Getting your news
Attempting to reconnect
Finding the latest in Climate
Hang in there while we load your news feed
Samsung
Data center news, project activity, and monthly briefings for Samsung.
Editor's picks
-
Financial Stability Report - July 2026
The Bank of England has published its July 2026 Financial Stability Report, warning that AI-related valuations, leverage, private credit risks and frontier AI cyber threats are increasing financial stability risks.
- The report says vulnerabilities in risky asset valuations, sovereign debt markets, private credit, and equity leverage remain, while frontier AI is creating new cyber and operational resilience risks.
- It also announces proposed reforms to the UK capital framework, including making buffers more releasable and consulting on changes to the leverage ratio; the report references earlier announcements and ongoing consultations rather than a single new deal.
- The report includes a system-wide exploratory scenario on private markets, states the UK has the largest data centre pipeline in Europe, and notes that AI data centre build-out will require significant investment and financing.
-
ST 10094 2026 ADD 3
The European Commission has published an Impact Assessment accompanying a Proposal for a Regulation to strengthen Europe’s semiconductor ecosystem (Chips Act 2.0), repealing Regulation (EU) 2023/1782.
- Main measures announced:Strategic Projects (EU-level co-funded industrial deployment via a proposed European Competitiveness Fund), clarification of the First-of-a-Kind (FOAK) scope, a voluntary Business-to-Business Semiconductor Supply Chain Platform with targeted pre-crisis information requests (PM5/PM6), and demand-side measures including innovation procurement and a recommended security-of-supply declaration in public procurement (PM8–PM10). The document sets out concrete implementation elements, candidate measures PM1–PM10, and governance via the European Semiconductor Board and the Chips JU.
- Background facts and concrete figures:EUR 80 billion in announced/planned manufacturing investments attributed to the Chips Act (Pillar II); 11 FOAK projects approved by end-2025 worth over EUR 31.6 billion (public + private); target to increase EU manufacturing capacity ~30% (from 1.07 million wspm in 2023 to ~1.39 million wspm by 2030); explicit references to AI-driven demand (AI chips to represent >70% of market by 2030) and data-centre capacity forecasts (quadrupling of EU data centre capacity 2025–2030).
-
Targeted Pressure: How Chinese Manufacturing Competition Impacts US States
The Information Technology and Innovation Foundation (ITIF) has published a report finding Chinese industrial policy is reshaping global manufacturing and harming industries across every U.S. state.
- Main finding & method: The ITIF report (June 1, 2026) analyzes one “national power industry” per state using County Business Patterns employment data, HS/SITC export proxies, and global market-share series to conclude that state-backed Chinese subsidies, export pushes, and overcapacity are driving down prices and pressuring U.S. producers in sectors such as semiconductors, batteries, aircraft, and fabricated metals.
- Key facts, numbers, and timelines:China plans ~$150 billion in semiconductor investment through 2030 vs. $52 billion under the U.S. CHIPS funding; the report cites $63.3 billion Chinese semiconductor spending in H1 2025, TSMC’s $165 billion U.S. investment announcement, GE Appliances’ $490 million Appliance Park investment (2025), and state/national export shares and HS-code trade series used throughout the analyses.
-
Empowering Telecom Core Networks Through Collaboration and Innovation
Dell and Samsung have announced successful validation of Samsung’s mobile core on the Dell PowerEdge R7625 server powered by AMD EPYC processors.
- Main announcement: Dell and Samsung completed validation testing of Samsung’s mobile core running on the Dell PowerEdge R7625 (2U rackmount) using AMD EPYC processors (up to 128 cores per CPU), demonstrating performance, scalability, and reliability for disaggregated telecom core deployments.
- Background and details: The validation emphasizes cache-sensitive performance using 4th Gen AMD EPYC with 3D V-Cache™, highlights readiness for open, disaggregated core infrastructures, and includes vendor statements from Boyoung Yoon (Samsung) and Derek Dicker (AMD) confirming platform-level interoperability and operator confidence. No monetary figures or implementation timelines were provided.
-
Q4 earnings call: Remarks from our CEO
Alphabet announced Q4 earnings and CEO Sundar Pichai summarized major product and infrastructure milestones including Gemini 3 adoption and a large 2026 CapEx plan.
