MulticoreWare collaborates with Arm to optimize and advance x265 video encoding on AWS Graviton4
SAN JOSE, Calif. -- October 1, 2024 -- MulticoreWare Inc., a pioneer in video encoding solutions and performance optimization services, announced a significant leap forward in x265 video encoding performance on Arm®-based AWS Graviton4 processors. In collaboration with Arm, MulticoreWare has delivered notable improvements in encoding speed, multi-stream density, and cost-efficiency, on Arm-based AWS Graviton4 as a leading option for cloud-based, high-density video encoding workloads.
By harnessing the Neon and SVE/SVE2 instruction sets from Arm, the x265 video codec has been optimized to deliver notable frames per second (FPS) improvements across different encoding presets, achieving gains of 29% in Ultrafast and 55% in very fast/faster modes.
These enhancements, targeting crucial encoding processes, such as motion estimation and DCT transformations, have resulted in substantial performance boosts, especially in multi-stream encoding situations.
"Video-encoding continues to be a key component for delivering high quality, efficient video compression in streaming, social media, and video conferencing applications in the cloud," said Bhumik Patel, Director of Software Ecosystem, Infrastructure Line of Business, Arm. "Our collaboration with MulticoreWare to add support for the latest Arm instruction sets into the libx265 codec shows the unique performance and cost-efficiency benefits that are possible when building on Arm."
AWS Graviton4 excels in x265 multi-stream encoding, outperforming x86-based processors by 50.24% in Ultrafast, 44.50% in very fast, and 52.53% in medium presets.
Shivakumar Narayanan, VP and GM, Media and Entertainment Business Unit at MulticoreWare, said: "Our joint efforts with Arm have brought remarkable improvements to x265 encoding on AWS Graviton4. This enables our customers to handle larger encoding workloads with fewer resources, driving significant operational savings while maintaining exceptional video quality."
MulticoreWare and Arm are advancing x265 encoding through ongoing innovations and future optimizations, enhancing functions and adapting to emerging hardware architecture.
This partnership aims to not only deliver immediate performance improvements but also ensure that x265 evolves in tandem with technological advancements, such as new instruction sets, multi-core scalability, and energy-efficient processing.
All these improvements will be available in the upcoming x265 v4.0 that will be released shortly.
About MulticoreWare
MulticoreWare delivers software IP Solutions and Engineering Services serving customers with solutions like HW Platform Compilers & Toolchains, SDK Libraries, Video codecs, and Algorithm & Data Engineering using vision & non-vision sensors (Radar, LiDAR, IMU, GPS). MulticoreWare's solutions are used in Automotive, Surveillance, Smart Health, IoT, Industrial, Robotics, Smart Cities. Their industry-leading video codec products (x266/x265/Ultraziq) are used in live streaming or VOD services across many broadcast customers.
Related Semiconductor IP
- Simulation VIP for Ethernet UEC
- Bluetooth® Low Energy 6.2 PHY IP with Channel Sounding
- Simulation VIP for UALink
- General use, integer-N 4GHz Hybrid Phase Locked Loop on TSMC 28HPC
- JPEG XL Encoder
Related News
- MulticoreWare demonstrates high quality 4K 10-bit real-time HEVC video encoding with x265
- Allegro DVT Introduces the Industry First Real-Time AV1 Video Encoder Hardware IP for 4K/UHD Video Encoding Applications
- SingMai offer universal encoding and decoding of analogue video
- MainConcept and Fraunhofer IIS collaborate on MPEG-H Audio and xHE-AAC encoding for video and audio streaming services
Latest News
- Mixel MIPI IP Integrated into Automotive Radar Processors Supporting Safety-critical Applications
- GlobalFoundries and Navitas Semiconductor Partner to Accelerate U.S. GaN Technology and Manufacturing for AI Datacenters and Critical Power Applications
- VLSI EXPERT selects Innatera Spiking Neural Processors to build industry-led neuromorphic talent pool
- SkyWater Technology and Silicon Quantum Computing Team to Advance Hybrid Quantum-Classical Computing
- Dnotitia Revolutionizes AI Storage at SC25: New VDPU Accelerator Delivers Up to 9x Performance Boost