DeepSeek gives China's chipmakers leg up in race for cheaper AI
The rise of DeepSeek's artificial intelligence (AI) models is seen providing some Chinese chipmakers such as Huawei a better chance to compete in the domestic market against more powerful U.S. processors.
Huawei and its Chinese peers have for years struggled to match Nvidia in building top-end chips that could compete with the U.S. firm's products for training models, a process where data is fed to algorithms to help them learn to make accurate decisions.
However, DeepSeek's models, which focus on "inference," or when an AI model produces conclusions, optimise computational efficiency rather than relying solely on raw processing power.
That is one reason why the model is expected to partly close the gap between what Chinese-made AI processors and their more powerful U.S. counterparts can do, analysts say.
Huawei, and other Chinese AI chipmakers such as Hygon, Tencent-backed EnFlame, Tsingmicro and Moore Threads have in recent weeks issued statements claiming products will support DeepSeek models, although few details have been released.
Huawei declined to comment. Moore Threads, Hygon EnFlame and Tsingmicro did not respond to Reuters queries seeking further comment.
Industry executives are now predicting that DeepSeek's open-source nature and its low fees could boost adoption of AI and the development of real-life applications for the technology, helping Chinese firms overcome U.S. export curbs on their most powerful chips.
Even before DeepSeek made headlines this year, products such as Huawei's Ascend 910B were seen by customers such as ByteDance as better suited for less computationally intensive "inference" tasks, the stage after training that involves trained AI models making predictions or performing tasks, such as through chatbots.
In China, dozens of companies from automakers to telecoms providers have announced plans to integrate DeepSeek's models with their products and operations.
"This development is very much aligned with the capability of Chinese AI chipset vendors," said Lian Jye Su, a chief analyst at tech research firm Omdia.
"Chinese AI chipsets struggle to compete with Nvidia's GPU (graphics processing unit) in AI training, but AI inference workloads are much more forgiving and require a lot more local and industry-specific understanding," he said.
NVIDIA STILL DOMINATES
However, Bernstein analyst Lin Qingyuan said while Chinese AI chips were cost-competitive for inferencing, this was limited to the Chinese market as Nvidia chips were still better even for inference tasks.
While U.S. export restrictions ban Nvidia's most advanced AI training chips from entering China, the company is still allowed to sell less powerful training chips that Chinese customers can use for inference tasks.
Nvidia published a blog post on Thursday about how inference time was rising as a new scaling law and argued that its chips will be necessary to make DeepSeek and other "reasoning" models more useful.
In addition to computing power, Nvidia's CUDA, a parallel computing platform that allows software developers to use Nvidia GPUs for general-purpose computing, not just AI or graphics, has become a crucial component of its dominance.
Previously, many Chinese AI chip companies did not directly challenge Nvidia by asking users to abandon CUDA but instead, claimed their chips were compatible with CUDA.
Huawei has been the most aggressive in its efforts to break away from Nvidia by offering a CUDA equivalent called Compute Architecture for Neural Networks (CANN), but experts said it faced obstacles in persuading developers to abandon CUDA.
"Software performance of Chinese AI chip firms is also lacking at this stage. CUDA has a rich library and a diverse range of software capability, which requires significant long-term investment," said Omdia's Su.
