DCD Zero Downtime: The Bi-Weekly Data Center Show

As 2030 approaches, the pressure on data centers to meet their carbon pledges continues to increase. This, compounded with the challenge of energy supply insecurity and rising costs, is driving data centers to look to new methods of powering their operations, which has triggered nuclear power to come to the fore. However, as a result of historical disasters as well as the upfront costs, there is hesitancy in the data center industry to make the move to nuclear. This podcast will take a deep dive into the future of nuclear in the data center market, and explore how the heavy regulation of this industry makes it a safe and reliable method of energy generation. However, despite there being some clear use cases of its viability, we are many years off from nuclear becoming a widely used power source owing to its economic feasibility. This begs the questions: when will it be deployed at scale? What is the practicality of SMRs? What technology will be used to bridge the gap as we wait for a nuclear solution?

As data center demands grow, hyperscalers are building out vast networks that will help define the future of how infrastructure is interconnected. In the latest episode of the DCD Zero Downtime podcast, we connect with Juniper Networks' group vice president of cloud-ready data center, Mike Bushong, to discuss what this means, and what we should expect next. Plus we discuss AI, silicon photonics, telecoms, the race to 800G, and more.

Recently there’s been a number of mobile operators switching off their 3G networks around the world, or scheduling to phase out this old technology. 2G has also been spoken about at length, but has a slightly longer shelf-life than 3G, as it provides a backbone that still supports the IoT industry today. But what does any of this mean, and why does it matter? Well, Wireless Logic Group Chief Product Officer Paul Bullock explains why the need to switch off 3G is necessary for the future of 5G. It’s not just 5G, he adds, noting that IoT will gain a lot from the switch off of this service, although he acknowledges that 2G will be more of a challenge to phase out. Tune in to listen to the full discussion...

As land and power in Virginia's Loudoun County become increasingly scarce, data center developers are seeking fresh ground on which build. While many are moving south within Virginia to the likes of Prince William County, Quantum Loophole is hoping to lure companies north into Maryland. A former aluminum smelting works, the company's maiden campus is reportedly luring cloud companies and the hyperscale developers serving them on a huge scale. CEO Josh Snowhorn talks us through the company's history and its 2,000+acre gigawatt project located just north from Loudoun across the Potomac river.

Hunter Newby doesn't claim to have invented the Meet-Me Room, but he created one of the first, in the iconic and historical Manhattan facility, 60 Hudson Street, in about 1998 At the time data centers had separate connections from multiple carriers, and linking between those carriers meant running a link between their connection points - and paying them heavily. Hunter set up a room where the networks all met up for physical (layer 0) connections - and the rest is history. It's a well-documented history because he went on to write a series of magazine articles "Meet Me In New York", "Meet Me In Chicago" etc, about interconnection facilities in major US cities. That series is archived on his site, alongside live data of connections for each of those cities from the open peering resource, PeeringDB. Today, there are still not enough Meet-Me rooms. Some US states with large cloud facilities don't have nearby carrier hotels, so local user traffic has to go out to a city like Chicago and back. Hunter's wants to foster neutral carrier hotels where there are none. Why is 60 Hudson Street the Ellis Island of the Internet? And did Hunter ever write "Meet Me In St Louis"? Listen in, and meet Hunter Newby..

Join us for an extended discussion with CTO and founder of North American Access Technologies Julius Neudorfer about the history of liquid cooling, and where he sees the market developing in the future.

Formed by DigitalBridge and Liberty Global, AtlasEdge has quickly become one of the Edge industry's largest players. We chat to COO Zahl Limbuwala about how it defines the Edge and where it places its servers.

For years, operators have been trying to make their data centers more efficient by optimizing the cooling and power distribution in the facility - and have more or less ignored the IT equipment in the racks. PUE - the most popular data center metric - simply optimized the ancillary equipment, trying to approach a "perfect" figure of 1.0, where all the power goes into the rack. But what if you could measure the power used by your IT equipment, run comparative tests, and find out how to reconfigure, replace or update your hardware to improve energy efficiency? Rich Kenny claims to be able to do that. Interact started as a project by refurbished hardware supplier TechBuyer, aiming to prove that refurbished hardware could be as good as new kit, and grew into a consultancy that compares hardware configurations, and gives guidance on how to do more with less hardware. Server manufacturers do their business on the basis that new hardware is always more efficient than old hardware, and colo providers would rather you didn't find out that you could do away with half the racks you are renting. Kenny is shining a light into a dark corner that has not had enough exposure, dissecting the 1 in that PUE figure. He could be about to step on some industry toes...

In this episode, we talk to Cloudflare's CTO John Graham-Cumming about how to respond to an outage at your company. We also discuss why the company bans the use of the word "Edge," and talk about the company's 'Supercloud.'

Can light be used to compute? We talk to Nick Harris, CEO of chip startup Lightmatter, about how the company hopes to harness photons for AI computing and as a data center interconnect.