- Main announcement: Alphabet reported Q4 results with annual revenues exceeding $400 billion, highlighted the launch and momentum of Gemini 3, and announced 2026 CapEx guidance of $175 to $185 billion; Sundar Pichai also noted intent to acquire Intersect (data center and energy infrastructure solutions) and detailed Cloud, YouTube and Waymo metrics.
- Additional details/background:Cloud is on an annual run rate of over $70 billion with a $240 billion backlog, YouTube annual revenues surpassed $60 billion, Google reported over 325 million paid subscriptions, Gemini App has 750 million MAU, and the company reduced Gemini serving unit costs by 78% over 2025.
-
MatKV: Trading Compute for Flash Storage in LLM Inference
Kun-Woo Shin et al. (Seoul National University and Samsung Electronics) propose MatKV, a system that precomputes and materializes key-value (KV) vectors of RAG objects on flash storage to avoid repeated GPU-based KV computation at inference time.
- Main announcement: The authors introduce MatKV, which precomputes KVs, stores them on inexpensive, fast, power-efficient flash SSDs, and reuses them at inference, claiming that MatKV reduces both inference time and power consumption by half for RAG workloads; experiments use Hugging Face’s Transformers across state-of-the-art GPUs and flash SSDs. (Paper submitted 20 Dec 2025; Accepted for publication in ICDE 2026.)
- Additional details / methods: The paper focuses on making the prefill phase efficient by materializing KVs; it demonstrates two optimizations: (1) concurrent decoding and KV loading where a GPU decodes while loading materialized KVs for the next instance to reduce load latency, and (2) enabling the use of low-end GPUs for decoding once KVs are loaded into GPU memory, with reported minimal impact on throughput and QA accuracy.
Recent news
-
The rise of lithium-ion batteries – EQ
The article is an analytical commentary on the rise of lithium-ion batteries and the policy and industrial conditions behind their global expansion.
- It says the global lithium-ion battery market is worth more than USD 150 billion and explains how the technology spread from publicly funded research in the United States, Europe and Japan to commercial products in Japan, then large-scale production in Korea and China.
- It highlights key milestones including Samsung overtaking Panasonic in 2011, China producing 85% of lithium-ion batteries globally, and battery prices falling 97% by 2025 versus 2000; it also notes battery use in EVs, power grids, data centres, drones and robotics.
- The piece also discusses policy support such as China’s subsidies, land use guarantees, preferential land pricing and state-backed capital, plus raw-material strategies involving lithium, graphite and other minerals and investment ties with the Democratic Republic of Congo and Indonesia.
-
🤖 La Machine #91: Mistral’s €3 Billion Mic Drop
Mistral AI has announced a €3 billion funding round at a post-money valuation above €21 billion.
- Samsung led the financing, which Mistral says is the largest private European tech equity round on record.
- The funds are intended for frontier research, more compute, and Mistral’s own infrastructure; the article also says Mistral has nearly doubled its valuation in a year and is still positioning itself against OpenAI and Anthropic.
- The piece is a newsletter-style news summary referencing a recent announcement rather than publishing Mistral’s original press release.
- The article also mentions Mistral’s sovereignty pitch and says competitive models are coming “very soon.”
-
Mistral AI Raises €3 Billion at €21 Billion Valuation to Keep Europe in the AI Race
Mistral AI has announced a €3 billion equity round at a post-money valuation of more than €21 billion. - The round is led by Samsung Electronics, with Scaleup Europe Fund (EQT) and PSG Equity as co-leads; new investors include Advent, BlackRock-managed funds, and the Grand Duchy of Luxembourg.
- The company says the financing will expand frontier research, fund more infrastructure including its own data centers, and support larger and faster models; Microsoft reportedly did not participate despite a prior infrastructure agreement in Europe.
- The article also says the round makes Mistral the largest equity round ever completed by a privately held European technology company and that its valuation has nearly doubled from about €12 billion in a year.
-
Nvidia unveils alternative high bandwidth technology to bolster AI cards
Nvidia has announced NVHBM, a customized High Bandwidth Memory architecture for AI accelerators and custom XPUs.
- The company says NVHBM is part of its NVLink Fusion platform and is aimed at hyperscalers and AI companies building custom accelerators; Amazon’s Annapurna Labs is named as the first announced partner.
