Intel's (INTC) Habana Labs Secures AWS Deal for Gaudi Processors

Intel Corporation’s INTC Habana Labs recently secured deal for its Gaudi processors from Amazon’s AMZN Amazon Web Services (“AWS”). AWS will utilize up to eight Gaudi accelerators for its upcoming Elastic Compute Cloud (“EC2”) instances for machine learning (“ML”) workloads.

AWS added that the Gaudi accelerators will offer high performance at relatively lower costs. Per AWS estimates, Gaudi chips enable up to 40% improved price performance compared with existing graphics processing unit-based EC2 instances used for ML tasks.

The accelerators will also support PyTorch and TensorFlow along with other such common frameworks. AWS’s EC2 instance based on Gaudi accelerators will be capable of processing 12,000 images per second based on the ResNet-50 model on TensorFlow.

The launch of Gaudi-powered EC2 instances will help AWS to reduce cost for its customers pertaining to deployment of AI-powered training models. Gaudi-based EC2 instances are expected to be available in the first half of 2021.

Intel Corporation Price and Consensus

 

Intel Corporation Price and Consensus

Intel Corporation price-consensus-chart | Intel Corporation Quote

 

AWS will also leverage Habana Labs’ SynapseAI Software Suite. The software solutions will enable developers create new training models as well as port exiting models to Gaudi accelerators from GPUs.

AWS deal is important for Intel as it expands its presence in the lucrative AI silicon market and fortifies its position against other chipmakers like NVIDIA NVDA and Advanced Micro Devices AMD.

Increasing Demand for AI-based Chips Bodes Well

Rapid proliferation of AI in industries like retail and automotive and manufacturing, which are also witnessing increasing cloud adoption will drive Cloud AI market, per Mordor Intelligence report. The Cloud AI market is expected to witness a CAGR of 20.3% between 2020 and 2025.

As cloud migration by enterprises picks up pace, it will boost the need for datacenters. The expansion of data center networks by operators is driving demand for the AI-powered GPUs, which are able to better execute complex workloads.

Intel continues to focus on accelerating its AI-based product portfolio and capitalise on the opportunity presented by cloud and data center markets. Intel’s acquisition of Habana Labs, which focuses on AI chip development, was in sync with this strategy.

Intel acquired Habana Labs in 2019 for nearly $2 billion. The acquisition has added standards-based programming environment to manage AI workloads as well as high-performance training processors to Intel’s portfolio.

Habana Labs’ Gaudi processors are designed for various training deep learning models including natural language processing (NLP), object detection and ML training. Gaudi-based large node training systems offer higher throughput compared with systems built with the equivalent number of GPUs.

Headwinds Persist

Intel is witnessing tough competition in the AI trained inference server chip market.

NVIDIA possesses a first-mover advantage in the AI chips domain. The company has an impressive product portfolio which positions it well to capitalize on the growing adoption of AI in various verticals. Moreover, the company teamed up with the prominent OEMs including Dell and Lenovo to build workstations that incorporate its new CUDAX AI accelerated software and Quadro RTX GPUs.

With the acquisition of ARM Holdings, NVIDIA will be able improve its inference technology, and development accelerators. The buyout will enable NVIDIA to provide end-to-end ecosystem of technology across several domains like data center, Internet of Things (IoT), mobile and autonomous vehicles.

Also, Xilinx’s Versal chips combined with AI inference capabilities pose a major concern. AMD recently announced its intent to acquire Xilinx in all stock deal worth $35 billion. The transaction is expected to conclude by the end of 2021.

Moreover, Intel’s manufacturing woes are a concern for investors. The company’s launch of chips based on 7 nanometer (nm) process has been delayed. Intel detected a major defect mode in 7 nm process, which caused yield degradation and hence led to the delay.

On the other hand, AMD is utilizing Taiwan Semiconductor Manufacturing Company’s 7 nm process technology, which will facilitate faster delivery of advanced 7 nm chips to market.

Shares of Intel, which carries a Zacks Rank #4 (Sell), have declined 16.6% compared with the industry’s growth of 36.6% on a year-to-date basis.

You can see the complete list of today’s Zacks #1 Rank (Strong Buy) stocks here.

These Stocks Are Poised to Soar Past the Pandemic

The COVID-19 outbreak has shifted consumer behavior dramatically, and a handful of high-tech companies have stepped up to keep America running. Right now, investors in these companies have a shot at serious profits. For example, Zoom jumped 108.5% in less than 4 months while most other stocks were sinking.

Our research shows that 5 cutting-edge stocks could skyrocket from the exponential increase in demand for “stay at home” technologies. This could be one of the biggest buying opportunities of this decade, especially for those who get in early.

See the 5 high-tech stocks now>>

Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report
 
NVIDIA Corporation (NVDA) : Free Stock Analysis Report
 
Amazon.com, Inc. (AMZN) : Free Stock Analysis Report
 
Intel Corporation (INTC) : Free Stock Analysis Report
 
Advanced Micro Devices, Inc. (AMD) : Free Stock Analysis Report
 
To read this article on Zacks.com click here.
 
Zacks Investment Research