AI News AI资讯 6h ago Updated 1h ago 更新于 1小时前 49

First Xiaomi, then the world: why Arm might give phone gaming a huge graphics boost 先小米,后全球:Arm为何可能给手机游戏带来巨大图形提升

Arm has developed the Mali G2-Ultra NX GPU, featuring a dedicated AI gaming graphics accelerator after five years of development, enabling hardware-accelerated neural graphics techniques previously exclusive to PC GPUs. The chip supports AI upscaling, frame generation, and ray reconstruction (dubbed "Neural Super Sampling," "Neural Frame Rate Upscaling," and "Neural Denoising"), potentially delivering up to 4x framerate improvements while consuming only 1.5 watts for the GPU. Xiaomi launched the Arm推出Mali G2-Ultra NX GPU,内置AI游戏图形加速器,性能较前代提升85%,功耗仅1.5瓦 技术对标Nvidia DLSS,支持AI超采样、帧生成、光线重建等"神经图形"技术,最高可实现4倍帧率提升 小米18 Fold首发搭载该芯片,首批适配游戏包括《Where Winds Meet》《Infinity Nikki》《Arena Breakout: Infinite》 技术将扩展至平板、Chromebook等设备,2027年初推出帧生成,去噪技术随后跟进 AI渲染替代传统渲染成为趋势,移动端游戏画质与性能瓶颈有望突破

72
Hot 热度
68
Quality 质量
70
Impact 影响力

Analysis 深度分析

TL;DR

  • Arm has developed the Mali G2-Ultra NX GPU, featuring a dedicated AI gaming graphics accelerator after five years of development, enabling hardware-accelerated neural graphics techniques previously exclusive to PC GPUs.
  • The chip supports AI upscaling, frame generation, and ray reconstruction (dubbed "Neural Super Sampling," "Neural Frame Rate Upscaling," and "Neural Denoising"), potentially delivering up to 4x framerate improvements while consuming only 1.5 watts for the GPU.
  • Xiaomi launched the first device with this technology—the Xiaomi 18 Fold in mainland China—packaging the GPU inside its custom Xring O3 chip, with the Xiaomi Pad 9 Pro Max tablet achieving 3.2K at 120fps.
  • Arm's EVP Chris Bergey confirmed the technology will expand beyond Chinese phones to tablets, Chromebooks, and potentially Google devices, with frame generation arriving in early 2027 and denoising following shortly after.
  • Three games are confirmed to support the technology at launch—Where Winds Meet, Infinity Nikki, and Arena Breakout: Infinite—with Unreal Engine and Unity providing integration support.

Why It Matters

Arm's entry into AI-accelerated mobile graphics represents a paradigm shift, bringing PC-class rendering techniques like DLSS and frame generation to the mobile ecosystem for the first time. This development could fundamentally change mobile gaming by enabling console-quality visuals at sustainable power levels, making high-fidelity gaming accessible on mainstream devices rather than niche gaming phones. For the industry, it signals Arm's aggressive push to compete directly with Nvidia's dominance in AI-enhanced graphics, potentially reshaping the mobile GPU landscape.

Technical Details

  • Architecture: The Mali G2-Ultra NX integrates a dedicated AI gaming graphics accelerator within Arm's GPU shader pipeline, described as bringing "Tensor-like capabilities" directly into the graphics processing unit.
  • Performance Claims: Arm reports an 85% performance increase over Xiaomi's previous GPU and a 14% improvement over Arm's prior flagship in non-AI gaming scenarios. When all three neural graphics techniques are combined, up to 4x framerate improvement is claimed.
  • Power Efficiency: The entire phone with these chips consumes no more than 2.5 watts, with only 1.5 watts allocated to the GPU—enabling sustained gaming sessions (e.g., 1080p at 60fps for three hours, compared to 30 minutes previously).
  • Neural Graphics Suite: Three core techniques are implemented: Neural Super Sampling (AI upscaling to render higher-resolution images), Neural Frame Rate Upscaling (generating intermediate frames between existing ones), and Neural Denoising (ray reconstruction for advanced lighting). Frame generation is slated for early 2027, with denoising arriving soon after.
  • Engine Support: Unreal Engine and Unity are confirmed participating game engines that will facilitate developer integration of Arm's neural graphics features.

