Hanwha Vision's newly launched Wisenet 9 system-on-chip separates camera processing into two independent neural processing units (NPUs), one for image enhancement and another for behavioral analytics, delivering three times the inference throughput of its predecessor while maintaining full NDAA compliance. The architecture represents a significant engineering shift in how surveillance cameras handle artificial intelligence workloads, particularly for US government agencies and contractors facing strict procurement restrictions.

What Makes Hanwha's Dual NPU Design Different From Single-Processor Cameras?

Traditional surveillance cameras force a single processor to time-share between multiple tasks, creating a fundamental trade-off: either prioritize image quality or prioritize analytics, but rarely both simultaneously. Hanwha's approach eliminates this compromise by dedicating separate neural processing units to distinct functions. The first NPU handles image enhancement tasks including extreme wide dynamic range (WDR) processing, low-light noise reduction, and advanced tone processing. The second NPU independently manages behavioral analytics such as object classification, attribute extraction, and anomaly detection.

This parallel processing architecture means cameras running Wisenet 9 can perform object detection and high-quality image capture at the same time without sacrificing performance in either domain. All processing happens at the edge, meaning inside the camera itself, rather than requiring data to be sent to cloud servers for analysis. This edge-based approach reduces bandwidth consumption and eliminates the latency gap that cloud-dependent systems introduce, a critical advantage for security applications where real-time response matters.

How Does NDAA Compliance Create Market Opportunity for Hanwha?

The National Defense Authorization Act (NDAA) Section 889, enacted in 2019, explicitly prohibits US federal agencies and contractors from procuring video surveillance equipment from five named Chinese manufacturers: Hikvision, Dahua, Huawei, ZTE, and Hytera. This ban extends to any equipment built on components from these companies. Hikvision, the world's largest surveillance camera manufacturer, cannot legally be sold to US government buyers, creating a significant market gap.

Hanwha Vision's Wisenet 9 is not built on any of the banned chipsets, as confirmed by the company's NDAA compliance documentation. Security integrators now position Hanwha as the most direct hardware replacement path for organizations replacing Hikvision systems under compliance mandates. According to a June 2026 assessment by Coram, a video surveillance software provider, multiple 2026 procurement guides for NDAA-compliant cameras list Hanwha as the leading alternative with equivalent or superior sensor performance.

The timing of Hanwha's restructuring amplifies this opportunity. On August 3, 2026, South Korean conglomerate Hanwha Group formally launched Hanwha Machinery & Services Holdings (Hanwha M&S), a standalone holding company that consolidates 57 technology and lifestyle affiliates under unified governance. Hanwha Vision, which manufactures the Wisenet 9 cameras, is now the centerpiece of this new entity. The company generated approximately 6 trillion Korean won (roughly $4.2 billion USD) in sales across its portfolio as of last year, with total assets of 11.3 trillion Korean won (approximately $7.8 billion USD).

What Technical Certifications Support Hanwha's Government Market Position?

Beyond NDAA compliance, Hanwha Vision has secured multiple security and governance certifications that matter to government procurement officers. The Wisenet 9 chipset itself carries FIPS 140-3 Level 3 security certification, a federal standard that validates cryptographic modules and their security properties. Additionally, Hanwha Vision holds ISO/IEC 42001 certification, the first international standard for AI management systems, which signals that its AI deployments are independently auditable against a defined governance framework.

These certifications reduce procurement risk for government agencies. Rather than conducting their own security audits, agencies can reference third-party validation. The combination of NDAA compliance, FIPS certification, and AI governance standards creates a defensible procurement position that competitors without this certification stack cannot easily replicate.

How Is the Broader Semiconductor IP Market Supporting NPU Development?

Hanwha's dual NPU architecture reflects a broader industry trend toward specialized processor designs. The global Semiconductor Intellectual Property (IP) market, which supplies pre-designed and verified processor cores and subsystems to chip manufacturers, was valued at $9.2 billion in 2025 and is projected to reach $24.3 billion by 2035, growing at a compound annual rate of 10.2 percent.

Processor IP, which includes CPU, GPU, and neural processing unit designs, dominated the design IP segment with a 46.2 percent market share. This growth reflects increasing demand for advanced processor architectures used in smartphones, artificial intelligence chips, automotive electronics, data centers, and connected devices. Asia Pacific held the largest regional share at 54.3 percent of the global market in 2025, representing approximately $5.0 billion.

