News related to
Nvidia

The performance figures emerged at Nvidia AI Day Seoul 2026, held on November 9 and 10, where both South Korean chipmakers presented applications of accelerated computing and AI to semiconductor development. Samsung’s fastest configuration delivered an approximately 3,800-fold improvement over a conventional processor-based calculation, although the comparison involved different computational methods.

Samsung demonstrated a thermal and mechanical simulation designed to predict how semiconductor structures respond to heat and physical stress. The original calculation required approximately 18 hours on a 32-core central processing unit using a detailed computational grid. An AI-assisted calculation finished in 41 seconds on one Nvidia H100 graphics processor and 17 seconds using four processors.

The single-GPU configuration produced an approximately 1,600-fold speed improvement, the figure highlighted in Nvidia’s official announcement. The faster four-GPU result illustrates additional acceleration through parallel processing, rather than an identical test conducted with unchanged hardware and numerical methods.

Samsung’s approach combines Nvidia’s PhysicsNeMo artificial intelligence framework with a technique known as AI homogenisation. This replaces highly detailed representations of complicated semiconductor structures with simplified computational models that preserve important physical properties.

The companies demonstrated how a structure containing approximately one billion fine-scale computational elements could be represented using roughly one million larger cells. AI models estimate effective material properties, allowing engineers to calculate thermal behaviour and mechanical deformation without repeatedly processing every microscopic feature.

Samsung reported 99.17% accuracy in effective vertical thermal conductivity for the demonstrated workflow. That measurement applies to the specified physical property and test conditions, rather than establishing equivalent accuracy across every aspect of semiconductor simulation.

The development addresses a significant engineering challenge as advanced semiconductor packaging introduces increasingly complicated arrangements of materials and components. Temperature variations and mechanical stress can affect chip reliability, making accurate simulations important before designs enter manufacturing.

Samsung’s Computational Science and Engineering team worked with Nvidia to develop an integrated scientific machine-learning pipeline. The process generates synthetic semiconductor layouts, produces simulation data through GPU-accelerated calculations and trains physics-aware models to estimate material behaviour.

SK hynix separately outlined its TCAD 3.0 programme, which integrates AI with technology computer-aided design software used to model semiconductor manufacturing processes and device characteristics before physical production.

Kim Jung-han, head of the company’s TCAD Intelligence team, presented an AI model developed with Nvidia PhysicsNeMo to predict changes in material profiles during etching. The system produces results comparable to conventional simulations at speeds described as several dozen times faster.

Etching involves removing selected materials from semiconductor wafers through tightly controlled manufacturing steps. Individual processes can involve numerous variables, including temperature, pressure and gas flow, making comprehensive simulation computationally demanding.

SK hynix also demonstrated chip-scale thermal and mechanical calculations running more than 30 times faster than a conventional 32-thread CPU configuration. Its approach distributes calculations across multiple graphics processors while reusing information obtained from smaller computational regions.

The company said performance increased almost proportionately as additional GPUs were introduced, suggesting that larger computing installations could handle increasingly demanding engineering workloads.

The two manufacturers’ performance figures reflect different simulations, hardware configurations and measurement conditions. Neither demonstration establishes that complete semiconductor development cycles can achieve equivalent reductions in duration.

Samsung is extending its Nvidia collaboration into semiconductor process modelling, optical simulation and materials research. Its sMOL research platform has achieved up to fivefold acceleration in selected quantum chemical calculations using Nvidia’s cuEST software.

The company is also evaluating Nvidia’s cuISS and cuDSS libraries for computationally intensive mathematical operations involved in semiconductor device modelling. These tools address different engineering workloads from the thermal simulation demonstrated in Seoul.

SK hynix’s broader collaboration with Nvidia includes accelerating existing simulation software and developing AI-assisted manufacturing models. The companies have also agreed to cooperate on advanced memory technologies and digital representations of semiconductor fabrication facilities.

Nvidia’s PhysicsNeMo framework supports machine-learning models that incorporate physical principles and simulation data. Its Warp software provides GPU-accelerated computing capabilities used to generate training information for engineering applications.

Samsung’s demonstrated pipeline uses these technologies together, with the trained models supplying simplified material properties for subsequent large-scale calculations. The companies are exploring software agents capable of coordinating simulation tasks and evaluating alternative engineering designs.

