45k user reviews & 1 million votes, over 29k web hosting brands & 85k plans since 2004!
X
Menu
WHTop → 📋 News → Groq Raises $650 Million for AI Inference Cloud
Jun, 2026 : Groq Raises $650 Million for AI Inference Cloud
📅 - Groq has secured $650 million in new funding as the AI infrastructure market increasingly pivots from model training toward inference, the stage where AI systems generate responses at scale. The company plans to use the capital to expand its global cloud footprint, deploy new NVIDIA-linked systems, and compete for a growing share of enterprise AI workloads worldwide.
The financing arrives at a moment when investors are recalibrating how they think about artificial intelligence infrastructure.
For the past several years, the industry's attention has largely centered on training. Bigger models. Larger GPU clusters. Ever-growing capital expenditures. Yet many infrastructure operators [...][... Check source for end of article ...]
📅 - U.S. Quantum Coalition Targets Future Space Infrastructure - A new U.S.-led coalition bringing together quantum computing specialists, aerospace companies, researchers, and investors is attempting to tackle a problem that has lingered for years: quantum technology has generated impressive demonstrations, but far fewer operational systems.
The newly launched America's Quantum Space Initiative aims to accelerate deployment of quantum technologies in future space infrastructure, defense platforms, navigation systems, communications networks, and exploration missions across government and commercial markets.
For all the excitement surrounding quantum computing, many of the most immediate opportunities in space may have little to do with processing [...]
📅 - MemryX Extends Cascade Platform to Edge AI Servers - MemryX is extending its Cascade platform into edge server infrastructure, adding a PCIe accelerator built on its MX3 architecture as AI inference moves closer to operational sites and developers confront a less glamorous problem than model design: deploying compute efficiently in constrained environments without blowing up power budgets, memory bandwidth, software consistency, or server density economics at scale.
The new server-focused product is the MemryX Cascade 100P PCIe Accelerator, a single-slot, passively cooled card using a 16-chip array of MX3 processors. It is aimed at high-density edge server deployments where operators want local AI inference without treating every new [...]
📅 - Upscale AI Raises $190M for AI Data Center Networking - Upscale AI has raised another $190 million, taking total funding to $500 million and its valuation to $2 billion, as investors chase the networking layer behind AI data centers - the part that decides whether expensive accelerators can actually work together across massive clusters, mixed suppliers, memory pools, and storage systems without choking on traffic today either.
The company wants to become the next Cisco for AI infrastructure. Ambitious. Also convenient shorthand for a market trying to work out whether the next infrastructure control point sits in chips, data center power, cloud orchestration, or the fabric connecting all of it.
Premji Invest led the Series A-1 round. New [...]
📅 - Schneider, HPE Lay Hybrid Cloud Groundwork for Software-Defined Automation - Schneider Electric and HPE are packaging industrial automation modernization as a managed service, combining Schneider's EcoStruxure Automation Expert with HPE SimpliVity infrastructure for factories, utilities, and other operators trying to update aging control systems without shutting production down. The offer targets a stubborn industrial problem - how to modernize operational technology while keeping plants running safely today.
Industrial automation has become one of the less glamorous bottlenecks in digital transformation. Everyone talks about AI in factories, autonomous plants, predictive maintenance, and software-defined operations. Then the conversation hits the control layer. [...]
📅 - Backblaze to Deliver $335M Storage Capacity to AI Firm CoreWeave - Backblaze has signed a five-year, $335 million multi-exabyte storage agreement with CoreWeave, adding lower-cost HDD-based capacity to parts of the AI cloud provider's managed storage infrastructure. The arrangement gives CoreWeave customers new object storage tiers without code changes, while underscoring a less glamorous constraint in AI infrastructure: where all the data actually sits during production workloads too.
For CoreWeave, the deal is not about headline GPU capacity. It is about the storage architecture surrounding that capacity. AI clouds are often discussed as if compute is the whole machine. It is not. Training runs, checkpoints, model outputs, retrieval-augmented [...]
📅 - Telefónica Germany Builds Private Cloud for Telco Workloads - Telefónica Germany is expanding its relationship with Tech Mahindra to build a private cloud platform for telecom and enterprise workloads, underscoring a broader shift underway across the telecom industry. Operators increasingly want cloud-native flexibility and automation, but many remain reluctant to hand critical network infrastructure entirely to public cloud providers amid sovereignty, control, and resilience concerns today.
For years, telecom operators were told they should behave more like cloud companies. Virtualize network functions. Automate operations. Abstract hardware. Build software-driven infrastructure capable of launching services faster and operating more [...]