Apresentação em tema: "NVIDIA MAXIMUS TECNOLOGIA MANUFATURA"— Transcrição da apresentação:
1 NVIDIA MAXIMUS TECNOLOGIA MANUFATURA DESIGN COM MAIS REALISMO,MAIOR IMPACTO EM SEUS PROJETOS
2 A visão: Design com realismo Cálculos físicos reais Interação em tempo realUtilização de texturas com características reais durante o design.Dinâmica de mecânicaDinâmica de fluídosSimulaçãoPropriedades físicasFoto-realismoIluminação Global
3 O Design no mundo virtual de hoje Análises físicas REAIS Design com interação+Aplicação de texturasDesign com interação
4 Etapas tradicionais no Desenvolvimento do Produto DesignDesignReviewSurfaceCreationComponentEngineeringAssemblyEngineeringDigital PrototypeDesign SignoffManufacture PlanningProductionRenderingComponent SimulationEngineering SimulationTest / Photo PrototypesPrototype ConstructionAplicações que exigem muitos cálculos computacionais migraram das Workstations para clusters de CPUEtapas de design com soluções e grupos separados demandam mais tempo e são caras para as empresas
5 Workstation Tradicional Design ou Simulação. Design ou Renderização. Nunca as duas.System CPUOpenGLGPU=
6 A solução! Desenho & Simulação Desenho Visualização Simulação Computação
7 A nova Workstation com a tecnologia NVIDIA MAXIMUS System CPUDesignGPUSimulateGPU
8 Reinventando os processos DesignSimulate (CPU)Design Time = xSimulation Time = yTraditionalWorkstationLets give you an example of where this technology applies in our core marketins.Simulation used to happen at the desktop and many of you remember that..it has migrated to servers..but now with maximus its time to return those cycles to the WS…In this real world example,where the engineer would be making the OR decision or even batching the sim out to a server, with the addition of Maximus..we can realize the genius of the AND…could have traditionally been a server..move some of the cycles back to the ws, where it used….Design Time = xDesign+SimulateMAXIMUSWorkstationSimulation Time = 1/10 y
9 Reinventando os processos DesignSimulate (CPU)Design Time = xSimulation Time = yTraditionalWorkstationLets give you an example of where this technology applies in our core marketins.Simulation used to happen at the desktop and many of you remember that..it has migrated to servers..but now with maximus its time to return those cycles to the WS…In this real world example,where the engineer would be making the OR decision or even batching the sim out to a server, with the addition of Maximus..we can realize the genius of the AND…could have traditionally been a server..move some of the cycles back to the ws, where it used….Design Time = xDesignDesign2Design3Design4++++SimulateSimulate2Simulate3Simulate4MAXIMUSWorkstationSimulation Time = 1/10 y
10 Alocação inteligente das GPUs Driver unificado Computação ParalelaNVIDIA® MAXIMUSComputação VisualAlocação inteligente das GPUsDriver unificadoCertificada para as apps profissionais
11 One NVIDIA® Quadro® 2000 GPU One NVIDIA® Tesla® C2075 GPU Dual CPUs Básica para CAECusto-benefícioMáximo DesempenhoSingle CPU32 GB RAM*One NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPUDual CPUs32 – 48 GB RAMOne NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPUDual CPUs48 GB RAM (or more)*One NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPU* Requerimento de memória baseado pelo tamanho dos projetos para minimizar disco de I/O.