- Nvidia says the design moves the memory controller into the HBM stack’s base die, and claims up to 30% greater bandwidth, 15% lower power, and 25% more compute-die area versus conventional HBM4E.
- The article frames this as a blog-post announcement published alongside Nvidia’s quarterly earnings call, not as a report of a separate commercial deal.
-
Nvidia, MediaTek Bring Custom Chips to AI Racks
Nvidia has announced that it is opening its rack-scale AI infrastructure and NVLink Fusion technology to custom accelerators from hyperscalers and other AI developers, and it has expanded its partnership with MediaTek to support pre-validated custom XPU designs.
- The platform lets customers design custom XPUs on a pre-validated NVLink Fusion foundation and integrate them into Nvidia-connected AI factory infrastructure built around the MGX ecosystem.
- Nvidia also invested $3.5 billion in convertible bonds issued by MediaTek; the companies did not name any customers or announce specific data center deployments.
- The broader context is Nvidia’s effort to make NVLink interoperable beyond its own processors, while rival approaches include Ethernet and UALink; the MediaTek partnership also extends to local AI computing and automotive systems.
-
Our renewable future: batteries to the rescue
East Anglia Bylines has published an opinion-style article arguing that grid-scale battery storage is rapidly improving and helping enable renewable energy transition.
- The article says battery storage costs have fallen sharply, with a grid-level battery energy storage system down by roughly two-thirds since 2022 and lithium carbonate at about one-third of its 2022 price.
- It cites Marek Kubik on market changes, including Saudi Arabia reaching about 18 GWh of grid-scale storage with 24 GWh more in the pipeline, and the United Arab Emirates building a 19 GWh solar/battery project; it also discusses LFP and sodium-ion batteries and mentions possible applications for data centres.
-
Security’s rising impact on office experience in Class A buildings: Kastle CEO
Kastle has announced a strategic expansion of its security platform through executive hires and a stronger push into data centers and other high-security venues.
- The company said on July 13 that it hired five senior executives: Matt Burke (COO), Neil Thomas (CRO), Chad Pierce (CIO), Lindsay McLain (CMO), and Steven Weingarth (VP of go-to-market).
- Kastle said it is working to become the category leader in integrated, intelligence-driven security platforms and that its majority stake in i2G Systems has opened opportunities in data centers, sports venues, universities, and hospital systems.
- The article also says Kastle supports Aliro and, in March, became the first company to roll out Aliro 1.0 credentials across Apple, Google, and Samsung mobile wallets and Aliro-enabled readers.
-
Meta could start production of Iris AI chip in September – report
Meta has reportedly planned to start production of its forthcoming AI chip Iris in September, according to a Reuters report citing an internal memo.
- Iris production is slated to begin in September; the chip was one of the AI chips Meta unveiled in March and is intended to support generative AI inferencing workloads.
- The memo says Iris took six weeks to test with no major issues; Meta worked with Broadcom on design and TSMC on manufacturing, and separately signed supply agreements with Samsung Electronics, SanDisk, and Sumitomo Electric for memory, flash storage, and fiber optic equipment.
- Reuters also reported Meta plans to deploy 7GW in 2026 and 14GW by the end of 2027; Meta CFO Susan Li said capex would rise to $120 billion-$135 billion this year due in part to higher component pricing and additional data center costs.
-
South Korea AI Chip Startup Rebellions Raised $400 Million at $2.3 Billion Valuation & Plans South Korea IPO in 2027 1H, Founded in 2020 by Sunghyun Park, Oh Jin-wook, Kim Hyo-eun & Shin Sung-ho, Investors Include Samsung, SK Hynix, SK Telecom, KT Corp, Korea National Growth Fund, Saudi Aramco (Wa’ed Ventures)
South Korea AI chip startup Rebellions has announced that it raised $400 million at a $2.3 billion valuation and plans a South Korea IPO in the first half of 2027.
- Raised $400 million at a $2.3 billion valuation; the company says it plans to IPO in South Korea in 1H 2027.
- Rebellions says it was founded in 2020 by Sunghyun Park, Oh Jin-wook, Kim Hyo-eun and Shin Sung-ho and counts Samsung, SK Hynix, SK Telecom, KT Corp, Korea National Growth Fund and Saudi Aramco (Wa’ed Ventures) among its investors.
