article thumbnail

What Adrian Did Next — Part 4 — how I helped Netflix launch on iPad and iPhone — 2007 to 2010

Adrian Cockcroft

We had some fun getting hardware figured out, and I used a 3D printer to make some cases, but the whole project was interrupted by the delivery of the iPhone by Apple in late 2007. Photo by Adrian2008 Netflix Hack Day Trophy for Instant Queue Add forNetflix In January 2010, exactly 15 years ago, Apple announced the iPad.

C++ 88
article thumbnail

Single-core memory bandwidth: Latency, Bandwidth, and Concurrency

John McCalpin

In my previous post , I reviewed historical data on single-core/single-thread memory bandwidth in multicore processors from Intel and AMD from 2010 to the present. The increase in single-core memory bandwidth has been rather slow overall, with Intel processors only showing about a 2x increase over the 13 years from 2010 to 2023.

Latency 71
Insiders

Sign Up for our Newsletter

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

article thumbnail

USENIX LISA2021 Computing Performance: On the Horizon

Brendan Gregg

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 on [slideshare] or as a [PDF]: I work on many areas of performance, but recently I've had a lot of demand to talk about BPF.

article thumbnail

Amazon EC2 Cluster GPU Instances - All Things Distributed

All Things Distributed

By Werner Vogels on 14 November 2010 04:00 PM. The early GPU systems were very vendor specific and mostly consisted of graphic operators implemented in hardware being able to operate on data streams in parallel. The Amazon.com 2010 Shareholder Letter Focusses on Technology. Comments (). General Purpose GPU programming.

AWS 125
article thumbnail

Platform Engineering Teams Done Right…

Adrian Cockcroft

We used this model effectively at Netflix when I was their cloud architect from 2010 through 2013. The layers of platforms start at the bottom with hardware choices such as which CPU architectures and vendors you want to use. There are three current underlying reasons for the platform engineering meme today.

article thumbnail

Expanding the Cloud - Cluster Compute Instances for Amazon EC2.

All Things Distributed

By Werner Vogels on 12 July 2010 05:00 PM. Additionally, many high-end HPC applications take advantage of knowing their in-house hardware platforms to achieve major speedup by exploiting the specific processor architecture. The Amazon.com 2010 Shareholder Letter Focusses on Technology. All Things Distributed. Comments ().

Cloud 96
article thumbnail

Structural Evolutions in Data

O'Reilly

Doubly so as hardware improved, eating away at the lower end of Hadoop-worthy work. ” There’s as much Keras, TensorFlow, and Torch today as there was Hadoop back in 2010-2012. Google goes a step further in offering compute instances with its specialized TPU hardware. Those algorithms packaged with scikit-learn?

Hardware 134