Improving Application Performance with NVMe Storage - Part 1
The Rise of AI and ML Driving Parallel Computing Requirements
April 29, 2019

Zivan Ori
E8 Storage

Share this

As computing technology and data algorithms have advanced over the years, the ways in which technology has been applied to real world challenges have grown more automated and autonomous. This has given rise to a completely new set of computing workloads for Machine Learning which drives Artificial Intelligence applications (aka AI / ML).

AI / ML can be applied across a broad spectrum of applications and industries. Financial analysis with real-time analytics is used for predicting investments and drives the FinTech industrys needs for high performance computing. Real-time image recognition is a key enabler for self-driving vehicles, while facial recognition is used by law enforcement across the globe. Manufacturing uses image recognition technology to spot defects in materials, organizations such as NOAA use satellite imagery to spot changes in weather, while social media platforms use image recognition to tag photos of friends and family.

What is common among these uses cases is the need for a high level of parallel computing power, coupled with a high-performance low latency architecture to enable parallel processing of data in real-time across the compute cluster. The "training" phase of machine learning is critical and can take an excessively long time, especially as the training data set grows exponentially to enable deep learning for AI.

With storage performance now recognized as a critical component of AI/ML application performance, the next step is to identify the ideal storage platform. Non-Volatile Memory Express (NVMe) based storage systems have gained traction as the storage media of choice to deliver the best throughput and latency. Shared NVMe storage systems unlock the performance of NVMe, and offer a strong alternative to using local NVMe SSDs inside of GPU nodes.

The Rise of GPUs for AI / ML

GPUs were originally created for high performance image creation, and are very efficient at manipulating computer graphics and image processing. Their highly parallel structure makes them much more efficient than general purpose CPUs for algorithms where the processing of large blocks is done in parallel. For this reason, GPUs have found strong adoption in the AI / ML use case as they allow for a high degree of parallel computing and current AI focused applications have been optimized to run on GPU based computing clusters.

With the powerful compute performance of GPUs, the bottleneck moves to other areas of the AI / ML architecture. For example, the volume of data required to feed machine learning requires massive parallel read access to shared files from the storage subsystem across all nodes in the GPU cluster. This creates a performance challenge that NVMe shared storage systems are ideally suited to address.

Shared NVMe Storage for High Performance Machine Learning (ML)

One of benefits of shared NVMe storage is the ability to create even deeper neural networks due to the inherent high performance of shared storage, opening the door for future models that cannot be achieved today with non-shared NVMe storage solutions.

Today, there are storage solutions that offer patented architectures built from the ground up to leverage NVMe. The key to performance and scalability is the separation of control and data path operations between the the storage controller software and the host-side agents. The storage controller software provides centralized control and management, while the agents manage data path operations with direct access to shared storage volumes.

While AI / ML workloads are run exclusively on the GPUs within the cluster, that doesn't mean that CPUs have been eliminated from the GPU clusters completely. The operating system and drivers still leverage the CPUs, but while the machine learning training is in progress, the CPU is relatively idle. This provides the perfect opportunity for an NVMe based storage architecture to leverage the idle CPU computing capacity for a high performance distributed storage approach.

With NVMe protocol supporting exponentially more connections per SSD, the storage agents use RDMA to give each GPU node a direct connection to the drives. This approach enables the agents to perform up to 90% of the data path operations between the GPU nodes and storage, reducing latency to be on par with local SSDs.

In this scenario, running the NVMe based storage agent on the idle CPU cores of the GPU nodes enables the NVMe based storage to deliver 10x better performance than competing all-flash solutions, while leveraging existing compute resources that are already installed and available to use.

Read Part 2: Local versus Shared Storage for Artificial Intelligence (AI) and Machine Learning (ML)

Zivan Ori is CEO and Co-Founder of E8 Storage
Share this

The Latest

March 01, 2024

As organizations continue to navigate the complexities of the digital era, which has been marked by exponential advancements in AI and technology, the strategic deployment of modern, practical applications has become indispensable for sustaining competitive advantage and realizing business goals. The Info-Tech Research Group report, Applications Priorities 2024, explores the following five initiatives for emerging and leading-edge technologies and practices that can enable IT and applications leaders to optimize their application portfolio and improve on capabilities needed to meet the ambitions of their organizations ...

February 29, 2024

Despite the growth in popularity of artificial intelligence (AI) and ML across a number of industries, there is still a huge amount of unrealized potential, with many businesses playing catch-up and still planning how ML solutions can best facilitate processes. Further progression could be limited without investment in specialized technical teams to drive development and integration ...

February 28, 2024

With over 200 streaming services to choose from, including multiple platforms featuring similar types of entertainment, users have little incentive to remain loyal to any given platform if it exhibits performance issues. Big names in streaming like Hulu, Amazon Prime and HBO Max invest thousands of hours into engineering observability and closed-loop monitoring to combat infrastructure and application issues, but smaller platforms struggle to remain competitive without access to the same resources ...

February 27, 2024

Generative AI has recently experienced unprecedented dramatic growth, making it one of the most exciting transformations the tech industry has seen in some time. However, this growth also poses a challenge for tech leaders who will be expected to deliver on the promise of new technology. In 2024, delivering tangible outcomes that meet the potential of AI, and setting up incubator projects for the future will be key tasks ...

February 26, 2024

SAP is a tool for automating business processes. Managing SAP solutions, especially with the shift to the cloud-based S/4HANA platform, can be intricate. To explore the concerns of SAP users during operational transformations and automation, a survey was conducted in mid-2023 by Digitate and Americas' SAP Users' Group ...

February 22, 2024

Some companies are just starting to dip their toes into developing AI capabilities, while (few) others can claim they have built a truly AI-first product. Regardless of where a company is on the AI journey, leaders must understand what it means to build every aspect of their product with AI in mind ...

February 21, 2024

Generative AI will usher in advantages within various industries. However, the technology is still nascent, and according to the recent Dynatrace survey there are many challenges and risks that organizations need to overcome to use this technology effectively ...

February 20, 2024

In today's digital era, monitoring and observability are indispensable in software and application development. Their efficacy lies in empowering developers to swiftly identify and address issues, enhance performance, and deliver flawless user experiences. Achieving these objectives requires meticulous planning, strategic implementation, and consistent ongoing maintenance. In this blog, we're sharing our five best practices to fortify your approach to application performance monitoring (APM) and observability ...

February 16, 2024

In MEAN TIME TO INSIGHT Episode 3, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at Enterprise Management Associates (EMA) discusses network security with Chris Steffen, VP of Research Covering Information Security, Risk, and Compliance Management at EMA ...

February 15, 2024

In a time where we're constantly bombarded with new buzzwords and technological advancements, it can be challenging for businesses to determine what is real, what is useful, and what they truly need. Over the years, we've witnessed the rise and fall of various tech trends, such as the promises (and fears) of AI becoming sentient and replacing humans to the declaration that data is the new oil. At the end of the day, one fundamental question remains: How can companies navigate through the tech buzz and make informed decisions for their future? ...