Databricks secured strategic funding at a $188 billion valuation, signing a financing term sheet led by existing investor Coatue with additional new and backers set to join; the round is expected to close late this summer. Unrevealed fresh capital will accelerate its AI roadmap: scaling Unity AI Gateway (multi-AI governance), Genie (AI coworker converting business data into reliable actionable insights), and Lakebase (serverless Postgres built for AI agents), alongside funding future AI acquisitions and deeper AI research.
“Enterprises shift from tokenmaxxing to valuemaxxing. They avoid wasting costly tokens on top-tier models for every workload and prioritize optimal outcomes per dollar, requiring flexible selection of task-fit AI,” said Ali Ghodsi, Databricks Co-founder and CEO. “This funding lets us advance our multi-AI strategy to meet surging customer demand, strengthening Unity AI Gateway, expanding Genie and refining Lakebase.”
Databricks released a blog titled Benchmarking Coding Agents on Databricks’ Multi-Million Line Codebase, evaluating diverse models and coding frameworks on real engineering tasks. Testing proved combined OpenAI, Anthropic and open-source stacks deliver peak performance while challenging common assumptions around cost, efficiency and model selection; all tested tools fall into three distinct performance tiers, with full details available in the blog.
The firm also named Lee Stockwell Head of Public Sector for the UK and Ireland.
Datadog expanded UK local data residency support by launching its full product suite on AWS London Region, a critical upgrade for regulated verticals including finance, healthcare, government and higher education that mandate in-country data storage and operational stability.
DDN appointed Michelle Rosen Chief Legal Officer amid its next phase of global AI infrastructure expansion covering enterprise AI, sovereign AI, cloud vendors and hyperscale deployments. Over six months, DDN widened its global footprint, hit one million deployed GPUs, launched major upgrades to its AI data intelligence platform, and records over 300% year-over-year growth as businesses ramp AI infrastructure investment.
Per Storage Review, Supermicro launched Kubernetes Edge AI Appliances co-developed with Red Hat and Everpure. These pre-integrated systems combine Supermicro edge servers, Red Hat OpenShift and Everpure’s Portworx Kubernetes storage platform to run AI inference, containers and VMs across distributed sites, delivering validated, pre-loaded deployments for standardized off-data-center AI infrastructure.
Portworx delivers Kubernetes-native storage without dedicated hardware arrays per location, aggregating local edge server storage into a software-defined layer. Supermicro’s design enables autonomous offline operation with high availability, local data protection and unified storage policies, ideal for repeatable rollouts and centralized management across hundreds to thousands of edge sites.
ExaGrid CEO Bill Andrews told Storage Newsletter its all-flash backup appliances earned positive initial feedback, yet industry-wide SSD price surges made flash storage unaffordable for backup use cases — a hurdle shared by competing all-flash backup hardware vendors.
Liqid, provider of software-defined memory and GPU pooling infrastructure, hired Daniel Berg as Chief Product Officer. Berg brings 30+ years product and engineering leadership experience from Gorilla Logic, JumpCloud, NetApp, Skype and Sun Microsystems, and will oversee end-to-end product management, strategy and engineering for Liqid Matrix and its disaggregated memory/GPU pooling stack.
Data integrity vendor Precisely appointed Vinesh Vis CRO and Jane Gilson CBO, both reporting directly to CEO Walid Abu-Hadba. Vis joins from Smarsh as former CRO and GM, with prior revenue leadership at Oracle and other high-growth tech firms. Gilson will manage Precisely’s global commercial arm covering channels, professional/strategic services, revenue operations, deal governance and go-to-market strategy, bringing two decades of global tech growth experience with executive roles at Microsoft, Google and CloudSphere CEO.
Citrini analyst Jukan reported Samsung Electronics held preliminary bank discussions for a potential ADR offering, following rival SK Hynix’s prior move.
A social media post cited a Goldman Sachs analysis covering Sandisk’s new New Business Model (NBM) contracts. The bank’s revised forecasts include five binding NBM deals: three signed in FQ3 generating ~$42 billion RPO, two more early FQ4 agreements with undisclosed RPO figures. Collectively, the five deals lock in roughly $62 billion minimum multi-year revenue, backed by over $11 billion in prepayments and financial guarantees to vastly improve Sandisk’s revenue predictability.
Per Wcctech, SK Hynix CEO Kwak Noh-jung warned the memory industry will face its worst-ever supply shortage in 2027, with unmet market demand persisting past 2030. He shared the outlook during SK Hynix’s Nasdaq listing ceremony in Times Square.
TrendForce’s latest memory pricing survey found manufacturers shift mature NAND fab capacity to high-layer 3D NAND, triggering severe MLC NAND shortages. Industrial, automotive and networking buyers are switching to SLC NAND, further tightening limited supply. SLC NAND demand surges for AI edge, datacenter and automotive electronics, rapidly widening supply-demand gaps. TrendForce forecasts SLC NAND contract prices will jump 120–170% H2 2026 versus H1, with potential larger hikes ahead.
WEKA confirmed CoreWeave (Neocloud parent) is a long-term customer: WEKA client Cohere runs NeuralMesh Axon on CoreWeave’s infrastructure, and WEKA completed its first large-scale Augmented Memory Grid production benchmarks on CoreWeave’s cloud, with supporting details in a dedicated blog post.
ZeroPoint Technologies debuted ZeroStream, novel hardware compression IP that boosts effective memory bandwidth to maximize throughput on existing memory hardware. Core capabilities: compatible with diverse workloads to raise token throughput from AI accelerators; fully lossless compression with zero quantization or model accuracy loss; up to 1.5x compression for LLM weights, average 2x compression for activations and KV cache; 20–35% higher effective bandwidth, peaking at 50% for select AI workloads.
Beijing Qianxing Jietong Technology Co., Ltd.
Sandy Yang/Global Strategy Director
WhatsApp / WeChat: +86 13426366826
Email: yangyd@qianxingdata.com
Website: www.qianxingdata.com/www.storagesserver.com
Business Focus:
ICT Product Distribution/System Integration & Services/Infrastructure Solutions
With 20+ years of IT distribution experience, we partner with leading global brands to deliver reliable products and professional services.
“Using Technology to Build an Intelligent World”Your Trusted ICT Product Service Provider!
Il tuo messaggio deve contenere da 20 a 3000 caratteri!
Italian