Samsung has made an example of its physics-based modelling work available through Nvidia’s PhysicsNeMo resources, alongside research covering semiconductor simulation techniques presented at the 2026 Design Automation Conference.

The Nikkei 225 closed at 58,775.50, gaining 1,340.73 points, while the broader Topix advanced 2.97% to 3,818.16. The rebound followed Friday’s 6% plunge, which left the Nikkei more than 20% below its June record, the threshold commonly used to define a bear market.

Buying extended across the Tokyo market rather than remaining confined to semiconductor companies. Of the Nikkei’s 225 constituents, 208 rose and 17 fell, indicating widespread demand after several sessions of selling. The recovery nevertheless left the benchmark substantially below its level at the beginning of last week.

Investors took encouragement from Friday’s gains on Wall Street, where technology stocks recovered after sharp declines linked to concerns about potential changes in United States policy towards artificial intelligence. The Nasdaq Composite gained 1.09% and the S&P 500 rose 0.32%, providing a firmer backdrop for Asian trading.

Nvidia and Tesla had recovered in that American session after investors reassessed the implications of Democratic victories in the US midterm elections. Market participants have questioned whether tighter scrutiny of artificial intelligence could slow the construction of data centres and affect demand for the advanced chips and equipment supplying them.

Those concerns remain particularly significant for Tokyo’s equity market, where semiconductor equipment manufacturers, technology investors and industrial suppliers carry substantial weight in benchmark movements. The sell-off had spread beyond companies directly involved in artificial intelligence, reflecting broader reductions in exposure to riskier assets.

Takuma Ikemoto, a market analyst at Tokai Tokyo Intelligence Laboratory, cautioned that uncertainty surrounding American policy could limit further advances once the initial wave of buying subsided. His assessment highlighted the distinction between a recovery from heavily sold conditions and a sustained improvement in expectations for technology earnings.

Individual stock movements showed that the rally was uneven despite the strong headline index gains. Consulting services company Baycurrent climbed 8.8% to a record closing high, making it the Nikkei’s strongest percentage performer. Construction group Shimizu gained 7.7%, while industrial manufacturer Mitsubishi Electric advanced 7.3%.

SoftBank Group, a prominent investor in artificial intelligence ventures, remained under pressure during the session, illustrating continuing investor caution towards businesses closely associated with the sector’s expansion. Its weakness contrasted with gains elsewhere and suggested that buyers were distinguishing between companies rather than returning indiscriminately to all technology-related shares.

Banking stocks also attracted demand as Japanese government bond yields moved higher. Rising yields can support banks’ lending margins, although their effect depends on funding costs and broader credit conditions. Automakers, retailers and non-ferrous metal producers contributed to the market’s advance, giving the recovery a wider base than the previous technology-led fluctuations.

The Topix’s larger percentage gain reflected the strength of shares across a broader range of companies. On the Tokyo Stock Exchange’s Prime Market, 1,412 issues advanced, compared with 120 declines and 13 unchanged. Thirty-one of the exchange’s 33 industry groups finished higher, with food and shipping among the exceptions.

Developments in the Middle East also shaped investors’ assessment of market risk. US President Donald Trump said on Sunday that further American attacks on Iran were highly unlikely while diplomatic discussions continued, with Qatar involved in mediation. The comments offered some reassurance, although accounts of attacks on vessels near the Strait of Hormuz underscored the fragility of the situation.

Disruption to energy shipments through the waterway has complicated the outlook for Japan, which depends heavily on imported fuel. Changes in oil prices influence corporate costs, household spending and inflation expectations, making the security of Gulf shipping an important consideration for investors alongside movements in technology shares.

Monday’s closing figures also showed the distance still to be recovered after the previous week’s losses. The Nikkei had finished November 2 at 63,081.86, well above Monday’s level despite the sizeable daily increase. The gap illustrated how a sharp one-day advance can coexist with substantial losses over a longer trading period. Investors therefore remained sensitive to changes in overseas equity markets, policy expectations and corporate earnings forecasts.

The session’s strongest performers included businesses outside the artificial intelligence supply chain, while the weaker areas showed that the rebound was not uniform. Shipping and food shares lagged the broader market as investors repositioned portfolios following the previous week’s abrupt decline.