- The company describes itself as an AI accelerator startup providing cards, servers and rack-scale systems for datacenters, with software support for PyTorch 2.x and vLLM.
- It says it has begun mass production of 5nm accelerators deployed across thousands of ATOM cards powering commercial AI services for telecom operators.
-
SK Hynix looking to raise $28bn with US IPO
SK Hynix has announced plans to sell 17.8 million shares in a US IPO, with the company expected to raise around $28 billion.
- In an SEC filing dated July 6, SK Hynix said it will offer American depositary receipts (ADRs), with 1 ADR = 0.1 common share; pricing is expected on Thursday, July 9, with trading to begin on Friday, July 10.
- The company said it is one of the world’s largest memory semiconductor companies, ranking second in DRAM and second in NAND flash by revenue; it also reported Q1 2026 revenue of $34.51 billion.
- The article also references a separate earlier announcement that SK Hynix, alongside Samsung, plans to invest $513 billion to build four memory fabs in South Korea and 81 trillion won ($52 billion) for HBM packaging fabs in the Chungcheong region, plus 5 GW of data center capacity and a possible additional 10 GW between 2029 and 2035.
-
Why India’s AI Boom Is Running On A Waiting List
The article analyzes how AI chip shortages and supply-chain constraints are reshaping how India’s cloud providers and startups source and use compute, rather than announcing a new deal.
- GPU delivery timelines have shifted from weeks to months, with large deployments historically taking 6-15 months in 2024 and dedicated cluster setups now taking about 4 months; lead times for next-generation enterprise AI GPUs are said to be 36-52 weeks.
- Supply bottlenecks extend beyond chips to CoWoS packaging, HBM memory, and networking components; the article also cites 2027-2028 as the expected period when new memory factory capacity from SK Hynix, Samsung, and Micron may materially improve supply.
- The story highlights how firms such as Murf.AI, Nurix, and CoRover are reserving capacity in advance, using mixed hardware fleets, and optimizing workloads to cope with scarcity, while IndiaAI-related sourcing is mentioned as part of the broader compute ecosystem.
-
Chip industry group warns US government against interventionist approach to memory chip shortage
SEMI has written to senior officials in the Trump administration urging the US government not to try to influence memory chip prices or production capacity.
- In a July 1 letter seen by Bloomberg News, SEMI said interventions that distort pricing or capacity decisions could prolong the demand downturn.
- The letter was addressed to Treasury Secretary Scott Bessent, Defense Secretary Pete Hegseth, Commerce Secretary Howard Lutnick, and Secretary of State Marco Rubio; SEMI also said tax breaks should be extended to encourage US production capacity increases.
-
South Korean Tech Giants to Build a $518 Billion Chipmaking Hub to Serve Soaring AI Demand
Samsung Electronics and SK Hynix announced a combined investment to build a new semiconductor hub in southwest South Korea.
- Main announcement: Samsung Electronics and SK Hynix will invest 800 trillion won (~$518 billion) to build a new chipmaking hub in southwest South Korea, with each company building two fabrication plants; Samsung said its new fabs will be built in Gwangju and potential sites include grounds of a military air base slated for relocation. The companies and President Lee Jae Myung publicly announced the plan together.
- Background and details: Government officials and company chairs said the project requires vast sites, sufficient power and water, and skilled workers; no completion timeline was specified. Officials highlighted the region’s renewable energy strengths and outlined a plan for a nationwide semiconductor ecosystem (southeast expanding components, Chungcheong for packaging, and data centers built across the country).
-
Singapore-based cloud service OrtCloud raises $1.7M pre-seed funding to advance AI-focused cloud infrastructure
OrtCloud has raised $1.7 million in a pre-seed funding round led by Golden Gate Ventures.
- Funding and purpose: The pre-seed round raised $1.7 million, led by Golden Gate Ventures with participation from Antler, to support development of OrtCloud’s deterministic virtual machine infrastructure for fixed workloads and AI-agent environments, and to scale product development, infrastructure, hiring, and go-to-market in the Asia-Pacific and U.S. markets.
- Product and market context: OrtCloud provisions fixed-resource virtual machine tiers (hosted cloud and on-premises options) targeting enterprises with data residency or network isolation needs; the article cites a $50 billion to over $80 billion Southeast Asia infrastructure opportunity and an agent-based workloads category worth over $20 billion.