Intel Releases Continuous Profiler to Increase CPU Performance

March 11, 2024 Published
A person stands in a data center, working on a laptop next to a row of tall, illuminated server racks. The room is bathed in blue and purple lighting, creating a modern and high-tech atmosphere.

Cloud-native code profiling is more important than ever to help applications run at peak efficiency and improve the customer experience while saving time and resources.

What’s New: Intel today announced the release of Continuous Profiler to open source -- serving as an example of the company’s open ecosystem approach to catalyze innovation and boost productivity for developers. The optimization agent is actively used by companies including ironSource, ShareChat and Snap Inc. to identify production bottlenecks and optimization opportunities. Developed by Intel® Granulate™ and contributed to the open source community, Continuous Profiler is a solution that combines multiple profilers into one view as a flame graph. This unified view offers developers, performance engineers and DevOps a continuous and autonomous way to identify runtime inefficiencies.

“Continuous Profiler has been at the heart of what we’ve been doing at Intel Granulate. By helping developers identify bottlenecks in the code, businesses can optimize their applications more easily and effectively.”

–Asaf Ezra, general manager of Intel Granulate
Why It Matters: Determining why central processing units (CPUs) are busy is a routine task for performance analysis in any testing and production environment. Continuous Profiler delivers a flame graph of the hottest code paths.“This visualized view makes it immediately obvious where CPU is consumed so you can find cost savings, eliminate bottlenecks, improve throughput, and reduce latency and performance regressions,” said Brendan Gregg, Intel Fellow. “In today's complex environment, however, flame graphs can unearth so many performance wins that it becomes laborious to apply them all. Intel Granulate automates this task, allowing companies to realize these performance wins now and in the future as Intel develops more optimizations.”

How It Works: The evolution of gProfiler, Continuous Profiler combines multiple sampling profilers to produce a unified visualization of what a CPU is spending time on. Easy-to-navigate flame graphs pinpoint performance regressions, such as garbage collection, deadlocks and others to help smooth deployments.

Among its unique features, Continuous Profiler allows DevOps to filter services based on container name, hostname or Kubernetes deployment object by simply selecting the resolution level within the platform. These native Kubernetes filters profile down from the deployment to pod level without having to deploy a profiler for each object. This allows teams to investigate the behavior of different deployments, pods, nodes and hosts across different regions and code.

The profiler is compatible with Intel Granulate’s continuous optimization services and can be deployed cluster-wide within minutes, supporting a range of programming languages without requiring code changes.

Continuous Profiler is SOC2-certified and held to Intel's high security standards, ensuring reliability and trust in its deployment.

What’s Next: With a simple user interface (UI) and seamless integration, Continuous Profiler will add modifications to meet unique developer needs, uncover bottlenecks on the runtime level and boost application performance for improved user experiences.

Together with open source contributors and the broader ecosystem, Intel will further refine the tool and help developers make applications faster, stronger and more cost-efficient.

About Intel at KubeCon EU: At KubeCon EU, Intel will have two booths. Being a part of the open source community means being an active contributor and Intel is active with over 100+ open internal projects, contributing to 300+ open source projects, and belonging to 700 foundations. Learn more at the Intel Open Ecosystem Booth #H5. Check out the Intel Enterprise Software Booth #J17 to see how Intel solutions help companies manage, secure and scale cloud-native workloads more efficiently and confidently – paving a faster way to AI.

More Context: For more information on the Intel Software Continuous Profiler, visit the Intel Granulate website and Continuous Profiler GitHub page. For more information, contact Granulate.io.

The Small Print:

Notices and Disclaimers:

Intel technologies may require enabled hardware, software or service activation. No product or component can be absolutely secure. Your costs and results may vary.

Related Posts

News Artificial IntelligenceImage of an Intel Arc 5800 graphics card against a digital background with a series of circular gradient patterns in purple, blue, and pink hues. The card features multiple display ports and a sleek, modern design.

Intel Launches Arc B-Series Graphics Cards

Intel Arc B580 and B570 GPUs offer modern gaming features with best-in-class value for performance.

December 3, 2024
News SoftwareLogo of PyTorch featuring a stylized red flame icon followed by the word PyTorch in black text.

Intel Boosts AI Development with Contributions to PyTorch 2.5

New features enhance the  programming experience for AI developers across data center and client hardware, expanding support for Intel GPUs.

October 17, 2024
News SoftwareA person sits smiling on a chair in a modern office setting with wooden walls and potted plants. To the right, the blue Intel logo is displayed prominently.

The Linchpin to Better Business Results with AI is a System Strategy

Making AI accessible across cloud, network, PC and edge requires a fundamental shift in the industry’s approach.

July 23, 2024
News Artificial IntelligenceA close-up image of a computer chip with a blue and black design. The chip has visible circuitry and details with a rectangular metallic border, emphasizing its technological complexity.

Intel’s Lunar Lake Processors Arriving Q3 2024

Lunar Lake processors will extend unmatched global AI PC scale and receive free updates to Copilot+ when available.

May 20, 2024