Suggestions
Owen Rogers
Analysing how cloud, data centers and AI are driving semiconductor supply and demand.
Owen Rogers is a prominent figure in the cloud computing and technology industry, currently serving as the Research Director of cloud, datacenter, and cloud AI services at TechInsights.12 He joined TechInsights in September 2023, bringing with him a wealth of experience and expertise in the field.2
Background and Education
Owen holds a PhD in Economics and Philosophy, which he completed in 2013 with a thesis focused on the economics of cloud computing.13 He also has a Master of Engineering degree in Computer Systems Engineering.3
Professional Experience
Prior to his role at TechInsights, Owen had a distinguished career that included:
- Product management in the hosting industry1
- Cloud industry analyst1
- Research Director of Cloud & Managed Services at 451 Research, where he led the firm's Digital Economics Unit1
Expertise and Contributions
Owen is known for his analytical skills and innovative approach to understanding technology markets. Some of his notable achievements and areas of focus include:
- Developing the Cloud Price Index, a benchmark indicator for public, private, and managed cloud costs1
- Analyzing how cloud, data centers, and cloud-based AI drive semiconductor supply and demand2
- Breaking through hype and speculation by creatively using data to uncover market realities1
Public Presence
Owen is recognized as a knowledgeable and engaging speaker in the industry.2 He maintains a presence on social media platforms like Twitter, where he shares insights on cloud computing, datacenters, and AI.4
At TechInsights, Owen continues to leverage his expertise to provide valuable insights into the interconnections between cloud computing, semiconductors, and emerging technologies like AI.
Highlights
AWS’s new Graviton 4 EC2 instances are now generally available in four AWS regions.
TechInsights ran a series of simple performance benchmarks on AWS’s r6g.large, r7g.large and r8g.large EC2 instances, based upon Graviton 2, 3, and 4 CPU respectively.
Although not comprehensive, our tests imply workloads running on Graviton 4 will perform better than those on Graviton 3, albeit at a greater instance cost per hour.
However, even when price is factored in, Graviton 4 demonstrates good value over Graviton 3.
Based on our limited set of MySQL benchmarks, Graviton 4 is 19% more price-performant than Graviton 2, and 4% more price-performant than Graviton 3.
More (free reg req): https://lnkd.in/eTiPxWgn
Ayush Jain
From 2025 through 2029, total cumulative power consumption of the world’s datacenter accelerators will be 2318 TWh, representing 1.5% of all the globes electricity. This consumption will produce 1.1 billion tons of carbon emissions.
Users of AI need to be aware that energy costs are likely to be a huge factor in total cost of ownership, and may impact return on investment. As adoption matures, users will need to use practices such as FinOps to optimize cost against value. Hyperscalers, in particular, must optimize to ensure they meet their ambitious carbon net zero targets.
More here (with free reg): https://lnkd.in/eWvSEvkY
