Ambassador Chen Mingjian Attended the 4th China-Tanzania Job Fair
Distributed by APO Group on behalf of Embassy of the People's Republic of China in the United Republic of Tanzania.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The National
7 hours ago
- The National
Trump says Nvidia's H20 is 'obsolete', yet there are worries about selling the GPU to China
US President Donald Trump explained his decision to allow Nvidia to sell its H20 graphics processing unit (GPU) to China on Monday by describing it as obsolete. More specifically, Mr Trump was asked about making a deal with Nvidia for the US government to receive a 15 per cent cut of sales to Beijing. 'I deal with Jensen [Huang], who is a great guy,' Mr Trump said, referring to Nvidia's chief executive and founder. 'The chip that we're talking about, the H20, it's an old chip and China already has it … but it's obsolete and it still has a market.' That begs the question, however, why there is so much demand from China for the H20, and why some elected officials and analysts are so concerned that Mr Trump will allow the chip to be sold to the country. By most standards, Mr Trump's description of the H20 as obsolete is incorrect – to describe a technology, particularly a piece of hardware like the H20 GPU, as obsolete is to imply that it's no longer supported by the tech world, making it completely impractical to use within an existing technology ecosystem. That's not the case with the H20, which was designed several years ago by Nvidia to comply with US regulations seeking to prevent powerful AI technologies from being used by countries it views as adversaries. In recent weeks, a group of Democratic senators sent a letter to White House officials criticising the move to allow the resumption of H20 sales to China, claiming that because of the GPU's improved power-efficiency and high-bandwidth memory chips, 'many PRC [People's Republic of China] firms reportedly prefer the H20 to other controlled chips'. That's hardly an indication a GPU design is obsolete. Late in July, in response to The National's inquires about concerns about the H20 being sold in China, Nvidia did not exactly describe the GPU as obsolete either. 'The H20 helps America win the support of developers worldwide, promoting America's economic and national security,' the statement read, emphasising the chip's deliberately modest attributes. 'It does not enhance anyone's military capabilities, and the US government has full visibility and authority over every H20 transactions.' Some Republicans have also described Nvidia's H20 as still relevant. Representative John Moolenaar, who chairs the House Select Committee on the Chinese Communist Party, recently opposed resuming exports of the H20 to China. 'The H20 is a stark improvement over what the Chinese can indigenously produce at scale,' he said during a speech at Purdue University's Krach Institute for Tech Diplomacy. 'When, not if, these chips are diverted to [China's] People's Liberation Army supercomputer, they will substantially upgrade their ability to run advanced AI models.' H20 paranoia abounds Regardless of whether or not one considers the H20 to be obsolete, there is a tremendous amount of speculation surrounding the chip, with some of the paranoia coming from China, where there is demand for it. This month, Nvidia sought to debunk reports that claimed some of its chips had a backdoor kill switch, to blunt any misuse that might compromise national security. Some members of the US Congress have even sought to make it mandatory for certain chips to contain kill switches. 'Nvidia GPUs do not and should not have kill switches and backdoors,' a blog post from the company read. 'No kill switches, and no spyware … that's not how trustworthy systems are built – and never will be.' Despite those assurances, however, recent reports have indicated China still has concerns. According to Bloomberg, Chinese officials are telling various companies within the country to not use Nvidia's H20 over fears that the chips might be compromised and vulnerable to being exploited. The Chinese embassy in Washington has not yet responded to The National's requests for comment. Trump praises Nvidia's Blackwell architecture Despite ample conjecture about the H20 coupled with the varying opinions about Nvidia's sales of it to China, during Mr Trump's news conference on Monday, he made clear that the company's other GPU offerings would likely not be exported to China anytime soon, at least not without a few changes. Mr Trump said Nvidia's highly touted Blackwell GPUs would need to be toned down before he would consider allowing them to be sold in China. 'It's super-duper advanced and I wouldn't make a deal with that,' he said, before quickly giving himself some wiggle room. 'Although it's possible that I would make a deal if they enhanced it in a negative way.' Mr Trump said based on his conversations with Mr Huang, he's impressed with the Blackwell platform, adding that in terms of processing power, none of Nvidia's competitors have it, 'and they probably won't for five years'. As for the 15 per cent cut that the Trump White House wants for the H20 sales to China, on Tuesday, press secretary Karoline Leavitt said similar approaches might be applied to other companies and products, although she said specifics were still being worked out by the Department of Commerce. Some have questioned the constitutionality of the move, referring to it as a "reverse tariff".