Building on the firm’s deep-rooted expertise in semiconductors and deep technology, Ondine Capital Fund III will invest across the full AI value chain—spanning AI foundation models, next-generation semiconductors, AI infrastructure, AI-native software, and physical AI applications.

Ondine Capital Fund III’s early portfolio highlights include:

• Recursive Superintelligence: Founded by former Salesforce Chief Scientist Richard Socher and built by veteran researchers from OpenAI, Google, and Meta, the startup focuses on recursive self-improving AI systems. It recently raised $650 million at a $4 billion pre-money valuation led by Google Ventures, with participation from Nvidia and AMD.
• Bot Auto: Founded by Xiaodi Hou (former co-founder and CTO of TuSimple), this Houston-based physical AI startup specializes in autonomous freight transportation. Bot Auto recently partnered with Ryan Transportation to launch driverless freight operations along the Houston-Dallas corridor.
• Salience Labs: Founded by Vaysh Kewada and Johannes Feldmann, the company develops silicon-photonics optical switching technology for AI data centers. Partnering with Tower Semiconductor, Salience Labs recently closed a $125 million financing round led by Bullhound Capital, with strategic backing from AMD Ventures and Intel Capital.

Partner Appointment to Boost Global Ecosystem

Coinciding with the first close, Ondine Capital announced the appointment of Brian Yen as Partner, reinforcing the firm’s global sourcing and founder network.

Prior to joining Ondine Capital, Brian Yen served as Principal and Head of the Taipei Office at Headline Asia, where he cultivated an extensive network across Japan, Taiwan, and Southeast Asia. His investment track record spans consumer technologies, agentic services, and physical AI, with notable investments including respond.io, forest inc., Braid, and iMotorbike. At Ondine Capital, Brian Yen will focus on backing early-stage founders building intelligent software and physical AI platforms.

“The convergence of AI and semiconductors is unleashing unprecedented opportunities across global tech,” said Brian Yen. “As these technologies mature, we anticipate monumental value creation throughout the entire stack. Through Fund III, we look forward to empowering visionary founders who are shaping the future of computing and industry applications.”

Ondine Capital continues to actively deploy capital and evaluate global investment opportunities in breakthrough, category-defining companies across the AI and semiconductor landscape.

Hashtag: #OndineCapital #RSI #BotAuto #SalienceLabs

https://www.ondinecap.com/

The issuer is solely responsible for the content of this announcement.

About Ondine Capital

Ondine Capital is a premier venture capital firm focused on frontier technologies, including semiconductors, artificial intelligence, enterprise software, clean energy, and emerging technology platforms. The firm backs visionary founders from inception, providing strategic support and access to a robust global network of industrial partners, investors, and entrepreneurs.

Solcis extends open security guardrails such as NVIDIA OpenShell and Microsoft’s AGT with the controls compliance professionals need, in a platform built to be self-hosted.

SINGAPORE – Media OutReach Newswire – 21 October 2026 – Solcis, a spin-out of 6clicks, today released Solcis Compliance Studio and the Solcis Policy Engine. Together they let organisations turn their policies and compliance obligations into enforceable rules for AI, check each action an AI agent takes against those rules as it happens, and keep a tamper-evident record for auditors and regulators. Solcis integrates with open runtime enforcement engines such as NVIDIA OpenShell and Microsoft’s Agent Governance Toolkit (AGT).

As organisations put AI agents to work, they must show that every automated action meets the same obligations as the rest of the business. Today, that is hard to prove. Solcis closes that gap across cybersecurity, privacy, financial crime, conduct and prudential regulation.

Solcis Compliance Studio
Solcis Compliance Studio is where compliance teams define, run and prove compliance for their AI agents. Studio has six core capabilities:

• Policy lifecycle and authoring: turns obligations into enforcement policies and manages them through deployment and updates.
• Universal evidence ledger: brings results from every enforcement point into one tamper-evident record.
• Compliance model management: recommends the best small language model (SLM) for each policy, benchmarked for compliance.
• Learning and drift detection: improves policies over time from observed decisions and human judgement, and detects control and agent drift.
• Compliance coverage and analytics: audit-ready reporting on compliance posture, including agent discovery.
• Integrations: connects to GRC, SIEM and ticketing systems, including 6clicks, ServiceNow IRM and Archer.