12 Renderização máxima com NVIDIA MAXIMUS Interação: Imagens & FilmesEstilo / Desenho IndustrialStylingDesign ReviewClass-A SurfaceDesignEngineeringAssemblyEngineeringDigital PrototypeDesign SignoffManufacture PlanningProductionRenderingTest / Photo PrototypesPrototype Construction
13 Processos Típicos para Estilo/Design Passos do desenvolvimento do estilo com a renderização fotorealística Offline (Devagar)StylingDesign 1aRenderDesign 1bRenderDesign 1cRenderDesign ReviewsIn product styling, design reviews are required to determine needed changes or to sign off on the design.In this slide set, each example assumes that a design goes through 3 revisions.On this slide, we show that until the offline rendering process is complete, the design review has to wait. But the stylist or designer can’t really start working on the next version until he gets the feedback and changes from the design review.Review 1aReview 1bReview 1cClass-A Surface DesignSurfaceDesign reviews demora até que as imagens e animações renderizemMudanças frequentes são desafios frequentesO processo de revisão de Design é caro e leva muito tempo
14 NVIDIA MAXIMUS para Design/Estilo Passos do desenvolvimento do estilo com a renderização fotorealística em Tempo-RealStylingDesign 1aDesign 1bDesign 1cDesign ReviewsWith GPU accelerated or real-time ray tracing, You are working or designing with photorealistic materials, without the need to send the models off to be rendered by a separate system, or even group.This allows the design revision process to go much faster, allowing you to more quickly get to the point where you start engineering the product.Review 1aReview 1bReview 1cClass-A Surface DesignCriação de Class-A SurfaceOs processos são executados ao mesmo tempo!
15 Exemplos de Rendering : CATIA V6 LiveRendering Otimizada pelo tecnologia NVIDIA Maximus Visão realística durante o processo de DesignDireto do CATIA V6Habilitado pelo NVIDIA iRayCorreção física de renderização
16 Customer Impact – Mercedes Benz “We always start with really ambitious ideas. Last year, we were forced to compromise, eliminating a lot of great features and dumbing things down because our computer technology wasn’t fast enough to perform the design and rendering we needed. This year was an entirely different story. The NVIDIA Maximus-powered system gave us the time to explore different options creatively, which shows in the quality of this year’s entry. With the parallel processing capabilities enabled by the NVIDIA Maximus systems, we can now be 10 times more creative. In a given amount of time, we can explore so many different options and get to a better end product.”Alan Barrington, Designer Mercedes-Benz Advanced Design Center California
17 Customer Impact – Cleat Inc. “The previous workstation took about ten minutes to change the color of a wheel. In the NVIDIA Maximus environment, NVIDIA iray runs smoothly and at high speed, thus improving our workflow. It can complete rendering in about 30 seconds. The amount of speed increase is overwhelming."Mr. Yamashita, President of Cleat
23 Customer Impact – Liquid Robotics “If you wanted to do anything else while running a simulation or modeling, you were out of luck. You either got a cup of coffee or worked on something in the shop once the computer was using all its processing power running one of these programs.”“The real advantage of the Maximus technology is flexibility and increased productivity. It’s a tremendous tool to allow my engineers to be flexible, to multitask, and to be more productive because they’re not waiting on computational power, period.”Tim Ong, VP Mechanical Engineering for Liquid Robotics.
24 Customer Impact - Astrobotics “The NVIDIA Maximus-powered system is like getting three people’s worth of use on a single machine. This system is a beast. We haven’t yet found anything it can’t handle, even simultaneous CAD, analysis, and additional number crunching in remote rendering jobs. We can do all our highly complex analysis or rendering without interrupting other work going on locally, and we are now able to do some very useful and interesting investigations with Ansys that wouldn’t have been possible before.”Jason Calaiaro, Director of Information Systems for Astrobotic Technology
25 Typical Small & Medium Businesses In-house Design and Engineering. Outsourced Simulation.I-DesignDesign 1aRenderReview 1aDesign 1bRenderReview 1bDesign 1cRenderReview 1cDesign ReviewsComponent SimulationDelaySimulate 1aDelaySimulate 1bDelaySimulate 1cOutsourced prototypeOutsourcedDesign reviews wait for image and animation renderingSimulation has to be outsourced to keep workstations available
26 Typical Small & Medium Businesses Local Simulation Removes Delay. Adds Confidence for No Prototype!I-DesignDesign ReviewsDesign 1aReview 1aDesign 1bReview 1bDesign 1cReview 1cComponent SimulationSimulate 1aSimulate 1bSimulate 1cManufacturingStraight to ProductionEverything done in-house without losing productivity. Often no prototypes required!