Zawya
10 hours ago
- Zawya
African Development Bank backs Artificial Intelligence (AI) training to boost Agenda 2063 implementation across Africa
The African Development Bank ( is supporting a new Artificial Intelligence (AI) training initiative to advance the implementation of Africa's continental development blueprint, Agenda 2063: The Africa We Want ( Through its Joint Secretariat Support Office (JSSO), the Bank provided technical and financial backing for the 5th Annual Training Workshop for African Union Member States on harnessing AI for effective monitoring, evaluation, and reporting on the Second Ten-Year Plan (2024–2033) of Agenda 2063. The five-day workshop, held in Lusaka, Zambia, was jointly organized by the African Union Commission (AUC) and the African Capacity Building Foundation (ACBF). The event brought together representatives from African Union Member States to strengthen their technical capacity in using AI tools and digital innovations to monitor, track, and report on the implementation of Agenda 2063. The training is part of broader efforts to enhance institutional and human capacity across the continent. Participants engaged in practical sessions on emerging AI platforms, including Ailyse, ChatGPT, Google AI Studio, Google Gemini, and Perplexity, exploring how these tools can improve data analysis, enhance decision-making, and promote evidence-based policy development in line with Agenda 2063 ( priorities. The Bank's support to the workshop reflects its broader commitment to strengthening results-based planning, monitoring, and accountability within the African Union framework, while promoting innovation and digital transformation as key enablers of Africa's development agenda. Speaking at the workshop, Abibu Tamu, Lead Programme Co-ordinator at African Development Bank Group, reaffirmed the Bank's dedication to working closely with the African Union Commission and strategic partners to accelerate the implementation of Agenda 2063, particularly its Second Ten-Year Implementation Plan. 'These tools are not only revolutionizing how data is collected, analysed, and reported, they are also enabling more targeted policy interventions and efficient resource allocation,' said Tamu. The workshop also served as a platform for peer learning and knowledge exchange, with participating countries sharing innovative approaches and best practices in national development planning and results-based reporting. Distributed by APO Group on behalf of African Development Bank Group (AfDB).
Zawya
11 hours ago
- Zawya
AI and data are reshaping Africa's retail sector
The African retail landscape is under immense pressure. Consumers are more digitally savvy and price-conscious than ever, and expect brands to offer convenience, personalisation, and value. About half of consumers will switch brands if their expectations are unmet, compelling retailers to continuously innovate and improve. Industry growth is tapering down across several African markets. South Africa's real retail growth has declined in recent years, constrained by economic stagnation and inflation. In East Africa, while GDP growth remains relatively strong, formal retail continues to be constrained by infrastructure gaps and logistical inefficiencies. In response, retailers across the continent are scaling their technology investments to boost competitiveness in an increasingly digital and data-driven industry. According to Gartner, African retailers are projected to invest $300m in cloud-based enterprise applications in 2025, with an expected growth of 10-12% over the next three years And yet, many still struggle to translate these investments into meaningful business outcomes. The triple disconnect holding retailers back 1. Fragmented systems, broken processes undermine scalability Despite deploying multiple best-of-breed applications and investing in a data lake to centralise data, many retailers fail to achieve differentiated customer engagement, real-time visibility and reliability in their operations. Typical constraints include disconnected systems, a lack of business context within data sets, and AI bolted-on instead being embedded within operational workflows. A pervasive industry pain point is the disconnected customer journeys caused by fragmented systems and broken processes. To illustrate: a customer may browse a product online, call the nearest store to check stock availability, visit in person and still find a different price to what was listed online. Or they may purchase the product online but be unable to return the product instore due to a disconnect between online and physical store data. This is caused when customer data, inventory, pricing, and orders are not synchronised across channels. And it's not just limited to customer experience - these fragmented processes are widespread across retail operations, often leading to operational inefficiencies and higher costs. At the root of this problem is a disconnected application landscape. In fact, 66% of organisations say application sprawl and complexity hinder their digital goals, resulting in process inefficiencies and a lack of reliability in retail operations. 2. Data is an opportunity, but a challenge too African retailers are sitting on mountains of data spanning point-of-sale systems, loyalty platforms, mobile apps, and third-party sources. But is this high volume, variety, and velocity of data being leveraged effectively to drive business outcomes? Unfortunately, the answer is no. Around 55% of business leaders cite poor data quality and fragmentation as the biggest obstacles to making data-driven decisions. Moreover, a disproportionate amount of time is spent on data management and building dashboards as against decision-making, underscoring the urgent need for more efficient processes. 3. AI without a solid foundation doesn't scale While there's palpable excitement in boardrooms over the potential and power of AI, execution remains limited. One study found that nearly three-quarters of organisations are struggling to scale AI projects beyond pilot projects. African retailers face even greater challenges, including poor-quality data, outdated IT infrastructure, and a lack of AI-ready processes. The result is AI initiatives that turn into isolated experiments that fail to deliver any meaningful ROI. Instead, retailers need to leverage a unified data platform that embeds AI directly into their business processes to scale AI initiatives effectively. The power of a unifying platform Leading retailers leverage a comprehensive, connected, and industry-specific suite of applications to confidently execute across their value chain. This helps ensure a consistent and reliable customer experience, minimising stock-outs, enabling fast, on-time delivery and delivering a seamless omni-channel journey. This reliability extends across store operations, finance, and support processes. The SAP Business Suite, for example, enables this efficiency through automation, actionable insights, and process optimisation, while the SAP Business Data Cloud provides seamless access to organisational data. Retailers also benefit from deep industry expertise and global context that helps them identify the correct application mix for integrating, harmonising and transforming business data across various applications. This also provides the foundation to scale AI effectively across retail operations through unified, context-rich data that brings together sales, inventory, promotional, customer behaviour and other data. By connecting end-to-end processes, retailers can also embed AI and enable seamless cross-functional workflows while leveraging innovations from a rich ecosystem of partnerships. Retailers leveraging SAP's connected platform and AI capabilities report a 20-30% reduction in customer churn, 30-50% fewer stockout, an up to 40% increase in workforce productivity and 20-30% reduction in IT spend. Realising Africa's data-rich, AI-powered retail future requires bold steps to shift beyond legacy systems and redefine operations. By utilising a powerful business suite, retailers can unlock new end-to-end capabilities that bring together applications, data and AI for unrivalled competitiveness. All rights reserved. © 2022. Provided by SyndiGate Media Inc. (
