Leaving Intel
A senior engineer reflects on his 3.5 years at Intel, highlighting contributions to AI flame graphs, eBPF, cloud strategy, and open-source performance tools.
BrendanGregg.com is the blog of Brendan Gregg, a performance engineer and computer scientist who shares deep expertise in system performance, observability, and kernel-level insights. His content covers topics like eBPF, Linux performance tools (perf, ftrace), flame graphs, and large-scale cloud systems analytics. Brendan writes with clarity about how to analyze bottlenecks, profile production systems, and tune infrastructure for speed and efficiency. The blog features both conceptual discussions and hands-on examples aimed at engineers tackling real performance problems in modern environments. Whether you are interested in kernel internals or high-level system design, BrendanGregg.com offers authoritative material grounded in years of experience.
13 articles from this blog
A senior engineer reflects on his 3.5 years at Intel, highlighting contributions to AI flame graphs, eBPF, cloud strategy, and open-source performance tools.
Brendan Gregg discusses AI agents trained on his performance engineering work, their limitations, and the ethical implications of creating 'Virtual Brendans'.
An Intel Fellow shares advice on how to give effective, constructive technical feedback to hardware vendors like Intel to influence product development.
Explains how hardware performance depends on software selection and tuning, using a three-stage rocket analogy for tech leaders.
Advice on when and why to form a computer performance engineering team, based on the author's experience at Netflix and Intel.
An engineer shares his 3-year experience of working remotely from Australia for a US firm, detailing the challenges of extreme timezone differences.
Analysis of Doom's gaming performance using GPU Flame Graphs and FlameScope for CPU/GPU correlation and shader compilation insights.
How eBPF technology can prevent system crashes like the massive July 2024 Windows outage caused by a faulty kernel driver update.
A list of essential Linux tools to pre-install for diagnosing performance issues and outages, including package names.
Explains how missing frame pointers in libc break profilers and debuggers, causing incorrect stack traces and misleading flame graphs.
Blog post about the new eBPF documentary, which tells the story of how the revolutionary Linux kernel technology was developed and accepted.
Explains why eBPF observability tools, designed for low overhead, are not suitable for security monitoring due to evasion risks.
Brendan Gregg's SREcon22 APAC keynote on the future of computing performance, covering new developments and predictions.