Industry Insight

  • Mobile Gaming Renaissance: The convergence of AI upscaling and frame generation on mobile devices will likely trigger a new era of mobile gaming, where flagship phones can deliver experiences previously requiring dedicated gaming hardware or consoles—potentially expanding the addressable market for premium mobile titles.
  • Arm vs. Nvidia Competition: Arm is explicitly following Nvidia's DLSS playbook, suggesting a strategic move to capture the mobile AI-graphics market before Nvidia can establish a foothold. This could pressure Nvidia to accelerate mobile partnerships or develop competing mobile-focused solutions.
  • Developer Integration Timeline: Since games won't automatically benefit and require explicit engine-level integration, early adopters like Where Winds Meet and Infinity Nikki will have a competitive advantage. Developers should prioritize Unreal Engine and Unity updates to leverage these capabilities as they roll out through 2027.

TL;DR

  • Arm推出Mali G2-Ultra NX GPU,内置AI游戏图形加速器,性能较前代提升85%,功耗仅1.5瓦
  • 技术对标Nvidia DLSS,支持AI超采样、帧生成、光线重建等"神经图形"技术,最高可实现4倍帧率提升
  • 小米18 Fold首发搭载该芯片,首批适配游戏包括《Where Winds Meet》《Infinity Nikki》《Arena Breakout: Infinite》
  • 技术将扩展至平板、Chromebook等设备,2027年初推出帧生成,去噪技术随后跟进
  • AI渲染替代传统渲染成为趋势,移动端游戏画质与性能瓶颈有望突破

为什么值得看

Arm将PC级图形技术(DLSS类方案)引入移动端,标志着手机/平板游戏画质迎来质变。这对游戏开发者、芯片厂商和移动设备用户均有重大影响,预示着移动端AI图形加速将成为行业新标准。

技术解析

  • Mali G2-Ultra NX架构:Arm五年研发成果,集成AI游戏图形加速器,支持Neural Super Sampling(神经超采样)、Neural Frame Rate Upscaling(神经帧率上采样)、Neural Denoising(神经去噪)三大核心技术,整机功耗控制在2.5瓦以内(GPU仅1.5瓦)。
  • 性能表现:相比小米此前使用的GPU性能提升85%,较Arm上一代旗舰在非AI游戏场景提升14%。三项技术同时启用时最高可实现4倍帧率提升,但帧生成会带来一定延迟。
  • 设备适配:小米18 Fold手机首发,小米Pad 9 Pro Max平板可实现3.2K分辨率120fps。技术不仅限于旗舰机型,Premium和Pro配置也将支持。
  • 游戏引擎支持:Unreal Engine和Unity等主流引擎已参与适配,开发者需在游戏内集成相关功能,玩家无法通过简单开关在任何游戏中启用。
  • 时间规划:首批游戏主要支持超采样技术,帧生成预计2027年初推出,去噪技术随后跟进。

行业启示

  • 移动端游戏进入AI图形时代:Arm复制Nvidia DLSS路径,表明AI加速图形渲染已从PC端下沉至移动端,未来1-2年将成为旗舰设备的标配能力。
  • 芯片厂商竞争格局变化:Arm通过差异化AI图形能力与高通、联发科形成区隔,同时与Nvidia、AMD在移动端展开间接竞争,芯片厂商需加速AI图形技术布局。
  • 游戏开发范式转变:开发者需重新考虑移动端游戏的渲染策略,AI超采样和帧生成可能成为新的性能优化标准,推动移动端游戏向主机级画质迈进。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Gaming 游戏 Chip 芯片 GPU GPU Product Launch 产品发布