Mirendil taps Google Cloud AI hypercomputer for research

datacenternews.asia Aug 7, 2026

Mirendil taps Google Cloud AI hypercomputer for research

Curated by Shubham · TechCapsules · Aug 7, 2026

AI-summarised brief · reviewed before publication

Frontier AI lab Mirendil has selected Google Cloud’s AI Hypercomputer to support its model development, gaining access to both Google TPU systems and Nvidia infrastructure. The partnership facilitates end-to-end training workflows, including large-scale reinforcement learning, by allowing Mirendil to match specific workloads to different hardware architectures. This strategic combination enables the startup to secure computing resources rapidly while optimizing performance across pre-training and post-training phases. Mirendil is currently operating a cluster of TPU v5P chips, with Nvidia-based systems scheduled for addition. The two companies collaborated on the infrastructure’s design, utilizing managed training clusters within the Gemini Enterprise Agent Platform to oversee both environments. This deployment provides Mirendil with access to the two primary AI hardware approaches used by developers. The arrangement supports Mirendil’s focus on building AI systems that accelerate research and development processes, addressing the growing dependency on specialized computing infrastructure for advanced model training.

💡 Why It Matters

· Mirendil’s dual-architecture strategy demonstrates a critical shift away from single-vendor dependency, allowing developers to mitigate supply chain risks while optimizing for specific computational needs.
· This move underscores how AI labs are leveraging hybrid cloud solutions to maintain agility in a market where hardware availability directly dictates research velocity.

Read full article on datacenternews.asia

Is AI Agent Pricing Getting Better? Grading My 2025 Predictions

Enterprise AI agent pricing remains complex, disrupting traditional per-user models as reasoning models and autonomous agents break established business structures. While predictions regarding consumption-based and action-based pricing have largely materialized, the widespread adoption of "AI credits" by major vendors like Salesforce, Microsoft, and SAP has introduced significant opacity. These credits simplify vendor billing but hinder customers’ ability to forecast costs or compare solutions across providers. Consequently, enterprises face difficulties in using price as a primary selection criterion. The industry continues to struggle with defining comparable metrics, making apples-to-apples evaluations nearly impossible without extensive negotiation. Analysts advise against pre-selecting vendors based solely on pricing models, noting that both SaaS and hyperscaler approaches remain in flux. As vendors adjust to maintain margins amid aggressive infrastructure build-outs, the lack of standardized pricing frameworks persists. This uncertainty complicates budgetary planning for organizations deploying agentic solutions, requiring deeper engagement between buyers and sellers to navigate the evolving landscape effectively.

💡 The opacity of AI credits prevents enterprises from accurately forecasting operational costs, forcing procurement teams to abandon standard competitive bidding processes. This shift empowers vendors to dictate terms, potentially locking organizations into expensive, non-comparable ecosystems before pricing standards stabilize.

nextplatform.com · Aug 6

Your Platform Was Built For A Different Era, And AI Just Exposed…

Google’s 2025 DORA study shows 90 % of enterprises now run internal platforms, yet many built for human‑centric, container‑based development are faltering under AI‑driven workloads. AI coding assistants have amplified code output, shifting bottlenecks from writing to delivering, while pipelines lack the capacity for the surge. New AI agents demand authentication, token regulation, GPU provisioning, scoped permissions and audit logging—capabilities most platforms do not natively provide. Rising AI infrastructure costs exacerbate cloud waste, with GPU instances and token usage escaping traditional FinOps controls. Additionally, AI expands attack surfaces, introducing risks such as prompt injection and model poisoning that existing SAST/DAST tools cannot detect, while regulatory pressures from the EU AI Act and U.S. executive orders tighten compliance demands. Platform Engineering 2.0 proposes an AI‑native, multi‑persona, cost‑aware evolution rather than a complete rebuild.

💡 Enterprises that fail to retrofit their platforms for AI workloads risk runaway cloud spend and security gaps, undermining the productivity gains AI promises.

forbes.com · Aug 6

What Are Hyperscalers? The Companies Driving The AI Infrastructure Boom

Major U.S. hyperscalers, including Amazon, Microsoft, Google, Meta, and Oracle, are projected to invest over $700 billion in AI computing infrastructure by 2026. This expenditure represents nearly double their spending from the previous year, driving a massive expansion of data center fleets worldwide. These facilities are critical for training advanced AI models and supporting daily digital services like streaming and cloud storage. However, this rapid growth places significant strain on local resources, particularly power grids and water supplies. Consequently, community pushback is increasing. In a notable example, Tucson, Arizona’s city council unanimously rejected Amazon’s Project Blue data center campus in 2025. As hyperscalers continue to expand, the location and construction methods of these facilities are becoming subjects of intense public negotiation and local regulatory scrutiny across the United States.

