Arm’s foundational technology is defining the future of computing. A future built by the greatest technology ecosystem in the world. A future built on Arm.
Cambridge, England
Based in United Kingdom
- Tweets 32,678
- Following 1,762
- Followers 101,321
- Likes 10,157
Congratulations to Unity China on the launch of Tuanjie Engine 1.11.0, featuring Arm Neural Super Sampling (NSS)!
Developers can now explore AI-powered image reconstruction to reduce GPU workloads and create richer gaming experiences with Arm Neural Technology.
Hear from Junbo Zhang, CEO of Unity China, on what this integration means for game development. ⬇️
1
2
11
850
Exciting to see @Vivo_GLOBAL continuing to push the boundaries of mobile innovation with the X500 Series!
Our collaboration with partners like vivo helps bring the next generation of intelligent, high-performance mobile experiences to life on Arm. 👏
Great moments never wait. ⚡
Meet vivo X500 Pro Max — built for everything you're about to create. ✨
#vivo #vivoX500Series #ZEISSCinematicsCreateFurther
1
2
19
2,321
👏Congratulations to @liquidai on the launch of LFM3, its new family of compact, open-weight models spanning text, vision, audio and decision-making!
We're proud to support LFM3 from day one, building on our collaboration with Liquid AI to advance efficient, on-device AI inference on Arm.
Early, provisional benchmarks of LFM3-1.7B (Q4_0) on Arm CPUs show a 3.1× speedup in prefill and 1.5× speedup in decode with SME2 enabled, achieving over 400 tokens/s prefill and over 50 tokens/s decode using just two CPU threads.
Introducing 𝙻𝙵𝙼𝟹: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture and new multimodal encoders, the LFM3 family is the ideal model suite for specialized intelligence.
> 6 model instances, across text, audio, vision, and omni modalities, generative and decision models.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱: text, 256k context window.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱: vision-language model with video-input support for spatial understanding
> 𝙻𝙵𝙼𝟹-𝙰𝚞𝚍𝚒𝚘-𝟸𝙱: tool calls, voice chat, reasoning in 18 languages.
> 𝙻𝙵𝙼𝟹-𝙾𝚖𝚗𝚒-𝟸𝙱: an omni model combines vision, audio, and language.
> 𝚍𝟷-𝟸𝙱 and 𝚍𝟷-𝙾𝚖𝚗𝚒-𝟸𝙱: decision models with typed outputs and calibrated probabilities in a single pass. Omni adds audio input.
> fine-tune on 1 GPU.
> a new tri-hybrid architecture
> architecture optimized for 14 processor classes across GPUs, CPUs, and NPUs.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱 is a beast. It performs competitively with Gemma 4 12B (6x larger), and Gemini 3.5 Flash Lite, and GPT-5.6 Luna, with 1.9B parameters.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱 is #1 on the Pareto frontier of device intelligence 𝚙𝚒𝚙𝚎𝚝𝚝𝚎.𝚕𝚒𝚚𝚞𝚒𝚍.𝚊𝚒
Every modality for specialized intelligence. Pick the one that fits your use case, customize it, and run it on-device.
1/ 🧵
5
10
93
5,485
The next measure of AI performance? Outcomes.
As agentic AI moves from individual models to multi-step workflows, performance increasingly depends on how efficiently the entire system works together.
Explore perspectives from leading companies across the industry on scaling agentic AI from cloud to edge. 💪 okt.to/V7I0ql
5
3
23
2,052
Making next-generation neural image compression practical on mobile! 📱
Arm SME2 helped Tencent Media Lab achieve 2× faster decoding in a key neural decoder module, reducing latency from 61 ms to 30 ms for 2K images on a reference smartphone.
See how we made it happen: okt.to/olhAIV
3
9
48
2,873
A strong software ecosystem is essential to bringing agentic AI to life.
See how open-source frameworks and tools come together on Arm AGI CPU to support agent orchestration, data retrieval, and tool execution in our multi-agent customer service demo.
🔗: okt.to/eSpGin
5
5
51
4,047
AI demand isn't slowing down - but the ability to supply enough compute is becoming a critical constraint.
Arm CMO Ami Badani joined @traderTVLIVE to discuss why power-efficient compute is essential to unlocking AI's potential across industries, and the opportunity ahead for Arm.
2
3
26
2,550
Scaling physical AI takes more than powerful hardware.
Hear from our Dermot O’Driscoll on @RoadtoAutonomy as he discusses how the Arm compute platform and ecosystem collaboration are helping accelerate the adoption of robotics and autonomous vehicles. 👇
piped.video/Hhe-gb6SrJQ?si=kq5k…
1
1
15
3,017
🎙️Tune in today from 10:30-10:45 a.m. PT as Ami Badani joins @traderTVLIVE to discuss Arm’s role in the AI ecosystem, the growing demand for power-efficient compute, and what’s next for AI.
🔴Watch live: nitter.siddz.com/i/broadcasts/1nxeLMAWb…
1
1
18
2,186
Agile Robots SE has unveiled Agile ONE SN, a new humanoid reference platform designed to help researchers and developers create, train and validate robot capabilities.
Powered by NVIDIA Jetson Thor, built on Arm Neoverse V3AE CPUs, Agile ONE SN delivers onboard compute for real-time control and reasoning, while bringing simulation, training and deployment workflows together to help reduce setup time from weeks to days.
Congrats to Agile Robots and @nvidia on the announcement! 👏okt.to/6WqOFm
1
2
20
2,449
From six months to as little as 14 days. That's the potential leap in AI factory deployment speed with NVIDIA's new DSX POD.
Powered by NVIDIA Vera Rubin infrastructure, built on Arm-based Vera CPUs, DSX POD is helping make the next generation of AI infrastructure faster to deploy and easier to scale.
Congrats, NVIDIA!
AI builders need computing capacity they can put to work sooner.
We introduced DSX POD at NVIDIA GTC Berlin: AI factories integrated and tested before delivery, enabling deployment at prepared sites in as little as 14 days.
1
29
2,735
As AI agents become more autonomous, security needs to be built into the infrastructure powering them.
NVIDIA's new OpenShell Reference Architecture helps make this possible, with Arm-based Vera CPUs powering agent execution and BlueField DPUs supporting isolation and policy enforcement.
Jensen introduces the OpenShell Reference Architecture to help businesses set boundaries for AI agents and stop them when they step outside those boundaries.
Securing agents around the world. 🔒
2
4
28
2,695
What does it take to accurately benchmark CPU performance for agentic AI?
Our white paper, developed in collaboration with @Meta, @fujitsulabs, and @OpenAI, explores a framework for measuring CPU performance across representative agentic AI workloads and the broader software stack.
🔗: okt.to/tXkKxq
3
8
43
3,057
F1 pit stops 🤝 Safer patient handovers. Who would've thought?
@GreatOrmondSt took inspiration from pit crews to improve patient handovers.
As Arm technology helps @AstonMartinF1 deliver their fastest pit stops, it gets us thinking: where else could advances in physical AI make a life-changing difference? okt.to/a7uLgd
3
11
1,969
From server-level performance to rack-level impact.
Independent testing from @Signal_65 highlights how Arm AGI CPU performs across key agentic AI workloads, demonstrating the importance of CPU performance as AI infrastructure scales.
Take a closer look: okt.to/LPaNF3
2
4
35
3,019