아놀드 GPU 렌더 - anoldeu GPU lendeo

아놀드 GPU를 탑재한 아놀드 6는 룩(look) 개발부터 최종 프레임 렌더링에 이르기까지, 아티스트 및 스튜디오가 콘텐츠를 생성하고 반복하는 작업을 단순화한다. 이를 통해, 빠른 속도로 진행되는 작업환경에 적응할 수 있는 속도와 유연성을 제공할 뿐 아니라, 렌더링 용량을 확장시켜 제작흐름의 변화를 수용할 수 있도록 한다. 또한, 아놀드의 직관적인 인터페이스를 통해 CPU와 GPU 렌더링 간에 원활한 전환이 가능하여, 클릭 한 번으로 동일한 설정을 유지할 수 있다. 아놀드 GPU는 엔비디아(NVIDIA)의 옵틱스(OptiX) 프레임워크를 기반으로 하며 엔비디아 RTX 기술을 활용하도록 최적화되었다.

크리스 비아노(Chris Vienneau), 오토데스크 미디어 및 엔터테인먼트 제품 담당 수석 총괄은 “콘텐츠 공급업체들이 콘텐츠 품질에 대해 높아지고 있는 고객의 기대치를 충족시키려 함에 따라, 렌더링 워크로드가 계속해서 증가하고 있다. 창조적인 작업 프로세스에서 속도와 상호작용은 그 어느때보다 중요해졌다. 아놀드 6는 향상된 성능을 제공하여 CPU 렌더러에서 알려진 것과 동일한 고품질 렌더링 결과로 로드를 가볍게 한다. 우리는 아놀드 GPU가 최신 RTX GPU와 RTX 서버에서 구동되도록 최적화하기 위해 엔비디아와 협력했다. 이 최신 업데이트를 신규 및 기존 아놀드 고객에게 제공할 수 있게 되어 매우 기쁘다”고 설명했다. 

리차드 케리스(Richard Kerris), 엔비디아 미디어 및 엔터테인먼트 담당 총괄은 “엔비디아 RTX GPU를 기반으로 하는 아놀드 GPU는 그래픽 전문가들이 작업속도를 높이고, 마감시간을 단축하는 동시에 새로운 아이디어를 탐구하는데 상당한 도움을 줄 것이다. 오토데스크는 아놀드 렌더러를 위해 RTX를 채택한 최초의 기업이었다. 아티스트들이 아놀드 6의 GPU가 구현해내는 새로운 차원의 상호작용성, 유연성, 그리고 획기적인 속도를 하루빨리 경험해 보길 기대한다”고 말했다. 

지난 3월 베타 버전으로 처음 공개된 아놀드 6는 조명, 셰이더 및 카메라 전반의 개선, 온디맨드 형식의 텍스처 로딩, 그리고 OSL(Open Shading Language) 및 OpenVDB 볼륨 지원으로 아놀드 GPU 툴세트를 완성했다. 또한, 개방형 표준 채택을 가속화하기 위한 오토데스크의 지속적인 노력의 일환으로, 현재 USD 생태계내에서 아놀드의 구성요소 모음을 온라인에서도 이용할 수 있다. 

아놀드 6의 주요 특징은 다음과 같다.

• 통합된 렌더러: CPU 렌더링과 GPU 렌더링 간의 원활한 전환이 가능하다.

• 조명, 카메라 등을 포함한 추가 개선사항: 아놀드 GPU는 이제 OSL, OpenVDB 볼륨, 온디맨드형 텍스처 로딩, 대부분의 LPE, 조명, 셰이더 및 모든 카메라를 지원한다. GPU 노이즈도 적응형 샘플링을 사용할 때 CPU 노이즈와 동등한 수준으로 개선되어, 사용된 렌더러에 관계없이 훨씬 빠르고 예측 가능한 결과를 산출할 수 있다. 

• 엔비디아 RTX 하드웨어에 최적화: 최신 기술을 통해 필요에 따라 렌더링 성능을 확장할 수 있다.

• 새로운 USD 구성요소: 이제 GitHub에서 아놀드 노드와 속성에 대한 히드라(Hydra) 렌더 델리게이트, 순차적 방식의 아놀드 USD, USD 스키마를 이용할 수 있다. 