The rise of specialized NPUs reflects a fundamental shift in chip design philosophy. Rather than building general-purpose processors that handle all workloads, manufacturers increasingly develop dedicated neural processing units optimized for specific AI inference tasks. This specialization allows better performance per watt of power consumed, a critical metric for surveillance cameras that often run continuously on limited power budgets.

Steps to Understanding NPU Architecture in Modern Surveillance Systems

• Separate Processing Domains: Modern surveillance NPUs divide workloads into specialized domains such as image enhancement, object detection, and behavioral analytics, allowing each domain to optimize for its specific task rather than compromising across multiple functions.
• Edge Processing Benefits: By performing AI inference locally inside the camera rather than sending data to cloud servers, edge-based NPUs reduce bandwidth requirements, eliminate cloud latency, and maintain privacy by keeping sensitive video data on-premises.
• Compliance and Certification: Government procurement increasingly requires third-party security certifications such as FIPS 140-3 and governance standards like ISO/IEC 42001, which validate that AI systems meet defined security and auditability requirements.
• Performance Metrics: NPU performance is measured in TOPS (tera operations per second), with Wisenet 9 delivering three times the inference throughput of its predecessor, enabling simultaneous high-quality image capture and real-time analytics.

What Does the Hanwha M&S Restructuring Mean for Investors and the Market?

The creation of Hanwha M&S is explicitly designed to address what analysts call the "conglomerate discount," the gap between a diversified conglomerate's market price and the sum of its parts. SK Securities analyst Choi Kwan-soon estimated Hanwha Corp.'s discount to net asset value at 64.1 percent before the restructuring, elevated partly because of governance uncertainty created by the pending spin-off itself. The analyst maintained a buy rating, forecasting a share price rebound once the restructuring is complete and governance uncertainty dissipates.

Hanwha M&S is scheduled for its Korea Exchange (Kospi) debut on August 25, 2026. The company's portfolio extends beyond surveillance cameras. Hanwha Semitech, the company's semiconductor equipment division, manufactures chip mounting and deposition equipment used in semiconductor assembly and testing. The company claims the number-one global market share position in its medium-speed chip mounter category and operates seven regional hubs and more than 35 authorized retailers across the US, Germany, China, India, and Vietnam.

This combination of surveillance camera technology and semiconductor manufacturing equipment positions Hanwha M&S across a widening band of the semiconductor value chain. As demand for specialized AI chips like NPU-based surveillance processors grows, the company benefits from both the end-product market (cameras) and the manufacturing equipment market (tools to produce those chips).

How Are Competing NPU Architectures Approaching the Same Problem?

While Hanwha focuses on surveillance-specific NPU design, other companies are pursuing broader neural processing strategies. Singapore-based startup Acrab, backed by CATL (the world's largest electric vehicle battery manufacturer), unveiled the Agent Box, a mini PC powered by a proprietary chip called G≡LIX 1 that combines a 20-core Arm CPU, a GPU rated at 3 TFLOPS (trillion floating-point operations per second), and a multicore neural processing unit tuned for large language model inference.

The G≡LIX 1 reaches 700 TOPS of processing power while maintaining a compact form factor suitable for desktop deployment. In internal testing, the company recorded a prefill rate of 1,416.8 tokens per second under a Gemma 26B A4B configuration, representing a claimed 7.5 times improvement in prefill speed over conventional consumer hardware like the Apple Mac Mini M4 Pro.

"Our goal is to give device makers and developers the foundation to bring agentic intelligence into many different products and environments," said Dr. Ken Phua, CEO of Acrab.

Dr. Ken Phua, CEO of Acrab

These competing approaches highlight a broader industry pattern: NPUs are becoming specialized for specific use cases rather than general-purpose. Hanwha's dual NPU design optimizes for surveillance analytics, while Acrab's G≡LIX 1 optimizes for large language model inference. As the semiconductor IP market grows at 10.2 percent annually, we can expect further specialization and proliferation of task-specific neural processing architectures across industries.

The convergence of NDAA compliance requirements, specialized NPU architectures, and growing government procurement demand creates a distinct market opportunity for companies like Hanwha that can deliver both technical performance and regulatory alignment. As Hanwha M&S prepares for its public market debut, investors will be watching whether the company's dual focus on surveillance technology and semiconductor manufacturing equipment can capture value across both the end-product and supply-chain layers of the AI chip ecosystem.