Hesam Nejati
My work spans HPC servers, distributed platforms, and cloud environments, with emphasis on performance modelling, low-level measurement, root-cause analysis, and intelligent analysis workflows for improving efficiency, resilience, and system behaviour.
High-Performance Computing · Distributed Systems · Cloud Infrastructure
I am a PhD researcher at University College Dublin working at the intersection of high-performance computing, distributed systems, and cloud infrastructure. My work focuses on performance analysis, energy-aware optimisation, observability, and reliability, combining rigorous systems research with practical engineering insight to improve how complex platforms are designed, measured, and operated.
Research focus
HPC, distributed systems, and cloud infrastructure
Work centered on performance analysis, energy efficiency, observability, and reliable system behavior in complex environments.
Measured outcomes
34%
Performance gains verified in controlled systems experiments
34% execution-time improvement in Linux and HPC performance studies, alongside 35% throughput gains in Cassandra cluster experiments.
Publications
6+
Published, submitted, and ongoing research across systems engineering
The public Scholar profile is one reference point, while the broader portfolio also includes submitted papers, active drafts, and applied research output beyond indexed records.
Applied engineering
Research turned into tools, systems, and operational ideas
From telemetry platforms to middleware and Linux tooling, the goal is to move from analysis to usable engineering output.
Policy pressure is reshaping infrastructure choices.
Efficiency, sustainability, and compute planning now move together.
The chart below is a visual summary of how this theme compares across three lenses: market momentum, energy importance, and performance or innovation pressure.
Market signal · Energy · Performance
snapshotMarket
68/100
Energy
82/100
Innovation
57/100
How Google, Microsoft And Amazon Are Racing To Solve The AI Energy Crisis
A current overview of how hyperscalers are responding to the power, cooling, and infrastructure strain created by fast-growing AI workloads.
These studies explain what was evaluated, in which environment, and what improvement or operational insight was achieved.
CPU Power
76 W
GPU Power
122 W
Forecast Load
68%
Workload distribution
Power forecast
Across modern computing infrastructure, from cloud platforms to high performance systems, there is a persistent and expensive blind spot. Systems are monitored, but not truly understood.
34%
Execution gain achieved
UCD / Linux
Primary test environment
Ingress
Request Router
Control
Coordinator
Consistency
Quorum Logic
Throughput
+35%
Replica sync
98.8%
Stale reads
-25%
Distributed systems are the backbone of modern applications, yet they are often managed rather than truly controlled. When behaviour becomes unpredictable, dashboards and partial metrics are not enough.
35%
Throughput improvement
25%
Stale reads reduced
Each one shows a clear technical purpose, the kind of problem it addresses, and the value it brings in practice.
Built to surface live kernel and system behaviour without the overhead of large observability stacks, helping shorten troubleshooting cycles and improve operational visibility.
2026
Year
In Progress
Mode
Building A
38.0 kW
HVAC
16.0 kW
Lighting
5.2 kW
Load by zone
Peak
40.0
Avg
29.0
Alert
1
In large scale infrastructure deployments, visibility is not optional. At Sadra Company, I worked on designing and deploying a distributed monitoring platform across 21 buildings, where the challenge was to collect, move, and trust data continuously at scale.
2014 - 2022
Year
Featured
Mode
This section brings together selected videos, written analysis, and public commentary focused on AI systems, distributed infrastructure, and energy-aware computing.
Useful as a high-level briefing on the real operational side of AI adoption: electricity supply, data-centre expansion, and the pressure to make compute growth sustainable.
A format for sharing short findings, snapshots from experiments, or paper milestones in a way that feels integrated with the rest of the site.
A compact section for X posts, especially useful once you begin publishing commentary around videos, conference notes, or systems experiments.