• 성능 향상: 더 빠른 하위 분할, 개선된 피지컬 스카이(Physical Sky) 셰이더, 유전체 미세면 다중 산란을 포함한 여러 가지 개선사항으로 효율성과 성능을 극대화한다.

아놀드 GPU가 탑재된 아놀드 6는 오토데스크 미디어 및 엔터테인먼트 콜렉션(Autodesk Media & Entertainment Collection)에서 독립형 구독 또는 엔드-투-엔드 크리에이티브 툴 형태로 이용 가능하다. 아놀드의 월간, 연간, 3년 단위의 구독은 오토데스크 e-스토어에서 이용할 수 있다. 아놀드 6에 대한 무료 30일 체험판도 사용 가능하다. 아놀드 GPU는 Autodesk Maya, Autodesk 3ds Max, Houdini, Cinema 4D 및 Katana에서 지원되는 모든 플러그인에서 사용할 수 있다. 

Supported Features and Known Limitations

arnoldrenderer.com

• Arnold GPU supports complex shading networks, SSS, hair, atmospherics, instancing, and procedurals.
• With the same settings, GPU renders will currently be noisier than CPU renders since GPU renders are “non-splitting” (i.e. one path per camera/AA sample). Accordingly, to achieve equivalent noise the AA sample count will need to be increased in GPU renders. Adaptive rendering is fully supported.
• standard_surface, standard_hair, and standard_volume are supported, with some limitations (see the table below).
• OSL is supported, with some limitations currently (see below).
• OpenVDB volumes are supported, with some limitations currently (see below).
• Volume displacement is supported, with some limitations currently (see below).
• Filename attribute tags are supported. Mipmap bias is not supported.
• Light linking is not supported on volumes.
• Limited AOV support.
• Trace sets are not supported.
• Custom procedurals, drivers, color managers are supported.
• Custom shaders, cameras, filters, BSDFs are not supported.
• Noice is not supported with Arnold GPU renders (because the variance filter is not supported on Arnold GPU)

System requirement - Arnold GPU

From version 6 onwards you are able to switch between CPU and GPU rendering. Arnold is optimised for NVIDIA RTX series graphics cards with CUDA cores. Take advantage of the new technology to improve your rendering performance. We can recommend the following graphics cards for optimal rendering results: NVIDIA Quadro RTX 4000 with 8GB DDR6 VRAM, NVIDIA Quadro RTX5000 with Cuda cores, NVIDIA Quadro RTX6000 with 24GB VRAM. NVIDIA Geforce series graphics cards, the RTX3060, RTX3070, RTX3080 and RTX3090 models demonstrate their power in Arnold 6 in particular by combining RTX-accelerated rendering and GPU-accelerated AI denoising for fast, interactive rendering. The result is noise-free images with very high fidelity. RTX graphics cards are almost twice as fast compared to NVIDIA’s Pascal series graphics cards.

Arnold GPU works on NVIDIA GPUs of the Ampere, Turing, Volta, Pascal, and Maxwell architectures. Multiple GPUs will improve performance, and NVLink can be used to connect multiple GPUs of the same architecture to share memory.

In general, Arnold is going to work on pretty much any 64-bit system where Houdini, Maya, Cinema 4D, 3ds Max, or Katana works. However, there are some minimum requirements:

• Windows 10 or later, with the Visual Studio 2019 redistributable.
• Linux with at least glibc 2.17 and libstdc++ 4.8.5 (gcc 4.8.5). This is equivalent to RHEL/CentOS 7.
• macOS 10.13 or later.
• CPUs need to support the SSE4.1 instruction set. Apple Mac models with M series chips are supported under Rosetta 2 mode.
• GPU rendering works on Windows and Linux only and requires an NVIDIA GPU of the Ampere, Turing, Volta, Pascal, or Maxwell architecture.
• Optix™ denoiser requires an NVidia GPU with CUDA™ Compute Capability 5.0 and above.

General setting

Pre-populating the GPU cache

The very first time you render with the GPU, the GPU renderer has to create a cache of shaders. This can delay the time to first pixel for your first render.

To avoid the one-time delay, we recommend that you run Pre-Populate GPU Cache before you do any renders. Note that pre-populating the cache can take up to 15 minutes.

