Strategic Foresight - Insights in Business and Technology

Deep Seek's GPU Infrastructure Initially acquired 10,000 GPUs in 2021 Estimated to have grown to around 50,000 GPUs in total Used 2,000 H800 GPUs specifically for V3 model pre-training Share infrastructure with their quantitative trading fund operations Initial Export Control Framework US government initially restricted two parameters : Computing power (FLOPS) Interconnect bandwidth between GPUs This two-factor restriction created an opportunity for optimisation H800 GPU Restrictions and Adaptations H800 was China's version of the H100 GPU Two key restriction factors from the US government: Chip compute (FLOPS) Interconnect bandwidth H800 was designed with: Full FLOPS capability (same as H100) Restricted interconnect bandwidth Deep Seek developed specialized SM (Streaming Multiprocessor) scheduling techniques to work around interconnect limitations Managed to achieve full GPU utilisation despite interconnect restrictions Export Control Evolution First Phase: Dual restric...

 Let's break down how Snowflake, Palantir, and Databricks work together in the AI world, using a technology stack analogy and real-world examples. The AI Technology Stack Think of building an AI-powered company like building a house. You need a solid foundation, a smart design, and skilled builders. Foundation (Data): Snowflake Layman's Terms: Snowflake is like the concrete foundation of your AI house. It stores all your data in one organized place, making it easy to access and use. It's not just storage; it's like a super-organized library where any information can be found instantly.   Technical Function: Snowflake is a cloud-based data warehouse. It allows companies to store vast amounts of structured and semi-structured data, making it readily available for analysis and AI model training. It handles the messy work of data organization and access.   Example: Imagine a retail company. Snowflake stores all their sales data, customer information, inven...

  The Future of Enterprise AI: Palantir's AIP Palantir's AI Platform (AIP) is revolutionising how enterprises harness data's power. By integrating, analysing, and visualising vast datasets, AIP enables organisations to uncover valuable insights and make informed decisions. What Does Palantir AIP Offer? At its core, Palantir AIP is an ontology-driven platform. This means it uses a structured knowledge graph to represent concepts, entities, and their relationships. This foundational layer allows AIP to: Integrate diverse data sources: Seamlessly combine data from various sources, including structured and unstructured data. Visualise complex relationships: Use powerful visualisation tools to explore connections and patterns within data. Support decision-making: Provide actionable insights to drive strategic decisions and optimise operations. Opportunities for Service Providers For service providers, Palantir AIP presents a wealth of opportunities: Skill Development: Invest in ...

The Future of Software is New SaaS - powered by Services, AI Agents, Sharing This POV is available for download below.

 Salesforce - Under The AI Cloud  -- Marc Known for his mission and marketing mantra, ”The End of Software" Revolutionised the CRM industry by introducing Cloud-based service.  Despite facing tough competition from companies like Oracle, SAP, and Microsoft, it has emerged as the leader in this field. Today it commands 23% of the Global Market Share. Strategically acquired companies like Tableau, Slack (Data & Analytics) and MuleSoft (PaaS) to build a platform-centric flywheel by integrating CRM and Business System applications. With a Market Cap of US $209Bn and a Robust Economic Moat, it can become a $500Bn company by the end of this decade. Strategic shift from Competition (MSFT 365, ORCL) to Cooperation (Google, OpenAI) Risks to AI Ambitions My other posts on  AI Value Chain , Microsoft Transformation - The King of AI in Software and Cloud   and how  Generative AI can transform Telecoms ,  Energy and Utilities Source:  Seekingalpha ,...

 NVIDIA - The Godfather of AI - Why the Market is Bullish Key Message Q2 FY23 -  The data centre business, driven by demand from cloud providers and internet companies, played a vital role, contributing $10.32 billion (76% of total revenue). Gaming revenue also grew positively to $2.49 billion. Products like the H100 tensor core GPU, DGX supercomputers, inference platforms, and AI Infrastructure-as-a-Service in the cloud are poised to transform AI delivery, partnered with major cloud providers. Today Nvidia commands 70% of the AI Chip market. Other Key players are AMD, ARM (Softbank) and Intel.  If you take the other Nvidia divisions separate them out and add them together, revenues were up by 9.9% to $3.18Bn. That’s another way of saying that the Nvidia datacentre business is now 3.2x bigger than the rest of Nvidia, and the datacentre business now accounts for 76.4% of the overall sales for the company. A single chip can cost upwards of US$40K  and Open AI used abou...

  Open AI - Why Google is Threatened If search without AI was an order of complexity - one, then search with AI is like an order of complexity two. Google is now launching its own #AI chatbot - BardAI which will respond to complex queries instead of simple questions served by google doodle. AI is now venturing into more complex tasks; hopefully, more players will participate and emerge. More info on Open AI can be found here

  My predictions on Macro Economic View of Australia FY23 How Humans and AI are Collaborating  The B2B Lead Gen Stumbling Blocks and How to Overcome Them Given the complex nature of B2B lead generation, it goes without saying that there are plenty of stumbling blocks that can hinder lead gen efforts. With this in mind, we’re diving into the common pain points marketers experience when filling the funnel while offering solutions that lead to better marketing results. 1. Marketing Without a Cohesive Lead Generation Strategy  2. Marketers are Focusing on Lead Quantity Instead of Quality  3. There are Gaps in Content Marketing Efforts  4. Content Marketing Isn’t Aligned with the B2B Buyer Journey  5. Marketing Efforts Aren’t Consistently Optimised  B2B buyers Transparency make Vendor Purchase Easier Research has shown that nearly two-thirds (65%) of B2B buyers felt vendors placed more emphasis on selling than they did on listening to their needs. Because o...

Top 10 Artificial Intelligence Companies in Media Ecosystem  Appier  GumGum  Sizmek by Amazon  IBM Watson  Dstillery  Invoca  Influential  Quantcast  Heuritech  NetBase Quid  More details here

  Bridging the gap: How and why product management differs from company to company Companies, like many in New York, that are introducing solutions to mature industries tend to be sales-driven, so product teams are informed by what customers will readily buy and are measured based on output and revenue. Companies, like many in the Bay Area, that are creating new categories are generally vision-driven, so product teams are more cross-functional and iterative, and are measured based on validated learnings and outcomes. The two models on each end of the spectrum manage risk differently and, with regard to startups, offer a varied range of results. Predicated on introducing solutions to previously unarticulated problems (e.g. Snapchat), Bay Area startups and product managers are tasked with launching moonshots. Failure is more frequent and expensive but success is exponential in size. How Marketing can assist in Maximising Growth Fortunately, there's a good rule of thumb called ...