Solcis Policy Engine
Organisations can enforce Studio policies through NVIDIA OpenShell, Microsoft’s AGT, or the Solcis Policy Engine, Solcis’s own open enforcement engine for general compliance and business rules. The Policy Engine reaches decisions in under 200 milliseconds by combining two types of check:

• Deterministic checks apply fixed rules that return the same result for the same input every time, giving regulators and auditors repeatable, evidence-backed answers.
• Inference-based checks use AI to judge context and ambiguous cases that fixed rules cannot capture. Checks are routed to the best-suited AI small language model (SLM).

Every agent action in its path receives an Allow, Deny or Escalate decision with a severity rating, inline and without slowing it down. Decisions are enforced through the Solcis SDK in an AI system’s action path, or the Solcis Gateway on model, tool and MCP endpoints.

“Most organisations know what the regulation says and what their systems do, but not how the two connect,” said Anthony Stevens, CEO and co-founder of Solcis. “The Solcis platform joins them up, from obligation to policy to enforcement to evidence. And for the organisations that need it most, keeping that chain on their own infrastructure isn’t a preference, it’s non-negotiable.”

Sovereign by design, available now
Solcis is built only to be self-hosted, so organisations keep full control of their AI compliance workloads, models and data: nothing Solcis processes, from policies to decision logs, ever leaves their environment.

Both products are available now, directly from Solcis and through partners, and from November 2026 on the Microsoft, AWS, Google Cloud, Oracle, NVIDIA and Red Hat marketplaces.

Compliance Studio is sold on a subscription or perpetual licence, with no usage metering. The Policy Engine is released under the Functional Source License, so its code is open to audit.

Learn more at solcis.ai.

Solcis, Solcis Compliance Studio and Solcis Policy Engine are trademarks of Solcis Pty Ltd.

NVIDIA and NVIDIA OpenShell are trademarks of NVIDIA Corporation in the U.S. and other countries. Microsoft is a trademark of the Microsoft group of companies. All other trademarks are the property of their respective owners.

Hashtag: #Solcis #AICompliance #AgenticAI #AgentCompliance

https://solcis.ai/
https://www.linkedin.com/company/solcis-ai-compliance

The issuer is solely responsible for the content of this announcement.

Solcis

Solcis builds sovereign AI compliance software that organisations self-host and control, with a mission to serve the auditors and regulators who hold them to account. Spun out of 6clicks, Solcis offers two products: Solcis Compliance Studio, which integrates with publicly available AI runtimes such as NVIDIA OpenShell and Microsoft’s Agent Governance Toolkit; and the Solcis Policy Engine, its own engine for general compliance and business rules.

The Nifty IT index climbed 11.7% during the week ended Friday, October 16, its biggest weekly advance since 2008. Meanwhile, an MSCI index tracking global semiconductor companies had fallen approximately 7.4% during the week, putting software exporters on course for their widest relative outperformance since 2004.

The divergence reflects a significant change in investor sentiment towards technology businesses. Semiconductor manufacturers, previously among the biggest beneficiaries of enthusiasm surrounding artificial intelligence, have faced selling pressure amid concerns about infrastructure spending, financing costs and the sustainability of earnings growth.

Software services companies, meanwhile, have attracted renewed interest following quarterly results that suggested corporate technology expenditure was stabilising. Investors have also responded to comparatively modest valuations after a prolonged period of weakness in the sector.

HCL Technologies helped initiate the recovery after reporting stronger quarterly performance and raising its full-year revenue growth guidance. Its shares surged 8.4% on Tuesday, their biggest single-session advance in more than six years, while the broader technology index gained 5.2%.

The company reported quarterly net profit of approximately Rs51.32 billion, representing growth of 10.9%. Revenue increased 2.3% sequentially to Rs356.2 billion, supported partly by demand for engineering and research services.

HCL Technologies also disclosed advanced artificial intelligence revenue of $219 million, indicating that established outsourcing companies are beginning to generate identifiable income from technology that investors had previously viewed principally as a competitive threat.

Tech Mahindra reinforced the positive sentiment with quarterly revenue growth of 2.4%, reaching approximately Rs161.8 billion. Net profit rose 8.3% to Rs16.1 billion, supported by improvements in its communications business.