Follow Emirates 24|7 on Google News.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Gulf Today
an hour ago
- Gulf Today
China's forex reserves in May up by a less-than-expected $3.6 billion
China's foreign exchange reserves rose by a less-than-expected $3.6 billion in May, official data showed on Saturday, as the dollar continued to weaken against other major currencies. The country's foreign exchange reserves, the world's largest, rose 0.11 per cent to $3.285 trillion last month, below the Reuters forecast of $3.292 trillion. They were $3.282 trillion in April. The increase in reserves was due to 'the combined effects of factors such as exchange rate conversion and asset price changes,' China's State Administration of Foreign Exchange said in a statement. The yuan weakened 1.05 per cent against the dollar in May, while the dollar slid 0.23 per cent against a basket of other major currencies. China's foreign exchange reserves rose by a less-than-expected $3 billion in May, official data showed on Saturday, as the dollar continued to weaken against other major currencies. The country's foreign exchange reserves, the world's largest, rose to $3.285 trillion last month, below the Reuters forecast of $3.292 trillion. They were $3.282 trillion in April. The yuan weakened 1.05 per cent against the dollar in May, while the dollar slid 0.23 per cent against a basket of other major currencies. Meanwhile China's central bank added gold to its reserves in May for the seventh straight month, official data from the People's Bank of China (PBOC) showed on Saturday. Spot prices for gold, often seen as a refuge from economic and geopolitical uncertainty, were steady in May after hitting an all-time high of $3,500 per ounce in April. China's gold reserves rose to 73.83 million fine troy ounces at the end of May from 73.77 million ounces at the end of April. Its gold reserves were valued at $241.99 billion at the end of last month, down from $243.59 billion at the end of April, the PBOC said. With bullion up 27% so far this year due to tariff war fears on top of 27% growth in 2024, gold market specialists say the PBOC's readiness to continue building up gold holdings despite high prices reflects Beijing's desire to diversify its foreign currency reserves. Officials at the PBOC have not publicly said what was driving the gold purchases. In 2024, the PBOC took a six-month pause after an 18-month-long gold purchasing spree, before resuming buying gold in November, when Donald Trump won the US presidential election. The world's central banks are on track to buy 1,000 metric tonnes of gold in 2025, which would be their fourth year of massive purchases as they diversify reserves from dollar-denominated assets into bullion, consultancy Metals Focus said this week. Meanwhile China is willing to accelerate the examination and approval of rare earth exports to European Union firms and will also deliver a verdict on its trade investigation of EU brandy imports by July 5, its commerce ministry said on Saturday. Price commitment consultations between China and the EU on Chinese-made electric vehicles exported to the EU have also entered a final stage but efforts from both sides are still needed, according to a statement on the Chinese commerce ministry's website. The issues were discussed between Chinese Commerce Minister Wang Wentao and EU Trade Commissioner Maros Sefcovic in Paris on Tuesday, according to the statement. The comments mark progress on matters that have vexed China's relationship with the European Union over the past year. Most recently, China's decision in April to suspend exports of a wide range of rare earths and related magnets has upended the supply chains central to automakers, aerospace manufacturers, semiconductor companies and military contractors around the world. The ministry said China attached great importance to the EU's concerns and 'was willing to establish a green channel for qualified applications to speed up the approval process.' Commerce Minister Wang during the meeting 'expressed the hope that the EU will meet us halfway and take effective measures to facilitate, safeguard and promote compliant trade in high-tech products to China,' according to the statement. Chinese anti-dumping measures that applied duties of up to 39 per cent on imports of European brandy - with French cognac bearing the brunt - have also strained relations between Paris and Beijing. The brandy duties were enforced days after the European Union took action against Chinese-made electric vehicle imports to shield its local industry, prompting France's President Emmanuel Macron to accuse Beijing of 'pure retaliation'. The Chinese duties have dented sales of brands including LVMH's Hennessy, Pernod Ricard's Martell and Remy Cointreau. Beijing was initially meant to make a final decision on the brandy duties by January, but extended the deadline to April and then again to July 5. Reuters
Al Etihad
2 hours ago
- Al Etihad
Beijing proposes leasing export of rare earths to EU
7 June 2025 14:35 Beijing (AFP)China has proposed establishing a 'green channel' to ease the export of rare earths to the European Union, the commerce ministry said Saturday, after Beijing restricted their sale has since April required licenses to export these strategic materials from China, which accounts for more than 60 percent of rare earth mining production and 92 percent of global refined output, according to the International Energy metals are used in a wide variety of products, including electric car batteries, and there has been criticism from industries about the way China's licenses have been issued.'Export control on rare earths and other items is an international practice,' the commerce ministry said in a statement.'China attaches great importance to Europe's concerns and is willing to establish a green channel for eligible applications, fast track the examination and approval, and instruct the working level to maintain timely communication on this,' a ministry statement comments were attributed to China's Commerce Minister Wang Wentao, who met Tuesday with EU trade commissioner Maros the bilateral talks, Wang said he hoped the bloc would turn 'take reciprocal action, adopt effective measures to facilitate, safeguard, and promote compliant trade of high-tech products with China,' according to the commerce two officials also discussed imports by European countries of Chinese electric vehicles, which the EU has hit with levies over allegedly unfair subsidies from Beijing.'The negotiation on the price commitment of electric vehicles between China and Europe has entered the final stage, but both sides still need to make efforts,' the commerce ministry said. The discussions will be followed by China hosting a summit with the EU next month, 50 years since Beijing and Brussels established diplomatic relations.
