Blog

How to Reduce Machine Learning Inference Cloud Cost

Learn all about how to leverage various techniques to reduce your machine learning training and cloud costs

How Does YOLOv6 Differ From YOLOv5 or YOLOX?

Dive into an in-depth analysis of the YOLOv6 detection model, and how it compares to other members of the YOLO family like YOLOv5 and YOLOX.

How to Identify the Bottlenecks in your NVIDIA DeepStream™ Pipeline

Have issues with your DeepStream pipeline? Learn how to identify and debug bottlenecks, and use Nsight Systems to speed up your application.

How To Build Your First DeepStream™ Pipeline

An overview of DeepStream and Gstreamer, including how to use some of the most popular plugins available that developers can use.

How to Maximize Your Model’s Performance on NVIDIA® Jetson AGX Orin™

Learn how AI teams can accelerate the performance of their deep learning models using Jetson AGX Orin and the Deci platform.

Intel Collaboration With Deci Boosts AI Performance on Intel Hardware

The Deci-Intel collaboration continues to uncover breakthrough performance for computer vision and NLP models on Intel hardware.

How to Scale Up Real-Time AI Video Analytics on the Edge

Video analytics are being deployed across verticals. Learn how to scale up real-time edge AI video analytics with a hardware-aware solution.

8 Questions and Practical Tips on Deep Learning Model Selection

In the process of selecting deep learning models? Read experts' thoughts on new architectures, tips, and best practices on choosing CNNs.

Introduction to Knowledge Distillation

Take a deep dive into knowledge distillation, its training process, and the problems this process helps to solve.

Challenges and Best Practices in Productizing Deep Learning Models

Read the highlights from our AMA session on AI productization including challenges and tips around deep learning model dev and deployment.

MLPerf: Intel and Deci Boost NLP Models – Reaching Faster and More Accurate Inference Performance

Deci’s DeciBert Models coupled with Intel's OpenVINO toolkit deliver better speed than BERT-Base, and higher accuracy than BERT-Large.

MLPerf: How Deci and Intel Achieved up to 16.8x Throughput Increase and +1.74% Accuracy Improvement

This marks another significant milestone in the ongoing Deci-Intel collaboration towards enabling deep learning outstanding inference on CPUs

Blog
Product

SuperGradients – The Simplest Way to Build Deep Learning Models

Looking to train your deep learning models? SuperGradients is an open source training library with SOTA models with unbeatable accuracy. Learn more here.

13 Free Resources and Model Zoos for Deep Learning and Computer Vision Models

Starting a deep learning project? Here are 13 free resources and model zoos for computer vision models that can help simplify your work.

Overview and Comparison of Neural Network Training Libraries

Explore common PyTorch-based neural network training libraries and tools before selecting the most appropriate one for your use case.

An Overview of State of the Art (SOTA) DNNs

SOTA DNNs can help handle image data. In this post, we cover important SOTA DNNs in classification, object detection & semantic segmentation.

Blog
Event

Top 12 Talks from NVIDIA GTC 2022 to Get You Started

Attending GTC 2022 and not sure which sessions to watch? Check out these top 12 talks for a comprehensive overview of the AI landscape today.

Can You Close the Performance Gap Between GPU and CPU for Deep Learning Models?

How can we optimize CPU for deep learning models' performance? This post discusses model efficiency and the gap between GPU and CPU inference. Read on.

5 Engineering Best Practices for Deep Learning Model Deployment on NVIDIA Jetson

Learn how you can build a model that is not too complex or large to run on the edge but still makes the most of the available hardware.

5 Must-Have Tricks When Training Neural Networks

Are you making the most of your neural network training? Here are the top items for improving your deep learning training bag of tricks.

How to Convert a PyTorch Model to TensorRT™ and Deploy it in 10 Minutes

Learn how to convert a PyTorch model to NVIDIA’s TensorRT™ model in just 10 minutes. It’s simple and you don’t need any prior knowledge.

How to Find the Best Hardware for Deep Learning Model Inference

Finding the right hardware for your model inference can be a daunting task, here are 4 key parameters to consider when you select hardware for inference.

