AMD Expands to Full AI Infrastructure with Helios Racks and Supercluster

Key Takeaways

• AMD deploys Helios racks in Europe and an Instinct-powered supercluster in India, expanding from chip sales to full AI infrastructure.
• Databricks launches AI Runtime (AIR), a serverless GPU platform used by Robinhood, Merck, and FactSet for faster deep learning training.
• OpenAI's models now available on Google Cloud, completing presence across all major U.S. cloud providers after Microsoft exclusivity ended.
• New York opens a $10 million AI Jobs Institute in Queens to train 40,000 lower-income residents for tech jobs by 2030.
• Gartner forecasts global IT spending will reach $7.35 trillion in 2027, with AI infrastructure driving the next $1 trillion in growth.
• IBM CEO Arvind Krishna says enterprises are reining in AI spending, turning to open-weight models at 10 to 25 percent of frontier model token costs.
• Bybit AI adds personalization and memory features, building user profiles across Markets, Trading, Earn, and Card services for its 80 million users.
• Tyler Technologies' Ask Fairfax AI assistant has answered over 40,000 resident questions in Fairfax County, Virginia, replacing a legacy chatbot.
• Michael Burry bets against Nvidia, Palantir, Tesla, and Micron, citing overvaluation and concerns about chip depreciation and memory cycles.
• LiteLLM suffered a supply chain attack on October 21, 2026, when a credential stealer was shipped through a compromised GitHub CLI token.

Fairfax launches AI assistant for Tyler Technologies

Tyler Technologies reported that Fairfax County, Virginia launched Ask Fairfax, an AI assistant that replaced a legacy chatbot. The tool has answered over 40,000 resident questions using official county website content. The deployment shows real usage for Tyler's Resident AI Assistant in public sector workloads like tax, vehicle registration, and court information.

Advanced Energy launches new AI power products

Advanced Energy Industries introduced M-CRPS power supplies, the Centra Smart Hub with remote sensors, and a single-whip option for its 33 kW ORv3 High-Power Rack shelf. The new 3,600 W OCP-compliant supplies target data center, AI, telecom, and semiconductor thermal processing workloads. The products focus on higher wattage, rack-level architectures and precise semiconductor temperature control tied to AI compute demand.

Ralph Lauren expands AI tools for customer engagement

Ralph Lauren has been investing in artificial intelligence, advanced analytics, and digital tools to improve customer engagement and online shopping. The company expanded its presence across major large language models and launched the RL mobile app in Korea. Management uses AI in areas such as personalization, search, and warehouse automation as part of its technology investments.

AMD secures AI infrastructure wins with Helios racks

Advanced Micro Devices reported new AI infrastructure wins including Helios rack deployments for European robotics data factories and an AMD Instinct powered AI supercluster in India. AMD also launched the open AMD Helion agentic AI rack and partnered with nVent for rack level cooling for the Helios platform. The announcements highlight AMD's shift from selling individual processors to supplying full AI data center systems.

SentinelOne launches AI security tools with partner support

SentinelOne rolled out Singularity Agentic Detection and Response and Singularity ASOC, expanding alliances with Lenovo, LevelBlue, and N-able. The new products aim to broaden AI-driven security coverage across endpoints, cloud, identities, and managed services. Singularity ASOC pulls together Purple, AI SIEM, Hyperautomation, and AI Data Pipelines into a more autonomous security operations center loop.

New York launches $10 million AI training hub in Queens

New York is spending $10 million to train 40,000 lower-income people for AI jobs. The new AI Jobs Institute will open in Queens and is run by the nonprofit Pursuit. Governor Kathy Hochul announced the program on October 22. The institute will teach people to use tools like ChatGPT and build AI agents. It aims to help New Yorkers get good-paying tech jobs by 2030.

Hochul announces AI training hub in Queens

Governor Kathy Hochul announced the AI Jobs Institute in Queens. The hub is designed by tech training group Pursuit and will train 40,000 New Yorkers by 2032. It will offer AI career courses and help connect workers with employers. NYC Council Speaker Julie Menin also held a hearing on AI rules. She wants New York City to remain the AI capital of the country.