27 * Memory requirements dictated by problem size to minimize disk I/O. Entry to CAE on GPUPrice-PerformanceMax PerformanceSingle CPU32 GB RAM*One NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPUDual CPUs32 – 48 GB RAMOne NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPUDual CPUs48 GB RAM (or more)*One NVIDIA® Quadro® GPUOne NVIDIA® Tesla® C2075 GPU* Memory requirements dictated by problem size to minimize disk I/O.
28 A New Era in Workstation Computing NVIDIA® Maximus™A New Era in Workstation ComputingGPU Horsepower10XCPU HorsepowerEfficient Multi-TaskingHigher PerformanceWidely SupportedTesla & Quadro boardcertificationsGlobal OEM systemAll leading CAD & CAE ISVsDesign AND SimulateDesign AND RenderSeamless ExperienceUp to 4x faster simulationUp to 9x faster rendering100% interactive design performance while rendering or simulatingImmediate Availability from Dell, HP, Lenovo, FTS - Nov 14th
30 Superior Performance NVIDIA® Quadro® 5000 GPU Quadro FX 4800 Here is a chart showing relative performance between generation of Quadro card from 3 years ago FX 1700 / 3700 and 4600 and the latest generation of Fermi cards.These results were obtained by running industry accepted standard graphics benchmark tests (SPECViewperf 11) for SolidWorks. Performance on actual models will vary depending on the model.As shown in the chart, you will realize substantial performance gains by upgrading from previous generations of Quadro solutions. Again, performance gains are more than 3x between an old FX1700 and a new Quadro Similar performance gains are available on the 4000 and the 5000.If you or your staff are spending a substantial amount of time using Solidworks there is a clear benefit to upgrade your graphic subsystem.Quadro FX 37003.3xQuadro 2000Quadro FX 1800Quadro FX 17003.2xResults obtained by running industry accepted standard graphics benchmark tests (SPECViewperf® 11) for SolidWorks, comparing Quadro 4000 or 2000 to Quadro FX 1700. The comparison is based on standard industry workstations representative of when each particular graphics product was launched. n: Resolution of 1900 x 1200, Windows Win 7, 64 Bit, System 1P XEON 3.3GHz, 6GB Memory. n-2: Resolution of 1900 x 1200, WinXP x64, 1x 3GHz Core2 Quad (QX9650) 4GB.
31 ANSYS and SIMULIA/Abaqus More than Twice the Design Candidates. Up to 4.2x Faster. with NVIDIA Tesla GPU8 CPU Cores + Tesla C20752 CPU Cores8 CPU CoresUp to 4.2x Fasterwith NVIDIA GPUSolidworks RealView 360 is available on premium version of Solidwork. RealView is a great way to obtain photorealistic rendering of Soldiworks models. RealView 360 is using a technique called ray tracing and uses the CPU to do the computation.For users that do not have access to Premium versions of Solidworks or want to free CPU cycles on their workstations GPU accelerated ray tracing packages like Bunkspeed can be an attractive solution.Bunkspeed is an easy to use interactive ray-tracing application that harnesses the processing power of the NVIDIA GPU. SolidWorks assemblies can be opened directly within Bunkspeed “Move” providing a simple design workflow. Bunkspeed is a relatively inexpensive add on ($495) to solidworks. With an easy-to-use intuitive interface, Bunkspeed quickly and accurately simulates the physical world. By easily integrating with SolidWorks Standard or Premium, you get a seamless workflow from design to photorealistic imagery without any data loss.Adding on Bunkspeed Move can be a cost effective way to add high quality photo-realistic rendering while reducing CPU cycles at the system level.Quadro professional solutions accelerate Bunkspeed. Ray tracing performance can be up to 3 time faster when using NVIDIA GPU instead of CPU in Bunkspeed. Creating and sharing photorealistic models has never been easier or faster.4 CPU Cores + Tesla C20752 CPU Cores4 CPU CoresBenchmark: V13sp-5 Model- Turbine Geometry 2.1M DOF, Static Nonlinear, Direct SparseCPU: 2 x Westmere Xeon 5670 at 2.93 GHzBenchmark: s4b Model- Engine Geometry, 5M DOF Static nonlinear, Direct SparseCPU: Westmere Xeon 5680 at 3.33 GHz