The VideoX Appliance NVR’s (formerly SHIELD | ENT) deliver enterprise class performance and protection and advanced video motion detection and analytics at breakthrough prices. All of the tests were performed with server-side video motion detection (VMD) with metadata collection enabled using Milestone XProtect Corporate 2019 R3 to ensure latest performance numbers.
Arxys Hardware Accelerated Video Decoding delivers breakthrough performance at a much lower price. Arxys’ Hardware Acceleration technology supports more full HD cameras with server-side Video Motion Detection for 24/7 Video recording with forensic-grade zero frame loss and video analytics built-in at no extra cost.
Tests compared two HPE DL360 G10 Xeon servers (1x Gold 5222 & 1x Silver 4208 CPU) against the The VideoX Appliance (formerly SHIELD | ENT) R12E Server (Xeon-E 2176 CPU). Memory, SSD’s, NIC, OS, Settings, software and storage equal in all systems. The performance differences are isolated to the performance characteristics of the CPU’s themselves and highlight how much better the Xeon-E CPUS’s process video and video analytics.
The VideoX Enterprise NVRs with Xeon-E CPU delivered 325% more total cameras at 440% lower cost per camera. Three times as many cameras with full video motion analytics on a system that cost 40% less.
Using the h.264 codec the VideoX server delivered impressive performance gains over the higher cost HPE system with Scalable CPU servers. When using the h.265 codec the VideoX server achieved 15% better performance. Where the HPE Servers using Scalable CPUs struggled to keep up with the enhanced video decoding demands of the h.265 CODEC and lost 25% of the camera performance.
Using a Compute Analytics Price (CAP) to demonstrate the system cost per camera analytics compute. In essence the cost per camera. Lower is better.
(System Cost / Cameras) = CAP
- VideoX R8M (Xeon-E) MSRP $6,595 / 325 cameras = 20 CAP
- HPE DL360R10 (Silver 4208 CPU) MSRP $8,942 / 100 cameras = 89 CAP
- HPE DL360R10 (Gold 5222 CPU) MSRP $9957 / 75 cameras = 133 CAP
Video streams used in this test were configured at 30 frames per second (FPS). Video streams were configured to use both the H.264 and H.265 video codecs. All tests were performed with 24/7 continuous recording with full Video Motion Detection and Metadata collection enabled for all videos to stress the overall system and the find the peak processing point of each system. Recorded video was purged after one hour of retention. Performance numbers were recorded after the system started to delete old video to further simulate a real-world scenario.