We take a tour of HPE’s Technology Renewal Centre to see how old tech is getting a second life
THE SO WHAT
Enterprise hardware is starting to look more like a circular asset class—HPE’s renewal pipeline turns decommissioned gear into a structured supply source. Infra leaders should factor certified refurbished into their capacity planning, especially for non-cutting-edge workloads where capex and lead times matter more than bleeding-edge performance.
READ THE SOURCE
MORE FROM THE WIRE
Deep & Emerging TechJapan’s GMO Internet Requests Government Cyber Defense Measures
A private operator asking the Japanese government for active cyber defense is another data point that large-scale breaches are outgrowing what individual firms can handle alone. If you run critical infrastructure or large consumer platforms, assume more state involvement in incident response and plan governance and disclosure accordingly.
Deep & Emerging TechThe UK plans to invest £100M to create the AI Biosecurity Lab to boost defenses against biological threats, supported by OpenAI, Anthropic, and Google DeepMind
AI labs are being pulled directly into national biosecurity architecture — £100M for the UK’s AIBL formalizes that AI capability is now part of biodefense infrastructure. If you touch biology, data, or models, expect tighter scrutiny on dual-use risk and more demand for auditable safety controls baked into your stack.
Deep & Emerging TechUK Vows £100 Million for Biosecurity Lab With OpenAI, Anthropic
A £100M AI Biosecurity Lab with OpenAI, Anthropic, and Google DeepMind formalizes bio-risk as a first-class AI policy and research domain. Healthcare, biotech, and cloud vendors touching biological data should expect more pre-clearance, audits, and specialized guardrails baked into deals and APIs.
Deep & Emerging Tech32 TB RAM for your GPUs: Marvell shows revolutionary Photonic Fabric Memory Module that uses light to link up to 16 GPUs and slash CAPEX — but Nvidia won't stick around doing nothing
Optically pooled 32 TB memory across up to 16 GPUs turns the GPU from a memory controller into a pure compute engine—if it ships on Marvell’s 2027 timeline, cluster design and capex math change. Infra teams planning multi-year buildouts should at least scenario-plan around disaggregated, photonic memory as a future option.