Brendan Gregg

JULY 4, 2021

This was a chance to talk about other things I've been working on, such as the present and future of hardware performance. The video is on [youtube]: The slides are [here] or as a [PDF]: first prev next last / permalink/zoom I work on many areas of performance, but recently I've had a lot of demand to talk about BPF. Ford, et al., “TCP

Performance 52

Performance Latency Hardware Storage 52

Brendan Gregg

AUGUST 8, 2017

They are demand on the system, albeit for software resources rather than hardware resources. ## Decomposing Linux load averages Can the Linux load average value be fully decomposed into components? Latency was acceptable and no one complained. Yes, I'd say so. They aren't idle. 3 tar termb$ iostat -x 60. [.] termc$ uptime.

Latency 111

Latency Metrics C++ Systems 111

Brendan Gregg

FEBRUARY 28, 2023

My personal opinion is that I don't see a widespread need for more capacity given horizontal scaling and servers that can already exceed 1 Tbyte of DRAM; bandwidth is also helpful, but I'd be concerned about the increased latency for adding a hop to more memory. Ford, et al., “TCP

Performance 52

Performance Latency Cache Google 52

Alex Russell

APRIL 29, 2021

WebGL 2 launched for other platforms on Chrome and Firefox in 2017. For heavily latency-sensitive use-cases like WebXR, this is a critical component in delivering a good experience. is access to hardware devices. This allows customisation and use of specialised features without custom, proprietary software for niche hardware.

Media 145

Media Games Education Engineering 145

O'Reilly Software

APRIL 4, 2018

A message-oriented implementation requires an efficient messaging backbone that facilitates the exchange of data in a reliable and secure way with the lowest latency possible. It enables unbounded scalability as more commodity or specialized hardware can be seamlessly added to existing clusters.

Architecture 80

Architecture Scalability Google Operating System 80

Alex Russell

OCTOBER 22, 2017

— Alex Russell (@slightlylate) October 4, 2017. It simulates a link with a 400ms RTT and 400-600Kbps of throughput (plus latency variability and simulated packet loss). Simulated packet loss and variable latency, however, can make benchmarking extremely difficult and slow. Maybe "ambush by JS"? First Load.

Performance 74

Performance Network Benchmarking Mobile 74

Alex Russell

JANUARY 30, 2024

HTML, CSS, images, and fonts can all be parsed and run at near wire speeds on low-end hardware, but JavaScript is at least three times more expensive, byte-for-byte. If you or your company are able to generate a credible worldwide latency estimate in the higher percentiles for next year's update, please get in touch.

Performance 98

Performance Network Mobile Speed 98

John McCalpin

JANUARY 22, 2018

A peculiar throughput limitation on Intel’s Xeon Phi x200 (Knights Landing) Introduction: In December 2017, my colleague Damon McDougall (now at AMD) asked for help in porting the fused multiply-add example code from a Colfax report ( [link] ) to the Xeon Phi x200 (Knights Landing) processors here at TACC.

Latency 40

Latency Hardware Code Testing 40

John McCalpin

JANUARY 22, 2018

Introduction: In December 2017, my colleague Damon McDougall (now at AMD) asked for help in porting the fused multiply-add example code from a Colfax report ( [link] ) to the Xeon Phi x200 (Knights Landing) processors here at TACC. A peculiar throughput limitation on Intel’s Xeon Phi x200 (Knights Landing). cmpl $1000000000, %eax.

Latency 40

Latency Hardware Code Testing 40

Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. As a Xen guest, this profile was gathered using perf(1) and the kernel's software cpu-clock soft interrupts, not the hardware NMI. A quick check of basic performance statistics showed over 30% higher CPU consumption.

Speed 40

Speed Java AWS Virtualization 40

Smashing Magazine

JANUARY 7, 2019

Estimated Input Latency tells us if we are hitting that threshold, and ideally, it should be below 50ms. On the other hand, we have hardware constraints on memory and CPU due to JavaScript parsing times (we’ll talk about them in detail later). Parsing and executing times vary significantly depending on the hardware of a device.

Performance 111

Performance Cache Network Metrics 111