Tensor Core GPU 架構(gòu)旨在將 AI 帶入每個行業(yè)。
人工智能的核心
人工智能不是由任何一個行業(yè)定義的。它存在于超級計算、醫(yī)療保健、金融服務(wù)、大數(shù)據(jù)分析和游戲領(lǐng)域。它是每個行業(yè)和市場的未來，因為每個企業(yè)都需要智能，而人工智能的引擎是英偉達(dá) GPU 計算平臺。
NVIDIA Volta 是人工智能背后的新驅(qū)動力。Volta 將推動每個行業(yè)的突破。人類的登月計劃，如根除癌癥、智能客戶體驗和自動駕駛汽車，都在人工智能的下一個時代觸手可及。
看視頻
突破性的創(chuàng)新
640 個張量核心
性能的指數(shù)級飛躍
每個行業(yè)都需要人工智能，隨著速度的巨大飛躍，人工智能現(xiàn)在可以應(yīng)用于每個行業(yè)。配備640
Tensor Cores，Volta 提供超過每秒 125 teraFLOPs (TFLOPS) 的深度學(xué)習(xí)性能，與上一代 NVIDIA Pascal? 架構(gòu)相比提高了 5 倍以上。
新的 GPU 架構(gòu)
專為現(xiàn)代計算機(jī)設(shè)計
人類面臨的最大挑戰(zhàn)將需要用于計算和數(shù)據(jù)科學(xué)的最強(qiáng)大的計算引擎。Volta 擁有超過 210 億個晶體管，是世界上最強(qiáng)大的 GPU 架構(gòu)。它將 NVIDIA ? CUDA ? 和 Tensor Core配對，以在 GPU 中提供 AI 超級計算機(jī)的性能。
下一代 NVLINK
快速解決方案的可擴(kuò)展性
Volta 使用下一代革命性的 NVIDIA NVLink ?高速互連技術(shù)。與上一代 NVLink 相比，這提供了 2 倍的吞吐量。這支持更高級的模型和數(shù)據(jù)并行方法，以實現(xiàn)強(qiáng)大的擴(kuò)展，以實現(xiàn)絕對最高的應(yīng)用程序性能。
VOLTA 優(yōu)化軟件
GPU 加速的框架和應(yīng)用程序
數(shù)據(jù)科學(xué)家經(jīng)常被迫在模型準(zhǔn)確性和更長的運(yùn)行時間之間進(jìn)行權(quán)衡。借助 Volta 優(yōu)化的 CUDA 和
cuDNN、NCCL 和
TensorRT等
NVIDIA 深度學(xué)習(xí) SDK 庫，業(yè)界頂級框架和應(yīng)用程序可以輕松利用 Volta 運(yùn)行
混合精度的強(qiáng)大功能。這推動數(shù)據(jù)科學(xué)家和研究人員比以前更快地發(fā)現(xiàn)。 NVIDIA Volta
The Tensor Core GPU Architecture designed to Bring AI to Every Industry.
THE CORE OF AI
AI is not defined by any one industry. It exists in fields of supercomputing, healthcare, financial services, big data analytics, and gaming. It is the future of every industry and market because every enterprise needs intelligence, and the engine of AI is the NVIDIA GPU computing platform.
NVIDIA Volta is the new driving force behind artificial intelligence. Volta will fuel breakthroughs in every industry. Humanity’s moonshots like eradicating cancer, intelligent customer experiences, and self-driving vehicles are within reach of this next era of AI.
WATCH VIDEO
GROUNDBREAKING INNOVATIONS
640 TENSOR CORES
An Exponential Leap in Performance
Every industry needs AI, and with this massive leap forward in speed, AI can now be applied to every industry. Equipped with 640
Tensor Cores, Volta delivers over 125 teraFLOPs per second (TFLOPS) of deep learning performance, over a 5X increase compared to prior generation NVIDIA Pascal? architecture.
NEW GPU ARCHITECTURE
Engineered for the Modern Computer
Humanity’s greatest challenges will require the most powerful computing engine for both computational and data science. With over 21 billion transistors, Volta is the most powerful GPU architecture the world has ever seen. It pairs NVIDIA? CUDA? and Tensor Cores to deliver the performance of an AI supercomputer in a GPU.
NEXT GENERATION NVLINK
Scalability for Rapid Time-to-Solution
Volta uses next generation revolutionary NVIDIA NVLink? high-speed interconnect technology. This delivers 2X the throughput, compared to the previous generation of NVLink. This enables more advanced model and data parallel approaches for strong scaling to achieve the absolute highest application performance.
VOLTA-OPTIMIZED SOFTWARE
GPU-Accelerated Frameworks and Applications
Data scientists are often forced to make trade-offs between model accuracy and longer run-times. With Volta-optimized CUDA and
NVIDIA Deep Learning SDK librarieslike
cuDNN, NCCL, and
TensorRT, the industry’s top frameworks and applications can easily tap into the power of Volta running
mixed precision. This propels data scientists and researchers towards discoveries faster than before.
