Tegra

  • Processadores
  • Dispositivos
  • Setor Automotivo
Documentos técnicos
header-image.jpg

GPU-Based Deep Learning Inference - A Performance and Power Analysis

NVIDIA GPUs already provide the platform of choice for Deep Learning Training today. This whitepaper investigates Deep Learning Inference on a Geforce Titan X and Tegra TX1 SoC. The results show that GPUs provide state-of-the art inference performance and energy efficiency, making them the platform of choice for anyone wanting to deploy a trained neural network in the field. View PDF 

NVIDIA Tegra X1 - NVIDIA'S New Mobile Superchip

Tegra X1 is NVIDIA's latest superchip, and has raised the bar again for mobile visual computing and energy-efficiency. Tegra X1 with its 256-core Maxwell GPU delivers 2x the performance and 2x the energy efficiency of Tegra K1. Tegra X1 supports the latest graphics APIs and further blurs the lines separating mobile, console and PC gaming. This paper explains the architectural features and capabilities of Tegra X1 and focuses in detail on mobile graphics and automotive technologies made possible by Tegra X1. View PDF 

NVIDIA TEGRA K1 – A NEW ERA IN MOBILE COMPUTING

NVIDIA's latest and most advanced mobile processor Tegra™ K1, creates a major discontinuity in the state of mobile graphics by bringing the powerful NVIDIA Kepler GPU architecture to mobile, and delivering tremendous visual computing capabilities and breakthrough power efficiency. With two decades of GPU industry leadership, the NVIDIA Tegra K1 mobile processor delivers the performance demanded by next-generation PC- and console-class mobile games, modern user interfaces, advanced visual computing applications, and high-resolution 4K displays, while delivering exceptional power efficiency required to bring the above capabilities to mobile devices. This whitepaper describes the architecture and benefits of Tegra K1 and Kepler GPU. View PDF 

NVIDIA Software Defined Radio Modem Technology – Enabling Next Generation LTE

Conventional wisdom, together with the massive amount of legacy investment by traditional modem vendors, dictates a fixed function approach to modem design. However the complexity of multi-mode, high-throughput cellular air-interfaces requires a revolution to lift us from convention. This revolution is brought about by the NVIDIA® Software Defined modem technology, the latest implementation of which is in the NVIDIA i500 and Tegra® 4i, discussed in this whitepaper. View PDF 

NVIDIA Tegra 4 Family GPU Architecture

The need for higher quality GPU subsystems in mobile devices and automotive infotainment systems will continue to grow as usage models rely increasingly on faster graphics processing capabilities. For the best user experiences, devices must provide highly responsive graphical user interfaces, fast Web browsing capability, visually rich 3D gaming, all while being able to drive higher resolution displays. The GPU subsystem within the NVIDIA® Tegra 4 family of processors is enabling all of the aforementioned visual computing capabilities in mobile devices, and is discussed in detail in this whitepaper. View PDF 

NVIDIA Tegra 4 Family CPU Architecture – Next Generation 4-PLUS-1™ Quad core

Smartphones and tablets are increasingly being used as personal computing devices. Today's mobile devices are no longer being used only for phone calls, messaging, and occasional Web browsing. Mobile applications for PC-class use cases such as photo editing, word processing, multi-tabbed Web browsing, modern graphics-rich gaming, and multi-tasking are now available and are pushing the performance requirements of mobile devices. The quad core CPUs and fifth Battery Saver Core in the NVIDIA® Tegra® 4 family of mobile SoCs discussed in this whitepaper, are architected to drive the next generation of mobile applications, and include several key enhancements that deliver higher performance along with great battery life.View PDF 

Chimera: The NVIDIA Computational Photography Architecture

The next wave in smartphone camera use will be fueled by computational photography. This will result in better images, capturing a scene based exactly on the subtleties and details your eyes see. The Chimera architecture harnesses the processing power built into the Tegra GPU, CPU, and ISP, enabling new computational photography features such as Always-on HDR for truly stunning photos, and advanced object tracking capability. View PDF 

NVIDIA Miracast Wireless Display Architecture

The group that brought Wi-Fi to the world will soon introduce their Miracast wireless display standard – enabling mobile devices to stream video and audio directly to large HDTVs without the need for cables or an existing wireless network. NVDIA will support this standard by bringing Tegra's outstanding multimedia capabilities to wireless display. This white paper looks at how NVIDIA's expertise in graphics and video processing can deliver an enriched wireless display experience. View PDF 

Arquitetura NVIDIA® DirectTouch™

A NVIDIA DirectTouch é uma arquitetura de toque com patente pendente que aprimora a capacidade de resposta a toques, transferindo parte do processamento de toques, normalmente realizado pelos controladores e módulos de toque, para o processador de aplicativos NVIDIA® Tegra® 3. A arquitetura também simplifica a implementação de hardware e interfaces de usuário baseados em toque, exigindo menos energia e, ao mesmo tempo, fornecendo desempenho mais expansível. Ver PDF 

Os benefícios das CPUS quad-core em dispositivos móveis

Esse documento técnico aborda como as CPU quad-core e a tecnologia de multiprocessamento simétrico (SMP) variável possibilitarão que os dispositivos móveis aumentem ainda mais a performance. Descubra como isso capacita os desenvolvedores de aplicativos e jogos a fornecer novas experiências móveis, ao mesmo tempo que aumenta a vida útil da bateria para os casos de uso mais populares. Ver PDF 

SMP variável – uma arquitetura de CPU de vários núcleos para um baixo consumo e uma alta performance

À medida que as exigências por performance dos aplicativos móveis aumentam, os fornecedores de SoC cada vez mais adotam arquiteturas de processadores com vários núcleos. Isso possibilita que eles ofereçam mais performance enquanto mantêm o consumo de energia dentro dos orçamentos móveis. Esse documento técnico fala sobre a tecnologia de multiprocessamento simétrico variável (vSMP) do Project Kal-El, que fornece novos níveis de performance quad-core, além de economias de energia que diminuem o consumo energético durante os estados de standby ativos. Ver PDF 

Disponibilizando gráficos avançados em dispositivos portáteis

O tamanho e a resolução de tela dos dispositivos móveis aumentaram rapidamente nos últimos anos, e a crescente popularidade dos tablets fará com que sua resolução e tamanho de tela se aproximem cada vez mais daqueles encontrados nos laptops. Esse documento técnico examina como os dispositivos móveis atuais e futuros precisarão de GPUs que não somente lidem com cargas maiores de processamento de pixels, como também permaneçam dentro dos orçamentos de energia móvel. Ver PDF 

Os benefícios de vários núcleos de CPU nos dispositivos móveis

Os fabricantes de CPUs de desktops fizeram a transição para as arquiteturas de processadores com vários núcleos há cinco anos, a fim de lidar com as crescentes demandas de performance e com o crescimento exponencial do consumo de energia dos processadores com um único núcleo. As CPUs atuais utilizam vários núcleos para concluir mais trabalhos de maneira mais rápida e consumindo menos energia que seus antecessores com um único núcleo. E os processadores móveis estão enfrentando os mesmos desafios de performance e consumo de energia. Esse documento técnico descreve como os dispositivos móveis realizarão a transição para CPUs com vários núcleos, a fim de aumentar ainda mais a performance e ampliar a vida útil da bateria. Ver PDF