Codensity™ G5

ASIC Video Processing Unit

The Codensity G5 architecture uniquely combines on-chip AV1, HEVC, and H.264 video encoding and AI processing engines to deliver scalability for metaverse, live streaming, and interactive applications.


G5 Video Processing Unit

8K Video Transcoder

Codensity G5 performance enables 8K live video encoding using AV1, HEVC, and H.264 video codec standards. Advanced codecs like AV1 and HEVC deliver superior quality with up to a 60% reduction in bitrate. Codecs like AV1 require 10x the processing power limiting the scalability of software encoders. This is the advantage of Codensity G5, making 8K high resolution in real-time possible for the first time.

AI Engine

Two Deep Neural Network engines capable of 18 Trillion Operations per Second (TOPS) enable object detection, classification, and segmentation to provide additional data to the encoding engine for image quality improvement and content-adaptive rate control for advanced performance and functionality. Seamlessly integrated for ROI encoding and background replacement. Additional features to be released.


Video applications bring I/O-intensive requirements and benefit from trade-offs in read/write performance, endurance, capacity, security, and data integrity. The Codensity G5 is based on programmable microprocessor architecture that allows customers to optimize the Flash Translation Layer (FTL) and pipeline processing to meet your application and system priorities.

PCIe 4.0

Today’s peripheral I/O performance is often limited by the PCIe bus interface between enterprise servers and connected modules. Codensity G5 supports the latest PCIe 4.0 standard, delivering 2X the throughput compared to the slower PCI 3.0 standard. This enables a 1RU NETINT Quadra VPU-powered server to encode 160 1080p60 live streams with broadcast quality.

20x YOUR Density Over Software Video Encoding

Codensity G5 Architecture

Codensity G5 ASIC

Codensity G5

NETINT Codensity G5 ASIC

NETINT Codensity G5 enables cloud data centers, edge computing companies, and content providers to deploy scalable high performance applications, while minimizing their data storage and video processing costs. We are focused on moving modern video workflows from software to hardware in the data center using purpose built ASIC technology.

We exist to reduce the TCO of hyperscale video platforms and services by converting the world from software-based video encoding to ASIC-based video encoding so that our customers can improve scalability, lower their carbon footprint, and operate more economically. 

Using ASICs you can reach environmental sustainability by reducing CO2 more than 80 times compared to software running in a data center.

Using Codensity G5, you can achieve a 20x or higher density and performance gain compared with software-based video encoding running on CPU. This leads to a corresponding TCO savings and vastly improved environmental impact by way of reduced CO2 emissions.

NETINT Codensity ASIC’s provide a massive cost advantage. Many customers report operational costs as low as $150 per channel for a 1080p broadcast live stream. Unlike software encoding where the operational costs for advanced codecs such as HEVC or AV1 greatly increase TCO, using Codensity ASICs, whether H.264, HEVC, or AV1, the operational cost advantages remain the same.