Wipro shares also advanced after the company reported quarterly revenue of Rs247.53 billion, exceeding market expectations. However, its deal bookings declined 45% from a year earlier to $2.6 billion, highlighting continued uncertainty surrounding demand for outsourcing contracts.

Wipro projected constant-currency revenue growth ranging from a decline of 1.5% to an increase of 0.5% for the following quarter. Management indicated that two delayed large contracts could be finalised, although their completion remains uncertain.

The contrasting figures illustrate why analysts remain cautious about treating the market recovery as evidence of a sustained improvement across the entire technology services industry.

Investors have nevertheless become increasingly sensitive to the valuation differences between software exporters and semiconductor manufacturers. Shares linked to artificial intelligence infrastructure had commanded substantial premiums based on expectations of expanding demand for processors, memory and computing capacity.

Those assumptions have come under pressure as investors question whether technology companies can generate sufficient returns from enormous commitments to data centres and related infrastructure.

Higher bond yields have compounded those concerns by increasing financing costs for capital-intensive businesses. Inflation pressures in the United States have also revived expectations that monetary policy could remain restrictive, reducing enthusiasm for highly valued growth stocks.

Semiconductor companies including Nvidia, Broadcom and Advanced Micro Devices have faced broader selling pressure, alongside memory manufacturers and businesses supplying artificial intelligence infrastructure.

However, the reversal does not necessarily indicate deteriorating demand for computing hardware. Analysts distinguish between concerns about share valuations and the underlying commercial requirements for advanced processors.

Software exporters face different challenges. Generative artificial intelligence could automate programming, testing and customer support functions, potentially reducing the labour-intensive contracts underpinning traditional outsourcing revenues.

At the same time, businesses adopting artificial intelligence require integration, cybersecurity, cloud management and technology consulting services, creating opportunities for established providers.

The balance between these competing forces remains important for revenue forecasts and profitability expectations.

Friday’s trading demonstrated the strength of the rotation. The Nifty IT index rose 3.9% to close around 31,924 points, while the broader Nifty 50 gained approximately 0.7% to 23,146.

The Sensex finished at 74,729, contributing to the benchmark indices’ strongest weekly performance in six months.

Despite the rebound, technology shares remain below their earlier peaks. Wipro had lost nearly 32% during the year before Friday’s close, compared with a decline of approximately 17% for the broader technology index before the week’s recovery.

Su held talks with Young Hyun Jun, Samsung’s vice chairman and chief executive of its Device Solutions division, during a visit to South Korea that also included discussions with SK Hynix chief executive Kwak Noh-jung. The meetings placed high-bandwidth memory, broader DRAM supplies and possible foundry co-operation at the centre of AMD’s efforts to secure capacity for its expanding AI business.

Speaking to reporters in Seoul, Su said AMD continued to explore opportunities with Samsung across both memory and contract chip manufacturing. She described Samsung and SK Hynix as critical partners for AMD’s data-centre ambitions and urged memory producers to add capacity as quickly as possible.

The talks come as shortages across parts of the memory market complicate expansion plans for chip designers and data-centre operators. Demand for high-bandwidth memory, or HBM, has surged because AI accelerators require substantially greater memory bandwidth than conventional computing processors, while strong demand has also tightened supplies of standard DRAM products.

Su said the supply environment remained tight and AMD was working with customers to optimise their memory requirements so available capacity could support more computing. She said the company had built a strong supply chain but was working with suppliers to accelerate its expansion.

AMD expects demand for AI computing to remain elevated for several years and plans to increase chip supply substantially in 2027. Su said during a stop in Taiwan this week that the company had extended its capacity-planning horizon from one or two years to between three and five years as it works with manufacturing partners.

That planning encompasses advanced wafers, packaging, substrates and memory, all of which have become potential constraints as companies race to build AI infrastructure. AMD has been expanding its Instinct accelerator business as it competes with Nvidia for spending by cloud providers and other large data-centre customers.

Samsung already occupies an important position in AMD’s supply plans. The companies signed a memorandum of understanding in March covering next-generation AI memory and computing technologies, with Samsung designated a primary HBM4 supplier for AMD’s Instinct MI455X graphics processor.

Their agreement also covers advanced DRAM products for AMD’s sixth-generation EPYC processors, code-named Venice, and its Helios rack-scale AI platform. Samsung has supplied HBM3E memory for AMD’s MI350X and MI355X accelerators, extending a technology relationship between the companies that spans graphics, mobile and computing products.

