Back to News
ProductAI Understanding briefing

VeriSilicon unveils low‑power wearable subsystem for next‑gen AI and AR glasses

VeriSilicon announced a new low‑power, high‑performance subsystem solution that bundles its Nano‑XR IP series with FLEXA architecture and DECNano compression to streamline AI, vision and display processing for upcoming smart‑glass devices.

4 min readRead the linked source
Source-page capture accompanying VeriSilicon unveils low‑power wearable subsystem for next‑gen AI and AR glasses
Source referenceSource recorded
Publisher
manilatimes.net
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Key terms

Memory (Agent Memory)
Stored context an AI agent uses across steps or sessions to improve continuity.
Latency
The time between sending a request and receiving the model's output.

What happened

VeriSilicon introduced a wearable subsystem solution built around its newly announced Nano‑XR IP family and supporting technologies. The offering combines image‑signal, video, neural‑network, graphics and display processing IPs with system‑level optimizations such as the FLEXA communication interface and DECNano compression. The company says the solution targets AI‑enabled augmented‑reality glasses, promising reduced power draw, lower memory bandwidth, and faster for tasks like eye‑tracking, gesture recognition, image enhancement and 3‑/6‑DoF display rendering.

VeriSilicon’s press release describes a subsystem architecture that integrates a suite of its low‑power IP blocks—GPU, NPU, VPU, DSP, ISP and display processing—into a single solution for AI/AR glasses. The Nano‑XR series includes four specific IPs: ISPNanoUltra‑XR for high‑quality photo/video capture, ISPPico‑XR for infrared sensor tasks such as eye‑tracking, DCNanoUltra‑XR for low‑ display pipelines supporting 3‑DoF/6‑DoF applications, and GCNanoOptimaV‑XR for enhanced graphics projection and memory efficiency.

At the system level, VeriSilicon highlights two proprietary technologies. The FLEXA interface enables segment‑level data exchange between IP blocks using on‑chip SRAM buffers, which reduces reliance on external PSRAM or DDR memory and cuts . DECNano compression provides visually lossless compression of image and video streams, further lowering external memory bandwidth and capacity requirements.

The company notes support for multiple operating systems—including RTOS, Linux and Android—allowing OEMs flexibility in software stacks. It also mentions ongoing collaborations with customers on custom smart‑glass chip designs that incorporate the Nano‑XR IPs, though no specific partners are named.

VeriSilicon positions the solution as a response to the growing demand for AI‑enabled wearables that must balance high computational workloads with strict power budgets. By offering a configurable IP portfolio and subsystem‑level optimizations, the firm claims it can help designers achieve “smart‑glass experiences” that were previously limited to larger, power‑hungry devices.

Source details: manilatimes.net

Why it matters

The announcement signals a maturing ecosystem for AI‑powered smart‑glass hardware, where integrated, low‑power silicon can enable richer on‑device AI without draining battery life. By bundling multiple processing blocks and optimizing data movement, VeriSilicon aims to lower the engineering effort and cost for OEMs developing next‑generation AR wearables. If adopted, the solution could accelerate the rollout of glasses that support real‑time AI vision, voice interaction and high‑fidelity graphics, expanding the market beyond niche enterprise tools toward broader consumer use. However, the press release does not disclose pricing, availability dates, or which customers have already committed to the technology, leaving the commercial impact uncertain.

AI‑driven AR glasses require on‑device processing for tasks such as real‑time object detection, gesture recognition and high‑resolution display rendering. Traditional approaches often rely on separate chips for each function, leading to higher power consumption and increased board complexity. VeriSilicon’s integrated subsystem promises to consolidate these functions, potentially reducing bill‑of‑materials costs and extending battery life.

The inclusion of low‑power NPU and DSP IPs means that even modest AI workloads—like wake‑word detection or basic image enhancement—can run continuously without draining the device, a key hurdle for consumer adoption of AR glasses.

If the FLEXA communication architecture and DECNano compression deliver the claimed reductions in external memory accesses, designers could use smaller, cheaper memory packages, further lowering device cost and size—critical factors for wearable form‑factors.

The announcement also reflects a broader industry trend toward domain‑specific silicon for wearables, where companies aim to differentiate through tightly integrated, power‑optimized solutions rather than generic processors.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Document Size:128K tokens
Needle Placement Depth (Location in document):50% into text
Attention Context Buffer Map:
Target Fact (50%)
Equivalent Pages~320Standard book pages
Retrieval Accuracy99.9%Needle recall score
RAM / KV Cache5.1 GBMemory overhead
Prompt CachingActive~80% discount on reuse
Core takeaway: Million-token context windows allow querying whole codebases or legal archives in one prompt. However, KV cache memory scales with context length, making prompt caching crucial for real-time production.
Interactive Concept Check+10 Points
AI Models Explained Quiz

Which component of an AI application is the machine-learning model itself?

What to watch next

Watch for OEM partnerships that confirm silicon adoption, details on tape‑out schedules, and any announced pricing or volume‑discount structures. Follow how the FLEXA and DECNano technologies perform in real‑world prototypes, especially regarding power savings versus competing solutions. Industry analysts will also track whether the Nano‑XR IPs enable new form‑factors or longer battery life for AR glasses, which could influence the competitive dynamics among wearable chipmakers.

OEM announcements confirming the use of VeriSilicon’s Nano‑XR IPs in upcoming AR glasses, especially any high‑profile consumer or enterprise products.

Technical validation results from prototype devices that measure actual power consumption, and memory bandwidth savings compared with existing solutions.

Details on the timeline for tape‑out and mass production, as well as any disclosed pricing models or volume commitments.

Competitive responses from other wearable‑chip vendors, such as Qualcomm, MediaTek or Apple, which may accelerate their own integrated AI/AR silicon roadmaps.

Related guides & quizzes

Found this useful?