The debate over whether telco network workloads can run on the public cloud has raged for years. While greenfield operators like DISH Wireless have demonstrated that a cloud-native 5G network is possible, critics dismiss these examples as outliers that don’t apply to established networks. But recent developments with brownfield operators and new infrastructure options are changing the conversation about what’s possible.
In this episode, I’m talking with Dr. Ishwar Parulkar, Chief Technologist for Telecom at AWS. With extensive experience creating cloud solutions for telecom networks, Ishwar is at the forefront of helping operators transition their networks to cloud infrastructure.
Listen now to hear:
Wanna talk AI and public cloud? Telco execs, set up a meeting with our team to learn how to tap the immense business value it can bring.
Dr. Ishwar Parulkar is the Chief Technologist for Telecom and Edge Cloud at Amazon Web Services. In this role he is responsible for setting AWS technology strategy, defining new cloud services and leading initiatives to enable AWS’ edge cloud offerings and next generation telecom networks and services. Prior to AWS, Dr. Parulkar was a Distinguished Engineer at Cisco and Chief Architect for business units responsible for telecom routing, mobile packet core, small cell and network orchestration products. He was a founding member of industry-wide initiatives around mobile edge computing and 5G and the conviction that cloud technology can transform the telecom field brought him to AWS in 2016. Before Cisco, Dr. Parulkar was a Distinguished Engineer at Sun Microsystems where he led the design data center computing infrastructure, including the first multi-core processor systems and first compute virtualization platforms in the industry. He started his career at Apple, where he worked on the Mac desktop/laptop product lines and on the Newton PDA technology, which had the seeds of the iPhone revolution.
The Telco in 20 podcast is ranked in the top 5% of all podcasts globally by Listen Notes! 🎉 We’ve also won a 2024 MarComm Award, a 2024 and 2025 Hermes Award, and are recognized as a TeckNexus Top 12 Telco and Tech Podcast, Forrester Top 100 Channel Podcast and Feedspot Top 10 Telecom Podcast.
If you enjoy the podcast, would you leave us a review? It takes you seconds to do in your app and it really makes a difference in helping to convince hard-to-get guests. And I love reading your feedback and reviews!
Get my FREE newsletter, delivered every two weeks, with curated content to help telco execs across the globe move to the public cloud.
Set up a meeting with our team to learn how to tap the immense business value it can bring.
Yes, absolutely. AWS has created what it calls an “edge continuum”—a complete portfolio of infrastructure options that can support all network workloads whether you’re a greenfield operator like DISH Wireless or a brownfield operator like Telefónica Germany. This includes AWS regions, local zones for metro-level deployments, Outpost servers for on-premise facilities, and cloud RAN servers for far edge locations. The infrastructure is specifically designed with high performance, high bandwidth, and optimized instances to run network functions at scale with reliability.
Working with Telefónica Germany taught AWS how to deploy cloud infrastructure in a brownfield scenario where operators already have existing data centers, fiber infrastructure, and internet peering agreements. Unlike greenfield networks that have more flexibility, brownfield operators need to work within their existing footprints. AWS created a portfolio approach so Telefónica could choose the right infrastructure—whether regions, local zones, or Outpost servers—based on specific deployment requirements, data center locations, and where functions like the UPF need to be placed for latency or peering considerations. The operator successfully completed a trial to load one million subscribers into the public cloud.
AWS Outpost brings public cloud infrastructure to a telco’s premises. At MWC, AWS announced two key products: Outpost servers that run on-prem for network workloads, and cloud RAN servers for far edge locations. These offer the same AWS services and APIs available in regions, but are optimized for network functions requiring high throughput. The Outpost servers can run functions like UPF that need to be close to endpoints for latency, while cloud RAN servers use ARM-based Graviton3 processors for power efficiency and include integrated L1 accelerators for running DU stacks.
DR spoke at Telecom TV’s DSP Leaders World Forum on June 3-4 in Windsor, England, where she participated in a panel discussion titled, “Leveraging AI Throughout The Network.” She also hosted a garden party at the event. Additionally, she encourages telcos to jumpstart their AI strategies with cloud-based solutions like Totogi’s BSS Magic, emphasizing that competitive advantage comes from implementing AI solutions faster than competitors, which the public cloud makes possible.
First, generative AI—how telcos can use it for productivity in internal systems and networks, and to create new products. Examples include SK Telecom building a specialized AI model trained with Anthropic’s Claude on AWS, and British Telecom using Amazon Q to enhance developer productivity. Second, modernizing the network by running network functions on the cloud to make operations easier, more dynamic and flexible. Third, designing networks specifically for AI applications, as traffic patterns have evolved from voice to video, and now to large-scale AI applications.
According to Ishwar Parulkar, it’s not yet clear whether the RAN edge makes sense for AI applications. AWS sees three key points where AI workloads should run: on the device itself (where modern phones and VR glasses can handle inference), in the middle edge or metro-level data centers (for latency considerations or data locality), and in the cloud (where scale and multi-tenancy make it cheaper). The far edge at the RAN level is very close to devices, and most applications that would run there can already be handled by the devices themselves. Chris Penrose at NVIDIA has discussed AI-RAN concepts on a previous Telco in 20 episode, but the business case for running AI workloads at the RAN edge remains unclear.
Apple Podcasts 
Spotify 
YouTube 