HBM4 is particularly important to the next generation of AI hardware because increasingly powerful accelerators need faster transfers of enormous datasets while controlling energy consumption. Memory manufacturers have consequently directed more investment towards HBM, leaving capacity across other memory categories under pressure.

Samsung has been seeking to increase its position in the high-margin HBM market while SK Hynix remains another major supplier to AI chip developers. Strong AI demand has supported memory prices and earnings, though the pace of conventional DRAM price increases is expected to moderate after steep gains earlier this year.

AMD’s discussions with Samsung could also extend beyond memory. Their March agreement provided for talks on a possible foundry partnership under which Samsung could manufacture future AMD products. Su said on Wednesday that AMD was continuing to examine opportunities in that area, without announcing a manufacturing contract.

The possibility would give AMD another potential production option as advanced semiconductor capacity remains heavily sought after. Taiwan Semiconductor Manufacturing Co remains a central manufacturing partner for AMD, and Su met suppliers in Taiwan before travelling to South Korea as part of her capacity discussions.

The Hangzhou-based AI developer released Ascend versions of several key components on September 30, including its DeepGEMM matrix-multiplication library and DeepEP communications software. The projects are designed to work with Huawei’s Ascend processors while keeping public interfaces aligned with versions built around Nvidia hardware, reducing the amount of code developers may need to rewrite when moving workloads.

That portability targets a critical part of Nvidia’s competitive advantage. Nvidia’s CUDA software platform has become deeply embedded in AI development, giving programmers mature libraries, tools and optimisation techniques built over years. Competing chips therefore have to overcome not only differences in raw computing performance but also the cost and complexity of shifting software away from CUDA.

DeepSeek’s DeepJIT project takes a further step by providing a shared runtime interface for Nvidia CUDA graphics processors and Huawei Ascend neural-processing units. It handles compilation, caching, loading and launching of specialised AI kernels while retaining backend-specific code where necessary. The approach does not make the two architectures interchangeable, but it can reduce friction for developers maintaining software for both.

DeepGEMM-Ascend is fully API-compatible with DeepSeek’s original DeepGEMM package and supports BF16, FP8 and FP4 calculations used in demanding AI workloads. DeepSeek says the library uses Ascend-specific optimisation methods to approach the hardware’s performance limits while shielding programmers from lower-level complexities including memory layouts, alignment constraints and address calculations.

Its DeepEP-Ascend library similarly aligns its public buffer interfaces with the Nvidia version while using Huawei’s communications technologies underneath. DeepSeek has also added Ascend support to DeepSelect, a specialised component used for sparse-attention and sampling operations, and released Ascend sparse-attention kernels for its model infrastructure.

Huawei, meanwhile, has been opening more of the software surrounding Ascend. Its CANN platform — the company’s counterpart to the programming infrastructure required to exploit AI accelerators — supports frameworks including PyTorch and works with open-source projects such as Triton, vLLM and veRL.

Huawei said last month that CANN had moved to sustained community-driven open-source development, with external programmers accounting for 61 per cent of its CANN developers. The company said the community had more than 5,200 monthly active developers, while more than 40 models had been natively pre-trained using Ascend and CANN.

The combination matters because China’s challenge to Nvidia has increasingly shifted from producing an alternative accelerator to assembling a usable full computing stack. Hardware can win orders only if model developers can train and run workloads efficiently without sacrificing too much engineering time or performance.

DeepSeek’s software also gives Huawei an influential independent developer helping test and improve that stack. The AI laboratory became prominent for building capable models with an emphasis on computational efficiency, and its infrastructure code is closely watched by developers seeking lower-cost approaches to training and inference.

The effort comes as US export controls continue to restrict China’s access to Nvidia’s most advanced processors, creating stronger incentives for technology groups to adopt domestic alternatives. Those restrictions have helped expand the strategic importance of Huawei’s Ascend line, even though China still faces constraints in advanced semiconductor manufacturing.

Nvidia nevertheless retains formidable advantages. CUDA has a vast installed developer base and an extensive collection of mature libraries, while Nvidia’s leading accelerators remain central to major AI training systems outside China. Compatibility layers and open-source libraries also require sustained testing, documentation and optimisation before they can match the reliability of an established platform across diverse workloads.