Intel and Deci Collaboration: Deci Breaks the AI Barrier (Again)

After the joint submission to MLPerf, Intel and Deci collaborated again to accelerate three off-the-shelf models: ResNet-50, ResNeXt101, and SSD MobileNet V1.

Exciting News: Deci Raises $21 Million in a Series A Funding Round

Join us on our journey towards making AI efficient, accessible, and scalable for all organizations.

Make the Most of Your Jetson’s Computing Power for Machine Learning Inference

The boost in inference speed of your Deep Learning model compared to your current runs can be as much as 8 times. By the end of this article, you’ll know how to apply it to your use case with minimal effort.

5 Factors that Impact the Inference Pipeline in Production + Hardware Usage Metrics

To achieve optimal performance, the entire inference pipeline in production needs to be fast—not only the model. Here are 5 ways to optimize it.

DeciNets: How Deci’s AutoNAC Automatically Extended the Efficient Frontier of Accuracy-Latency for NVIDIA Cloud and Edge GPUs

Learn how AutoNAC creates proprietary DeciNet architectures that extend the efficient frontiers of latency/accuracy tradeoffs on NVIDIA cloud and edge GPUs.

Blog
Product

Infery – Run Inference for Optimized Models with Only 3 Lines of Python Code

Infery is a Python runtime engine that transforms running inference on optimized models into a light and easy process.

How to Convert a PyTorch Model to ONNX in 5 Minutes

Learn about ONNX and how to convert a ResNet50 model to ONNX. Then, try optimizing it to reduce its latency and increase its throughput.

The Object Detection Landscape: Accuracy vs Runtime

Learn about the different types and technical implementations of object detection algorithms, a key domain of deep learning and computer vision.

Podcast: Key Benefits of Using an Inference Acceleration Platform

Listen to this episode of The Data Exchange Podcast to know the reasons why you should optimize your deep learning inference platform.

LISTEN NOW

Graph Compilers for Deep Learning: Definition, Pros & Cons, and Popular Examples

Let’s take deeper dive into how graph compilers work to better understand how, when used correctly, they can offer enormous amounts of acceleration.

How to Automatically Double YOLOv5 Performance in Less Than 15 Minutes

In this post, you'll learn how the Deci platform can optimize your machine learning models. We use the YOLOv5 in our example, but the platform allows you to optimize any model.

The Future of Video Conferencing Looks Promising with AI

Remote working trends accelerated, prompting a surge in demand for video conferencing apps and innovative communication techniques.

How Deci and Intel Hit 11.8x Inference Acceleration at MLPerf

MLPerf provides fair and standardized benchmarks for measuring inference performance of machine and deep learning. Our submission to MLPerf proved that our AutoNAC technology reduces runtime while preserving the accuracy of the base model.

The Next Innovation Wave: AI Building Better AI

One of the reasons why most businesses struggle to obtain real value from AI is the high algorithmic complexity of deep learning models.

Blog
Product

Deci’s Deep Learning Acceleration Platform is Now Available

Take your models to production faster than ever. The platform combines multiple algorithmic and software techniques in an automatic and managed manner.

Toward Defining AutoML for Deep Learning

To date, there is no widely-accepted definition of AutoML, and the industry is using this term to refer to many different things.

Hardware for Deep Learning Inference: How to Choose the Best One for Your Scenario

Deep neural networks have become common practice in many machine learning applications.

An Introduction to the Inference Stack and Inference Acceleration Techniques

Deep neural networks can achieve state-of-the-art performance on many machine learning tasks.

Blog
Product

Deci RTiC – The Case for Containerization of AI Inference

One of the main obstacles impeding the effective utilization of AI models, is their deployment on compute resources, within their target environments.

The Correct Way to Measure Inference Time of Deep Neural Networks

Accurately measuring the inference time of neural networks is not as trivial as it sounds. In this post, we review some of the main issues that should be addressed to measure latency time correctly.