Databricks launches AI Runtime for GPU training

Databricks made AI Runtime (AIR) generally available on October 21, 2026. AIR is a serverless GPU platform built into the Databricks Lakehouse for training deep learning models. Companies like Robinhood, Merck, and FactSet already use it. Databricks says AIR cuts training time and works with tools like PyTorch and Ray. It also integrates with Lakeflow and Unity Catalog for governance.

3 AI stocks that could split in 2027

Nvidia and Broadcom split their stock shares in 2024 after big AI rallies. Now SanDisk, GE Vernova, and AMD are trading at high prices between $530 and $1,300. SanDisk sells flash memory for AI data centers. GE Vernova makes power plant equipment boosted by AI demand. AMD has seen data center sales double and could follow Nvidia's example with a stock split.

Bybit AI adds personalization and memory features

Bybit launched new features for Bybit AI on October 22, 2026. The tool now builds a user profile based on assets, positions, and risk profile. It connects Markets, Trading, Earn, and Card services within Bybit. Users can share preferences that Bybit AI remembers over time. The platform also introduces Featured entry points for tasks like wealth planning. Bybit serves over 80 million users worldwide.

IBM CEO says big companies are rethinking runaway AI spending

IBM CEO Arvind Krishna says larger enterprises are becoming more aware of the real return on their AI investments. Companies worry that some AI spending is out of control with no real return. Businesses can use open-weight models at 10 to 25 percent of the token cost of frontier models. IBM's on-premises infrastructure helps protect data by running models locally or on private clouds.

Gartner predicts 15 percent growth in global IT spending by 2027

Gartner forecasts worldwide IT spending will reach $7.35 trillion in 2027, up 15 percent from 2026. AI infrastructure, AI-optimized servers, AI-optimized IaaS, and AI software platforms will drive the next $1 trillion in growth. Hardware markets face growing pressure from memory-related inflation. Organizations are seeking to industrialize AI deployment through these spending areas.

LiteLLM suspects GitHub token leak led to malicious package attack

On October 21, 2026, LiteLLM's release pipeline was compromised and a credential stealer was shipped to users. The poisoned version ran a stealer routine called privcollect to gather system, browser, cloud, and AI tool secrets. The attacker used GitHub CLI credentials to disable branch protections and push malicious changes. LiteLLM said Claude Code may embed GitHub OAuth tokens in responses that get logged. The maintainers hardened their release pipeline after the incident.

OpenAI brings its AI models to Google Cloud after Microsoft deal

OpenAI announced its models will be available through Google Cloud for developers. This puts OpenAI's models on all three major U.S. cloud providers. Amazon started hosting OpenAI models in April after exclusivity ended with Microsoft. The move supports OpenAI's push to expand distribution to enterprise developers beyond its consumer roots. Google Cloud reported annualized revenue growth of about 82 percent in the second quarter driven by AI demand.

Michael Burry bets against four major AI stocks

Michael Burry, founder of Scion Asset Management, is betting against Nvidia, Palantir Technologies, Tesla, and Micron Technology. He criticizes Nvidia's circular financing and chip depreciation methods. He questions Palantir's high price-to-sales ratio and Tesla's high price with lower profits. He also raises concerns about Micron's memory cycle. Burry believes these stocks are overvalued and may experience a pullback.

Sources

• Is Fairfax AI Rollout Altering The Investment Case For Tyler Technologies (TYL)?
• Advanced Energy Stock Is In Focus After New AI Power Products Launch
• Is AI Expansion Altering The Investment Case For Ralph Lauren (RL)?
• Did AI Infrastructure Wins Just Shift AMD Stock's Investment Narrative?
• How Investors May Respond To SentinelOne (S) AI Security Launch
• No city employs more tech workers than New York. Now it’s spending $10 million to train 40,000 people for AI jobs
• Hochul announces AI-focused training hub in Queens
• Databricks Makes AI Runtime Generally Available for GPU Training
• 3 AI Stocks That Could Split in 2027
• Bybit AI Empowers Users with Their Personal Financial Edge, Adding Personalization and Memory
• Big companies are rethinking 'out of control' AI spending — and turning to cheaper models
• Gartner forecasts 15% global IT spending growth, recaps top tech predictions
• LiteLLM Suspects GitHub Token Leak in AI Gateway Logs Behind Malicious Package Release
• OpenAI to bring AI models to Google's cloud, in further expansion beyond Microsoft
• Michael Burry Is Betting Against These 4 AI Stocks. Is He Right?