As the name implies, CG uses computers to generate or enhance still or moving images for a multitude of uses that include art projects, architecture, print media, simulations, movies, videos, and video games. CG made its movie debut in the early 70s and became more mainstream in the 80s. Music videos began incorporating CG around the same time, and full-length feature films were being entirely created using CG content by 1995.[1] Today, a growing number of industries, applications, and use cases are adopting CG for an ever-growing array of purposes. This blog focuses on media and entertainment, and particularly on rendering high-resolution, high-quality moving images or animations that are often delivered digitally to a spectrum of end-user devices that range from desktop and laptop computers to TVs, tablets, and phones. I'll talk a little about rendering and streaming, and then show you some examples of why 4th Gen AMD EPYC™ processors are ideal for media and entertainment workloads.
Rendering and Streaming
The process of converting digital assets such as 3D models to finished still or animated images is called rendering. Today's high-end 4K and 8K renderings can require almost 8.3 million pixels per still image and over 33 million pixels per animated frame. Rendering places a tremendous burden on computer processors that are being asked to handle larger resolutions and deliver higher quality results faster than ever.
Further, the days of planning your week around your favorite TV shows or wrestling to program a VCR are in the past. There is a growing trend toward delivering still and animated content on demand as opposed to on a fixed schedule. YouTube is a great example; you can watch any video at any time you want, and that content will appear on your chosen device in a format and resolution appropriate to that device. The process of delivering content from a server to an end device on demand is called streaming. Streaming also places a tremendous burden on computer processors that must fetch the desired content, translate that content into a format usable by the end device, and then transmit that content to the end device in real time without interruption.
Delivering “EPYC” Content
How do 4th Gen AMD EPYC processors meet this demand from the M & E sector? This section showcases some truly “epyc” results.
- Autodesk® Arnold: Autodesk Arnold is an advanced Monte Carlo ray tracing renderer designed for VFX and animation production. It is designed to work with some of the top tools used by digital artists such as Maya, Houdini, 3ds Max, Cinema 4D, and Katana via plugins. AMD tested the 4th Gen AMD EPYC 9654 processor against both the 3rd Gen AMD EPYC 7763 and Intel® Xeon® Platinum 8380 processors. Each of these processors represents the “top of stack” for that processor family or generation. The 3rd Gen AMD EPYC 7763 showed competitive uplifts of ~1.45x and ~1.56x respectively over the Intel Xeon Platinum 8380 rendering the gtcrobot and sima scenes. These are both impressive results, but the 4th Gen AMD EPYC 9654 processor delivers whopping ~2.74x and ~2.42 respective uplifts rendering the same scenes compared to the Intel Xeon Platinum 8380. Dividing the 4th Gen AMD EPYC uplifts by the 3rd Gen AMD EPYC uplifts yields approximate generational uplifts of ~1.88x and ~1.55x, respectively.[2]
Figure 1: Comparing 4th Gen vs. 3rd Gen AMD EPYC and Intel® Xeon® processors running Autodesk Arnold
- Chaos® V-Ray®: Chaos V-Ray5 is a 3D rendering plugin that works seamlessly with major 3D design and CAD programs, such as 3ds Max, Cinema 4D, Houdini, Maya, Nuke, Revit, Rhino, SketchUp, and Unreal. V-ray allows artists and designers to create and share projects with real-time ray tracing and the ability to render high-quality 3D visualizations. It is widely used for film and television productions, advertising, and architectural visualizations. The 4th Gen AMD EPYC 9654 processor delivers an overwhelming, world record ~1.98x generational performance uplift over the 3rd Gen AMD EPYC 7763 processor. [3]
Figure 2: Comparing 4th Gen vs. 3rd Gen AMD EPYC processors running Chaos V-Ray, in vsamples
- FFmpeg: FFmpeg is a multimedia framework that can encode, decode, transcode, stream, filter and play just about any type of video in virtually any format, from legacy to ultramodern. FFmpeg works on a wide variety of operating systems, environments, and configurations. 3rd Gen AMD EPYC 7763 processors show a strong ~1.25x uplift compared to Intel Xeon 8380 Platinum processors, and 4th Gen AMD EPYC 9654 processors show an even more impressive ~1.35x uplift. Dividing the 4th Gen AMD EPYC uplift by the 3rd Gen AMD EPYC uplift gives us a generational uplift of ~1.08x.[4]
Figure 3: Comparing 4th Gen vs. 3rd Gen AMD EPYC and Intel® Xeon® processors running FFmpeg
- Synamedia® Virtual Digital Content Manager (vDCM): Synamedia vDCM provides advanced virtualized, software-based video, audio, and metadata processing for live delivery across many video formats. Broadcasters, content providers, and service providers can offer excellent viewing experiences that include high picture quality and multi-screen transcoding delivered with high bandwidth efficiency. Synamedia testing showed that the 4th Gen AMD EPYC 9654 processor delivers an average of approximately 74% generational video encoding performance uplifts over the 3rd Gen AMD EPYC 7763 processor using only 50% more processor cores across a wide range of video bitrates, resolutions, framerates, and formats. A single dual-processor system powered by 4th Gen AMD EPYC 9654 processor) can concurrently transcode two 8K video streams at 60 frames per second.[5]
Figure 4: Synamedia vDCM generational performance uplifts
- Visionular AV1 Codec: The Visionular AV1 codec is an advanced video coding format that offers high performance and fidelity in rendering. In Visionular testing, the 4th Generation AMD EPYC 9654 processor encodes 8 concurrent video streams at ~1.66x higher frame rate of the 3rd Gen AMD EPYC 7763 processor while encoding the crowd_run scene. For the Tears_of_Steel scene, the 4th Gen AMD EPYC 9654 processor encodes 8 concurrent streams at a ~1.55x higher frame rate of the 3rd Gen AMD EPYC 7763 processor. These results indicate a generational performance uplift of about double with only 50% more cores.[6]
Figure 5: Visionular generational performance uplifts
Conclusion
AMD EPYC processors have already made powerful inroads into the Media and Entertainment industry and how 4th Gen AMD EPYC processors continue to deliver performance gains in this vertical. 4th Gen AMD EPYC processors are ideal for media and entertainment workloads from rendering to encoding, decoding, and transcoding still and animated content. Our impressive performance allows you to render higher-resolution content and deliver streaming content more rapidly. You may also be able to process your current workloads using fewer servers, thanks to the sheer core and thread density of these processors. Either way, 4th Gen AMD EPYC processors can help you deliver today’s and tomorrow’s digital media more quicklyand/or with higher quality than ever before. Now that’s entertainment!