💡 The rejection of Amazon’s Project Blue in Tucson demonstrates that local municipalities now hold veto power over trillion-dollar tech expansions. This shift forces hyperscalers to treat community relations as a critical operational risk rather than a peripheral concern.

nypost.com · Aug 5

NASA chief Jared Isaacman tells ‘Pod Force One’ about his NIMBY solution…

NASA Administrator Jared Isaacman proposed placing data centers in orbit during a Wednesday appearance on “Pod Force One.” He argued that space‑based facilities could tap “free solar power” from fusion reactors, eliminating neighborhood noise complaints and reliance on terrestrial electricity. Isaacman cited the growing backlash against more than 1,500 data centers under construction, noting a July Fox News poll that shows 70 % of Americans oppose local AI‑related facilities. He linked public support for a lunar return—also at 70 %—to potential acceptance of orbital data hubs. While acknowledging that the economics may not materialize for five to ten years, Isaacman compared the timeline to Amazon’s early e‑commerce model, expressing confidence that space‑based data centers will eventually become viable.

💡 Orbiting data centers could bypass entrenched local opposition, unlocking new capacity for AI workloads without the political friction that stalls ground‑based projects.

evertiq.com · Aug 5

AI server shipments forecast raised to 31% YoY in 2026

TrendForce has revised its 2026 AI server shipment forecast upward to nearly 31% year-over-year growth, driven by surging infrastructure investments from major cloud service providers. The research firm estimates that the combined capital expenditures of nine leading CSPs, including Google, Amazon, Meta, Microsoft, Oracle, ByteDance, Tencent, Alibaba, and Baidu, will exceed USD 886.7 billion in 2026. North American hyperscalers account for nearly 90% of this spending. Demand for NVIDIA rack-scale platforms has strengthened, while Google and AWS prepare to ramp up next-generation in-house ASIC production in the second half of the year. Chinese providers are simultaneously accelerating domestic AI solution deployments. This spending supports generative AI and large-scale model requirements, encompassing GPU clusters and liquid-cooling infrastructure. Looking ahead, TrendForce projects the top nine CSPs’ combined capital expenditures will reach approximately USD 1.3 trillion in 2027, representing nearly 50% year-over-year growth, indicating sustained investment momentum despite a higher comparison baseline.

💡 The dominance of North American hyperscalers in capital expenditure underscores a widening technological gap with Chinese providers, who are forced to accelerate domestic alternatives amid supply constraints. This bifurcation in infrastructure strategy will likely fragment the global AI hardware ecosystem, creating distinct supply chains for Western and Chinese markets.

bmmagazine.co.uk · Aug 5

Amazon tops $3trn as AWS sales grow 36.7 per cent

Amazon’s market capitalization surpassed $3 trillion on Monday, becoming the fifth company to achieve this milestone after Apple, Microsoft, Nvidia, and Alphabet. The surge followed second-quarter earnings revealing that Amazon Web Services (AWS) grew net sales by 36.7 percent to $42.2 billion, marking its fastest expansion in over four years. Group net sales rose 20 percent to $200.6 billion, while operating income increased 43 percent to $27.5 billion. Shares closed at a record high of $284.02, up 5 percent for the day and 23 percent year-to-date. CEO Andy Jassy highlighted booming AI and chip businesses, each exceeding $25 billion in run rates. Amazon also raised its 2026 capital spending forecast to approximately $220 billion. The rally reflects strong investor confidence in the company’s cloud infrastructure and artificial intelligence capabilities, with AWS securing major contracts, including a $38 billion deal with OpenAI.

💡 Amazon’s performance dispels fears that hyperscalers are reducing AI infrastructure spending, validating the sector's continued growth trajectory. This confidence stabilizes broader market sentiment by confirming that major tech firms remain committed to heavy capital investment in artificial intelligence.