Recommender systems, the economic engines of the internet, are getting a new turbocharger: the NVIDIA Grace Hopper Superchip.
Every day, recommenders serve up trillions of search results, ads, products, music and news stories to billions of people. They’re among the most important AI models of our time considering they’re incredibly constructive at finding in the internet’s pandemonium the pearls users want.
These machine learning pipelines run on data, terabytes of it. The increasingly data recommenders consume, the increasingly well-judged their results and the increasingly return on investment they deliver.
To process this data tsunami, companies are once raising accelerated computing to personalize services for their customers. Grace Hopper will take their advances to the next level.
GPUs Drive 16% Increasingly Engagement
Pinterest, the image-sharing social media company, was worldly-wise to move to 100x larger recommender models by raising NVIDIA GPUs. That increased engagement by 16% for its increasingly than 400 million users.
“Normally, we would be happy with a 2% increase, and 16% is just a beginning,” a software engineer at the visitor said in a recent blog. “We see spare gains — it opens a lot of doors for opportunities.”
The next generation of the NVIDIA AI platform promises plane greater gains for companies processing massive datasets with super-sized recommender models.
Because data is the fuel of AI, Grace Hopper is designed to pump increasingly data through recommender systems than any other processor on the planet.
NVLink Accelerates Grace Hopper
Grace Hopper achieves this considering it’s a superchip — two fries in one unit, sharing a superfast chip-to-chip interconnect. It’s an Arm-based NVIDIA Grace CPU and a Hopper GPU that communicate over NVIDIA NVLink-C2C.
What’s more, NVLink moreover connects many superchips into a super system, a computing cluster built to run terabyte-class recommender systems.
NVLink carries data at a whopping 900 gigabytes per second — 7x the bandwidth of PCIe Gen 5, the interconnect most leading whet upcoming systems will use.
That ways Grace Hopper feeds recommenders 7x increasingly of the embeddings — data tables packed with context — that they need to personalize results for users.
More Memory, Greater Efficiency
The Grace CPU uses LPDDR5X, a type of memory that strikes the optimal wastefulness of bandwidth, energy efficiency, topics and forfeit for recommender systems and other taxing workloads. It provides 50% increasingly bandwidth while using an eighth of the power per gigabyte of traditional DDR5 memory subsystems.
Any Hopper GPU in a cluster can wangle Grace’s memory over NVLink. It’s a full-length of Grace Hopper that provides the largest pools of GPU memory ever.
In addition, NVLink-C2C requires just 1.3 picojoules per bit transferred, giving it increasingly than 5x the energy efficiency of PCIe Gen 5.
The overall result is recommenders get a remoter up to 4x increasingly performance and greater efficiency using Grace Hopper than using Hopper with traditional CPUs (see orchestration below).
All the Software You Need
The Grace Hopper Superchip runs the full stack of NVIDIA AI software used in some of the world’s largest recommender systems today.
NVIDIA Merlin is the rocket fuel of recommenders, a hodgepodge of models, methods and libraries for towers AI systems that can provide largest predictions and increase clicks.
NVIDIA Merlin HugeCTR, a recommender framework, helps users process massive datasets fast wideness distributed GPU clusters with help from the NVIDIA Collective Communications Library.
Learn increasingly well-nigh Grace Hopper and NVLink in this technical blog. Watch this GTC session to learn increasingly well-nigh towers recommender systems.
You can moreover hear NVIDIA CEO and co-founder Jensen Huang provide perspective on recommenders here or watch the full GTC keynote below.
The post Why the New NVIDIA Grace Hopper Superchip Is Ideal for Next-Gen Recommender Systems appeared first on NVIDIA Blog.