Raghu Nambiar is a Corporate Vice President of Data Center Ecosystems and Solutions for AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
References
- https://en.wikipedia.org/wiki/Computer-generated_imagery
- Please see https://www.amd.com/system/files/documents/epyc-9004-pb-arnold.pdf.
- SP5-038B: V-Ray based on published scores from https://benchmark.chaos.com/v5/vray as of 1/10/2023. Comparison of 2P AMD EPYC 9654 (209,102 max vsamples, https://benchmark.chaos.com/v5/vray/38531) is 2.24x the performance of published 2P Intel Xeon Platinum 8490H (93,210 max vsamples, https://benchmark.chaos.com/v5/vray/29746). 2P EPYC 7763 (109,248 vsamples, https://benchmark.chaos.com/v5/vray/12888) shown at 1.17x and 2P Intel Xeon Platinum 8380 (62,919 vsamples, https://benchmark.chaos.com/v5/vray/29746) shown at 0.68x for reference. Chaos®, V-Ray® and Phoenix FD® are registered trademarks of Chaos Software EOOD in Bulgaria and/or other countries. NOTE: Red text only needs to be included with charts that show the 7763/8380.
- Please see https://www.amd.com/system/files/documents/epyc-9004-pb-ffmpeg.pdf.
- See https://www.synamedia.com/whitepapers-reports/scaling-video-encoding-performance/ (registration required). Results not verified by AMD.
- SP5-132: Testing performed by Visionular on Aurora1 AV1 version 1.3.10, showing performance running 8 concurrent encoding streams using the crowd_run scene (https://media.xiph.org/video/derf/y4m/crowd_run_2160p50.y4m) and the Tears of Steel scene (https://visionular-release.s3.ap-northeast-1.amazonaws.com/y4m/tearsofsteel-2160p-40s.y4m), with each system using 128 threads. All performance data was provided by Visionular. Results are specific to Visionular and may not be typical. Systems configurations: CPU: Two AMD EPYC 9654 CPUs, 2.4 GHz | 3.7 GHz, 96 cores per socket (192 cores per node), TDP 400 W, 384 MiB L3 per CPU; RAM: 1.5TiB (24) Dual-Rank DDR5-4800 64GiB DIMMs, 1DPC; NIC: Broadcom Inc. NetXtreme BCM5720 Gigabit Ethernet (tg3: 5.15.0-48-generic); Storage: Samsung SSD 970 EVO Plus 500GB | Four 2.9TB NVMe drives; BIOS: AMI Core Ver. 5.25, Project Ver. RTI1000F and Default BIOS setting (SMT=on, X2APIC=on, IOMMU=Auto, APBDIS=Auto, Fixed SOC P-state=Auto, Determinism=power, NPS=1, DF C-states=Auto); OS: Ubuntu 20.04.4, OS Settings Default versus: CPU: Two AMD EPYC 7763 CPUs, 2.45 GHz | 3.5 GHz, 64 cores per socket (128 cores per node), TDP 280 W, 256 MiB L3 per CPU; RAM: 2TiB (32) Dual-Rank DDR4-3200 64GiB DIMMs, 1DPC; NIC: Broadcom Inc. NetXtreme BCM5720 Gigabit Ethernet (tg3: 5.15.0-48-generic); Storage: Samsung SSD 970 EVO Plus 500GB | Four 2.9TB NVMe drives; BIOS: AMI Core Ver. 4, Project Ver. RYM 1008B and Default BIOS setting (SMT=on, X2APIC=on, IOMMU=Auto, APBDIS=Auto, Fixed SOC P-state=Auto, Determinism=power, NPS=1, DF C-states=Auto); OS: Ubuntu 20.04.4, OS settings Default.