Showcasing a vision of technological innovation through GPUs in areas such as artificial intelligence, autonomous driving, and graphics

NVIDIA showcased various technologies that suggest the future of computing at the NVIDIA GPU Technology Conference (GTC), which was held successfully in San Jose, Silicon Valley, USA from March 26 to 29 (local time).

The four-day GTC 2018 attracted a record crowd of about 8,300 scientists, engineers, businesspeople, and media from around the world, and was filled with 600 sessions and seminars and about 150 exhibition booths. Founder and CEO Jensen Huang drew attention with NVIDIA's continuous innovation and incredible technological prowess in a keynote speech lasting about two hours on the 27th local time. The live broadcast of the keynote speech reached 550,000 viewers, indicating high interest in the future of computing.

At GTC 2018, NVIDIA presented an amazing vision of technological innovation through GPU, including autonomous driving, artificial intelligence (AI), graphics, and new platforms. Moore's law has been accepted as a fact for about 30 years, but the recent CPU development speed is gradually slowing down. On the other hand, GPU computing is developing at a speed that surpasses Moore's law, and is introducing innovation across the industry.

NVIDIA DRIVE Constellation is a suite of two server-based computing platforms that enable testing of autonomous vehicles in lifelike simulations, providing a safer, more scalable way to get self-driving cars on the road. The first server runs NVIDIA DRIVE Sim software to simulate a self-driving car’s sensors, such as cameras, lidar and radar, while the second server’s powerful NVIDIA DRIVE Pegasus AI car computer runs the entire self-driving car software stack and processes the simulated data as if it came from the sensors of a real car driving on the road.

At GTC 2018, CEO Jensen Huang unveiled a series of performance improvements to the company’s world-leading deep learning compute platform, announcing a 10x performance improvement over the previous generation in just six months on deep learning workloads. “The advances in deep learning we’re announcing today are just a glimpse of what’s to come,” he said in his keynote. “We’re dramatically accelerating the performance of NVIDIA’s deep learning platform at a rate that far outpaces Moore’s law, creating breakthroughs that will drive transformational change in healthcare, transportation, scientific exploration, and countless other areas.”

Key enhancements to the NVIDIA platform, which has been adopted by many of the world’s leading cloud service providers and server manufacturers, include a doubling of the memory of the NVIDIA Tesla V100, the world’s most powerful data center GPU, and a groundbreaking GPU interconnect fabric, NVIDIA NVSwitch™, along with an updated and optimized software stack.

NVIDIA has made another breakthrough in deep learning computing with the launch of NVIDIA DGX-2, the first single server capable of delivering two petaflops of compute power. DGX-2 delivers the deep learning processing power of 300 servers occupying 15 racks in a data center, but is 60 times smaller and 18 times more power efficient.

NVIDIA has introduced a new version of its TensorRT inference software, TensorRT 4, which can be used to rapidly optimize, validate, and deploy trained neural networks on hyperscale data centers, embedded, and automotive GPU platforms. It provides deep learning inference capabilities up to 190x faster than CPUs for common applications such as computer vision, neural network-based machine translation, automatic speech recognition, speech synthesis, and recommendation systems.

NVIDIA also announced the NVIDIA Quadro GV100 GPU with NVIDIA RTX technology, the first to deliver real-time ray tracing to millions of artists and designers around the world. NVIDIA RTX, combined with the powerful Quadro GV100 GPU, brings compute-intensive ray tracing to professional design and content creation applications in real time, representing the biggest advancement in computer graphics since the introduction of shader programs nearly 20 years ago. RTX technology is the culmination of nearly a decade of NVIDIA research, combining new GPU architectures, algorithms, and deep learning in a way that only NVIDIA can do.