We published nearly 100 technical blogs this year on the NVIDIA Developer Blog to help developers across a variety of industries develop their GPU-accelerated applications and the site had millions of page views. Below is a list of the most viewed posts from 2019 and some that you might have missed.
Most Viewed New Posts
- Jetson Nano Brings AI Computing to Everyone – Dustin Franklin, a Developer Evangelist on the Jetson team at NVIDIA, wrote the most viewed blog of the year that covers all the technical details you need to know about the Jetson Nano Developer Kit which was announced earlier this year at GTC.
- ArchiGAN: a Generative Stack for Apartment Building Design – Harvard Masters student Stanislas Chaillou summarized his thesis describing how you can leverage GANs to design floor plans, and entire buildings.
- NVIDIA Accelerates Real Time Speech to Text Transcription 3500x with Kaldi – Co-authored by NVIDIA researchers and engineers, learn how the GPU-accelerated Kaldi Speech Recognition Toolkit can be used to unlock extreme inference speedups for ASR workloads.
- GPUDirect Storage: A Direct Path Between Storage and GPU Memory – Just as GPUDirect RDMA (Remote Direct Memory Address) improved bandwidth and latency when moving data directly between a network interface card (NIC) and GPU memory, a new technology called GPUDirect Storage enables a direct data path between local or remote storage.
- Turing H.264 Video Encoding Speed and Quality – NVIDIA Turing hardware provides tremendous speed-ups for matrix computations, but the GPU also incorporates multimedia features such as an improved NVENC unit to deliver better compression and image quality in video codecs.
Five You May Have Missed
- NVIDIA Clocks World’s Fastest BERT Training Time and Largest Transformer Based Model, Paving Path For Advanced Conversational AI – The blog covers the latest advances at NVIDIA on two state-of-the-art NLP networks: BERT, and an 8.3 billion parameter version of a GPT-2 model known as GPT-2 8B, the largest Transformer-based network ever trained.
- Introduction to Ray Tracing in Unreal Engine 4.22 – NVIDIA Graphics engineers shared some tips and tricks for taking advantage of real-time ray tracing in UE4 4.22 which is currently in Preview and will launch in Final release soon.
- Object Detection on GPUs in 10 Minutes – Understand the components needed to setup an end-to-end object detection inference pipeline, how to apply different optimizations on GPUs, and how to perform inference in FP16 and INT8 precision on your pipelines.
- Real-Time Natural Language Understanding with BERT Using TensorRT – Step-by-step instructions explain how it is possible to perform BERT inference in 2.2 ms using TensorRT on T4 GPUs.
- Object Detection and Lane Segmentation Using Multiple Accelerators with DRIVE AGX – Co-authored by engineers from the NVIDIA autonomous driving group, this blog post dives into an application that runs two deep learning models concurrently to do both object recognition and ego-lane segmentation on an image.
Oldies, but Goodies
- An Even Easier Introduction to CUDA – You’ve heard about CUDA and you are interested in learning how to use it in your own applications. If you are a C or C++ programmer, this blog post should give you a good start.
- NVIDIA Turing Architecture In-Depth – Technical deep dive into the Turing GPU architecture that implements a new Hybrid Rendering model that combines real-time ray tracing, rasterization, AI, and simulation.
- Numba: High-Performance Python with CUDA Acceleration – Numba provides Python developers with an easy entry into GPU-accelerated computing and a path for using increasingly sophisticated CUDA code with a minimum of new syntax and jargon.
Thank you to all the developers that continue to educate themselves by reading posts on the NVIDIA Developer Blog to increase your technical knowledge of products from NVIDIA and our partners. Please comment below on a subject you would like us to cover in the upcoming year. See you in 2020!