs at CES, though at this point it’s clearly targeting the mid-range, not NVIDIA’s RTX 5090. Notably, AMD will debut a new AI-powered scaling technology in RDNA 4 GPUs, FidelityFX Super Resolution 4 (FSR 4). This should finally give AMD a way to directly compete with NVIDA’s DLSS, which has been looking better than earlier versions of FSR for years. The first RDNA 4 cards, the Radeon RX 9070 and 9070 XT, will arrive sometime in the first quarter.

Intel’s presence at CES 2024 was more muted than the competition, but loyalists will likely appreciate the new Core Ultra 200HX chips. Although they’re downgrading NPU performance from recent AI PC hardware (12 TOPS down from 48 TOPS), the Core Ultra 9 285HX looks like a 24-core beast. It will be interesting to see how it competes against AMD’s Ryzen AI 300 hardware, though it probably won’t stand a chance against the Ryzen AI Max when paired with a discrete GPU.