We Are Intel
Learn about Intel culture and the individuals who do something wonderful everyday
1015 Discussions

CCG Team Spotlight: How Intel’s Own Telemetry Data Helps Uncover and Solve Complex Issues

We_Are_Intel
Employee
0 0 2,851

We recently spoke with experts from our Data Collection & Analytics team to discuss the importance of telemetry data, an ambitious collaboration to help glean insights from a full petabyte of data and how this is all impacting Intel’s future.

First, what is telemetry data?

Telemetry data is used by Operating Software Vendors (OSVs) like Apple and Microsoft to improve user experiences and help prevent system errors. Intel needed its own telemetry collection program to solve end-user experience issues outside of a lab and en-masse, improve its products performance gen-over-gen, and to collaborate effectively with Original Equipment Manufacturers (OEMs).

Telemetry data helps us understand if users are receiving the same experience and performance, measured in the lab, during their daily PC use in the real world. This data can alert Intel and our OEMs to high or low errors and help us find the root cause. For example, does a system really offer 15 hours of battery life? Does it really wake in less than a second as promised by our instant wake? Does it overheat? Do certain programs impact CPU and memory more than they should? Overall, does the system perform the same outside of the lab? A positive user experience and reliable performance is imperative to our competitive strategy.

 

A petabyte of data

Once it started collecting the data from users who opt-in, the team had a big problem—a petabyte of telemetry data from 39 million Intel systems to be exact. The small-but-mighty Data Collection & Analytics (DCA) team, in the Client Platform Engineering team, analyzes real-world telemetry data from Intel systems, an opt-in rate of 30-40%; this means over 6 million consumers send data daily.

The team analyzes system health across CPU, RAM, storage, Wi-Fi, web usage, power consumption, load times, programs, event logs and thermals to drive informed business decisions. These insights significantly help Intel, and our customers, improve user experience, system performance, and more. Yes, the DCA team is comprised of talented engineers and data scientists, but with such an extraordinary amount of data to process, they needed help.

We_Are_Intel_3-1677258133459.png

 

DCA’s solution

The team reached out to University of California San Diego (UCSD) to partner with professors and students who are studying the latest methods of data science. UCSD has received ~$75 million for a new Data Science Institute, offering a Bachelor’s, Master’s and PhD program dedicated to the art of data science. First, the DCA team needed to ensure data privacy and security with a strong legal framework for sharing data to protect Intel and our customers. The program started small in 2019 with one intern, one professor, three students, and a small amount of funding.

We_Are_Intel_4-1677258133475.png

The initial objectives were to

  1. Collaborate with Intel subject matter experts and university students and professors
  2. Build upon Intel’s technical pipeline of PEs to recruit future Intel employees via internships
  3. Apply data science techniques to create actionable insights
  4. And publish analytics, insights, and results for Intel and the industry

 

Three years later, the results are incredible

The program has been extended to the University of Washington, another highly ranked data science university, and in 2020, the Telemetry Data Science Center of Excellence (Telemetry CoE) was born when Intel Labs contributed Membership-level funding for the next three years.

As of 2022, the Telemetry CoE has achieved:

We_Are_Intel_0-1677258133384.png

*Note: Internships and hiring were conducted prior to the current fiscal environment.

Some of the past and future topics include:

  1. Happiness Project - identifying sources and root causes of everything from slow system responses to blue-screen-of-death
  2. Web-Based Computing - studying power and cycles consumed by browser and web-based applications which are now responsible for more than 60% of PC CPU cycles worldwide
  3. COVID Time Series Analysis - enabling us to better understand and predict PC usage patterns
  4. Carbon Impact Studies - determining the Scope-3 carbon footprint impact as required by Intel’s RISE 2030 goals

We_Are_Intel_2-1677258133432.png

 

One of the pillars of Fearless Innovation is “to continuously improve, enabling our teams to be more curious, bold and innovative”. This team’s focus on driving innovation and results has led to a world-class collaboration and provides long-term insights for our competitive strategy. The team looks forward to collaborating with other universities, industry collaborators, and growing the Telemetry CoE’s impact.

We_Are_Intel_5-1677258133497.pngWe_Are_Intel_1-1677258133411.png