GPU-Accelerated Frameworks and Applications Volta adds support for new synchronization patterns. Pascal and Volta include support for new Cooperative Launch APIs that support synchronization amongst CUDA thread blocks. Basic Cooperative Groups functionality is supported on all NVIDIA GPUs since Kepler. Cooperative Groups allows developers to express the granularity at which threads are communicating, helping them to express richer, more efficient parallel decompositions. Cooperative Groups and New Cooperative Launch APIs Cooperative Groups is a new programming model introduced in CUDA 9 for organizing groups of communicating threads.On IBM Power platforms, new Address Translation Services (ATS) support allows the GPU to access the CPU’s page tables directly. Enhanced Unified Memory and Address Translation Services GV100 Unified Memory technology in Volta GV100 includes new access counters to allow more accurate migration of memory pages to the processor that accesses the pages most frequently, improving efficiency for accessing memory ranges shared between processors.The new NVIDIA DGX-1 with V100 AI supercomputer uses NVLink to deliver greater scalability for ultra-fast deep learning training. NVLink now supports CPU mastering and cache coherence capabilities with IBM Power 9 CPU-based servers. GV100 supports up to 6 NVLink links at 25 GB/s for a total of 300 GB/s. Second-Generation NVLink™ The second generation of NVIDIA’s NVLink high-speed interconnect delivers higher bandwidth, more links, and improved scalability for multi-GPU and multi-GPU/CPU system configurations.
Volta MPS also triples the maximum number of MPS clients from 16 on Pascal to 48 on Volta.
Volta Multi-Process Service Volta Multi-Process Service (MPS) is a new feature of the Volta GV100 architecture providing hardware acceleration of critical components of the CUDA MPS server, enabling improved performance, isolation, and better quality of service (QoS) for multiple compute applications sharing the GPU.The combination of both a new generation HBM2 memory from Samsung, and a new generation memory controller in Volta, provides 1.5x delivered memory bandwidth versus Pascal GP100 and greater than 95% memory bandwidth efficiency running many workloads. HBM2 Memory: Faster, Higher Efficiency Volta’s highly tuned 16GB HBM2 memory subsystem delivers 900 GB/sec peak memory bandwidth.Finally, a new combined L1 Data Cache and Shared Memory subsystem significantly improves performance while also simplifying programming. Volta’s new independent thread scheduling capability enables finer-grain synchronization and cooperation between parallel threads. With independent, parallel integer and floating point datapaths, the Volta SM is also much more efficient on workloads with a mix of computation and addressing calculations. New Tensor Cores designed specifically for deep learning deliver up to 12x higher peak TFLOPs for training. The new Volta SM is 50% more energy efficient than the previous generation Pascal design, enabling major boosts in FP32 and FP64 performance in the same power envelope.
New Streaming Multiprocessor (SM) Architecture Optimized for Deep Learning Volta features a major new redesign of the SM processor architecture that is at the center of the GPU. This table 1 compares NVIDIA ® Tesla ® accelerators over the past 5 years. (Measured on pre-production Tesla V100.) COMPARISONS Right: Given a target latency per image of 7ms, Tesla V100 is able to perform inference using the ResNet-50 deep neural network 3.7x faster than Tesla P100. Figure 1: Left: Tesla V100 trains the ResNet-50 deep neural network 2.4x faster than Tesla P100.
0 kommentar(er)