The cache only needs to be re-populated after installing a new Arnold version, updating to a new NVIDIA driver, or changing the hardware configuration of GPUs on the system.

Selecting a Render Device

You can easily switch between CPU and GPU with a single click in the Render Settings > System.

Matching Noise on CPU and GPU

Matching noise can take a little experimentation because Arnold GPU uses Camera (AA) sampling only. You should  also use Adaptive sampling. Here are some guidelines:

• Set the Max. Camera (AA) in the range of 30 to 50 (depending on the scene, you might go closer to 100). In general, the max samples should be a large value. A large max samples means that the quality is controlled by the noise falling under the threshold, instead of by clamping to the max AA.
• Set the Adaptive Threshold to something like 0.015 or 0.02. For a noise-free render, lower the threshold value, maybe even as far as 0.010.
• Set the Camera (AA) samples to around 3 or 4. One of the few reasons to go higher with AA is for motion blur. The higher the number of Camera (AA) samples, the less of a speedup you’ll get from adaptive sampling.

Get your own remote workstation with iRender service!

With the mentioned configuration requirements, you can consider investing in yourself a workstation to work effectively with Arnold projects. Or another economical and effective option is to use iRender’s cloud rendering service. You will immediately have yourself a workstation with a powerful configuration from 1 to 8 GPU RTX3090, your working environment will be saved and ready to use whenever you need it.

To simplify, we provide the servers and you will access the server you want via Remote Desktop. When you access, it is just like a new and blank computer so you have to install the software and plugins you need just one time and they will be saved and ready to use anytime.

iRender high-end hardware configuration

• Offering the most powerful graphic card currently, RTX 3090. The servers range from Single and Multi-GPU servers: 1/2/4/6/8 x RTX 3090 with 24 GB vRAM capacity, fitting to the heaviest images and scenes. NVLink/SLI requested for bigger vRAM.
• A RAM capacity of 128/256 GB.
• Storage (NVMe SSD): 512GB/1TB.
• Intel Xeon W-2245 or AMD Ryzen Threadripper Pro 3955WX CPU with a high clock speed of 3.90GHz.
• Additionally, iRender provides NVLink (Request) which will help you increase the amount of VRAM to 48GB. This is a technology co-developed by Nvidia and IBM with the aim of expanding the data bandwidth between the GPU and CPU 5 to 12 times faster than the PCI Express interface. These servers are sure to satisfy Arnold artists/ studios with very complex and large scenes.

Below are our Multi-GPU server package with one of the best GPU and CPU in the market.

Hopefully the above information will help you better control the render quality and speed and have a better experience when working with Arnold.

At iRender, we have developed multiple RTX 3090 workstations specifically optimized and configured for GPU rendering workflows, from single- GPU 1x RTX 3090 to Multi-GPU 2/4/6/8 x RTX 3090, you can choose one server that is suitable for your demand and your software to get started. You are guaranteed to experience maximum performance for your budget. Users will remotely connect to our server, install their software only one time and easily do any intensive tasks like using their local computers. We offer clients the powerful processor Intel Xeon W-2245 @ 3.90GHz / AMD Ryzen Threadripper Pro 3955WX @ 3.90GHz and up to 1 TB SSD to load your scene quickly.

Register an account today for FREE TESTING COUPON to experience our service. Or contact us via WhatsApp: (+84) 912 785 500 for advice and support.

Thank you & Happy Rendering!

Reference source: arnoldautodesk.com

Related Posts

The latest creative news from Maya Cloud Rendering, Redshift Cloud Rendering, Cinema 4D Cloud Rendering , 3D VFX Plugins & Cloud Rendering.

05 Oct 2022

6 types of light in V-Ray for Maya

While you can use Maya standard lights with V-Ray, V-Ray includes a set of lights designed specifically for rendering with…

15 Sep 2022

Using your GPU in V-Ray for Maya

Chaos releases visualization tools that enable creators in various creative fields to produce photorealistic visuals and animation. The company’s physically…

09 Sep 2022

Common errors and solutions for Maya users with iRender

Maya is a professional 3D animation, modeling, simulation, and rendering toolset, designed for creating realistic characters and blockbuster-worthy effects. In…