Are All FLOPs Created Equal? A Comparison of FLOPs vs Runtime

Analyses of efficient models often assume that FLOPs can serve as an accurate proxy for efficiency, including run-time. But this is wrong.

Efficient Inference in Deep Learning – Where is the Problem?

Convolutional neural networks can provide superior solutions for many computer vision tasks such as image classification, segmentation, and detection.

The New Electricity Requires Faster Wiring

We’re in the middle of a vast technological revolution, which is transforming life as effectively as the industrial revolution did in the 19th century.

Featured Blog

SuperGradients – The Simplest Way to Build Deep Learning Models

DeciNets: How Deci’s AutoNAC Automatically Extended the Efficient Frontier of Accuracy-Latency for NVIDIA Cloud and Edge GPUs

How to Reduce Machine Learning Inference Cloud Cost

How Does YOLOv6 Differ From YOLOv5 or YOLOX?

How to Identify the Bottlenecks in your NVIDIA DeepStream™ Pipeline

How To Build Your First DeepStream™ Pipeline

How to Maximize Your Model’s Performance on NVIDIA® Jetson AGX Orin™

Intel Collaboration With Deci Boosts AI Performance on Intel Hardware

How to Scale Up Real-Time AI Video Analytics on the Edge

8 Questions and Practical Tips on Deep Learning Model Selection

Introduction to Knowledge Distillation

Challenges and Best Practices in Productizing Deep Learning Models

MLPerf: Intel and Deci Boost NLP Models – Reaching Faster and More Accurate Inference Performance

MLPerf: How Deci and Intel Achieved up to 16.8x Throughput Increase and +1.74% Accuracy Improvement

SuperGradients – The Simplest Way to Build Deep Learning Models

13 Free Resources and Model Zoos for Deep Learning and Computer Vision Models

Overview and Comparison of Neural Network Training Libraries

An Overview of State of the Art (SOTA) DNNs

Top 12 Talks from NVIDIA GTC 2022 to Get You Started

Can You Close the Performance Gap Between GPU and CPU for Deep Learning Models?

5 Engineering Best Practices for Deep Learning Model Deployment on NVIDIA Jetson

5 Must-Have Tricks When Training Neural Networks

How to Convert a PyTorch Model to TensorRT™ and Deploy it in 10 Minutes

How to Find the Best Hardware for Deep Learning Model Inference

Intel and Deci Collaboration: Deci Breaks the AI Barrier (Again)

Exciting News: Deci Raises $21 Million in a Series A Funding Round

Make the Most of Your Jetson’s Computing Power for Machine Learning Inference

5 Factors that Impact the Inference Pipeline in Production + Hardware Usage Metrics

DeciNets: How Deci’s AutoNAC Automatically Extended the Efficient Frontier of Accuracy-Latency for NVIDIA Cloud and Edge GPUs

Infery – Run Inference for Optimized Models with Only 3 Lines of Python Code

How to Convert a PyTorch Model to ONNX in 5 Minutes

The Object Detection Landscape: Accuracy vs Runtime

Podcast: Key Benefits of Using an Inference Acceleration Platform

Graph Compilers for Deep Learning: Definition, Pros & Cons, and Popular Examples

How to Automatically Double YOLOv5 Performance in Less Than 15 Minutes

The Future of Video Conferencing Looks Promising with AI

How Deci and Intel Hit 11.8x Inference Acceleration at MLPerf

The Next Innovation Wave: AI Building Better AI

Deci’s Deep Learning Acceleration Platform is Now Available

Toward Defining AutoML for Deep Learning

Hardware for Deep Learning Inference: How to Choose the Best One for Your Scenario

An Introduction to the Inference Stack and Inference Acceleration Techniques

Deci RTiC – The Case for Containerization of AI Inference

The Correct Way to Measure Inference Time of Deep Neural Networks

Are All FLOPs Created Equal? A Comparison of FLOPs vs Runtime

Efficient Inference in Deep Learning – Where is the Problem?

The New Electricity Requires Faster Wiring

Unleash Your Deep Learning Models

Unleash Your
Deep Learning Models