Big Blue might be a little late to the AI acceleration game, but it has a captive audience in its System z mainframe and Power Systems servers. …
IBM Shows Off Next-Gen AI Acceleration, On Chip DPU For Big Iron was written by Timothy Prickett Morgan at The Next Platform.
Hardware is always the star of Nvidia’s GPU Technology Conference, and this year we got previews of “Blackwell” datacenter GPUs, the cornerstone of a 2025 platform that includes “Grace” CPUs, the NVLink Switch 5 chip, the Bluefield-3 DPU, and other components, all of which Nvidia is talking about again this week at the Hot Chips 2024 conference. …
Nvidia Rolls Out Blueprints For The Next Wave Of Generative AI was written by Jeffrey Burt at The Next Platform.
COMMISSIONED Organizations must consider many things before deploying generative AI services, from choosing models and tech stacks to selecting relevant use cases. …
Get Your Data House In Order To Unlock Your Generative AI Strategy was written by Timothy Prickett Morgan at The Next Platform.
Shipping netlab release 1.9.0 included running 36 hours of integration tests, including fifteen VXLAN/EVPN tests covering:
All tests included one or two devices under test and one or more FRR containers1 running EVPN/VXLAN with the devices under test. The results were phenomenal; apart from a few exceptions, everything Just Worked™️.
Effects are multiplicative, not additive, when it comes to increasing compute engine performance. …
Bechtolsheim Outlines Scaling XPU Performance By 100X By 2028 was written by Timothy Prickett Morgan at The Next Platform.
I’ve been working on new material over at Rule 11 Academy. This month’s posts are:
This brings us up to a total of 39 lessons. Each lesson should be about 15 minutes, so about 10 hours of material so far. The trial membership will take you through the end of the year. After the first of the year, the trial membership will last 2 months.
I love bashing SRv6, so it’s only fair to post a (technical) counterview, this time coming as a comment from Henk Smit.
There are several benefits of SRv6 that I’ve heard of.
Hi all, welcome to the 'Network CI/CD' blog series. To kick things off, let's ask the question, "Why do we even need a CI/CD pipeline for networks?" Instead of diving straight into technical definitions or showing you how to build a CI/CD pipeline, which might make you lose interest, we’ll focus on the reasons behind it. Why should network teams even consider implementing CI/CD?
In this post, we’ll talk about the benefits and the problems it solves, so you can see why it's worth learning. Let's get to it.
Even though I call it the “traditional way,” most of us (myself included) still make changes via the CLI. So, let’s imagine you and two colleagues are managing a campus network with 10 access switches. One of your tasks is to configure VLANs on all of Continue reading
When you are designing applications that run across the scale of an entire datacenter and that are comprised of hundreds to thousands of microservices running on countless individual servers and that have to be called within a matter of microseconds to give the illusion of a monolithic application, building fully connected, high bi-section bandwidth Clos networks is a must. …
This AI Network Has No Spine – And That’s A Good Thing was written by Timothy Prickett Morgan at The Next Platform.
Nvidia hit a rare patch of bad news earlier this month when reports started circulating claiming that the company’s much-anticipated “Blackwell” GPU accelerators could be delayed by as much as three months due to design flaws. …
When Nvidia Says Hot Chips, It Means Hot Platforms was written by Jeffrey Burt at The Next Platform.
Many network operators think the idea of building rather than buying is something that’s out of reach–but is it? Join Steve Dodd, Eyvonne, Tom, and Russ as we discuss the positive and negative aspects of build versus buy, what operators get wrong, and what operators don’t often expect.
I once again loved this episode of the Art of Network Engineering featuring Mike Bushong. He is a very astute judge of character as well as how to apply social skills to your tech role. Definitely listen to the above episode if you’re interested in countering cognitive biases.
In the episode, he told a great story of how he had a pivotal career moment with one of his managers that led to some important introspection. I won’t tell his story but the summary is that he had taken on way too much work and way too many roles and he blew up at his manager because of the stress. She leveled him with a quote that rang true for me:
“No one knows everything you’re working on. They just see that the thing that’s important to them is late.”
That’s not the verbatim quote but that’s how I remember it. It’s definitely something that I’ve been thinking about since the previous episode when he mentioned it the first time.
The odds are good that we’re all doing way too many things right now. Whether it’s doing more work in our role or taking on way Continue reading
The 2024 Democratic National Convention (DNC) wrapped up on Thursday, August 22, in Chicago, Illinois. Since our blog post about Internet trends during the first presidential debate between President Joe Biden and former President Donald Trump on June 27, the presidential race has fundamentally changed. We experienced the attempted assassination of Trump, the Republican National Convention (RNC), Biden’s late July withdrawal from the race, and Vice President Kamala Harris being selected as the Democratic nominee and participating in her party’s convention this week. Here, we’ll examine trends more focused on DNS traffic to news and candidate-related sites, cyberattacks targeting politically-related organizations, and spam and malicious emails mentioning the candidates’ names.
Over 60 more national elections are scheduled to take place across the world this year, and we have been monitoring them as they occur. Our goal is to provide a neutral analysis of their impact on Internet behavior, which often mirrors human activities. Significant events, such as the total eclipse in Mexico, the United States, and Canada, and the Paris 2024 Olympics, have had an impact on Internet traffic. Our ongoing election report on Cloudflare Radar includes updates from recent elections in the European Union, France, Continue reading
Pulumi, like Terraform and OpenTofu, has the ability to store its state in a supported backend. You can store the state in one of the blob/object storage services offered by the major cloud providers, via Pulumi’s SaaS offering (called Pulumi Cloud), or even locally. It’s this last option I’ll explore a little bit in this post, where I’ll show you how to configure Pulumi to store the state in the project directory instead of somewhere else.
Let me start with this disclaimer: If you’re working with a team of folks on IaC for your project or employer, don’t do this. Storing project state locally with your project will just make life difficult for you. Instead, just accept that you need to store the state somewhere that your whole team can access it. Howver, if you are a “team of one” then you might find this interesting or useful.
Pulumi supports a “local” backend, which means storing stack state information locally on the same system where Pulumi is running. By default, Pulumi will store the state information in the ${HOME}/.pulumi folder.
It’s possible to configure the location the local backend uses with the PULUMI_BACKEND_URL environment variable (see this page for Continue reading