Gulf Business
5 hours ago
- Gulf Business
Executive education: Accenture's Abir Habbal on preparing AI-savvy leaders
Image: Supplied The AI revolution isn't coming — it's already here. But for most business leaders, bridging the gap between AI hype and actionable strategy remains a challenge. Enter the Generative AI Scholars Programme, a joint initiative by Accenture and Stanford University, now being rolled out across the Middle East. In this interview, Abir Habbal, Data and AI Strategy & Consulting lead at Accenture in the Middle East, explains what makes this programme more than just another executive course. From bite-sized modules grounded in Stanford's academic legacy to real-world applications tailored for the region's ambitious digital visions, this programme is designed to turn C-suite curiosity into capability. Habbal also shares why the UAE and Saudi Arabia are ideal launchpads, highlights the top misconceptions around AI adoption, and offers a glimpse into the real business impact already emerging from this new generation of AI-savvy leaders. In your view, what role can executive education like this play in accelerating the region's digital transformation — and what's still needed to close the gap between ambition and impact? Executive education programmes are instrumental in accelerating the Middle East's AI transformation by comprehensively equipping leaders for the AI era. They are crucial for building AI-literate leadership, empowering decision-makers across the GCC to not only understand AI's potential but also to strategically apply it within their organisations. This involves a fundamental shift in mindset: moving from viewing AI purely as an IT function to recognsing it as a catalyst for business model transformation and sustainable growth. By fostering AI fluency, these programmes bridge the communication gap between the c-suite and technical teams, enabling more productive dialogues and streamlined decision-making, which is vital for the region's ambitious national AI strategies. However, to fully close the gap between this ambition and tangible impact, several critical elements are still needed. Beyond leadership, there must be a broader investment in role-based AI training across all levels of the workforce, ensuring everyone understands how AI integrates into their daily tasks and contributes to organisational goals. Furthermore, the emphasis on responsible AI must be woven into the fabric of every initiative, with clear governance models addressing data privacy, bias, and transparency from inception, as this builds crucial trust for widespread adoption. Ultimately, sustained success hinges on a commitment to scaling what works, fostering a culture of continuous learning and experimentation, and aligning AI innovation directly with national economic diversification and digital transformation priorities across the region. There's growing awareness of AI across industries — but a notable gap when it comes to implementation. What are the most common misconceptions or barriers you see among business leaders trying to adopt AI? Despite high awareness, a significant gap exists between AI understanding and real-world implementation among business leaders. Accenture's research highlights that while 84 per cent of c-suite executives believe they must leverage AI to achieve their growth objectives, only 15 per cent feel their organization is truly ready to scale it. The most common misconceptions and barriers include: Uncertainty about where to start: The rapidly evolving AI landscape makes it challenging to differentiate hype from achievable business value. Lack of leadership alignment: If c-suite leaders (CIO, CFO, CHRO) don't operate from a shared understanding, AI initiatives often stall at the pilot stage. Organisational unreadiness: Many businesses lack the foundational data infrastructure, skilled talent, or robust governance frameworks needed for responsible AI scaling. As Accenture's research indicates, 70-80 per cent of AI initiatives never move beyond the pilot phase, and while many focus on technical capabilities, successful AI implementation is predominantly a people and process challenge. Misconception of AI as purely an IT initiative: Leaders often fail to see AI as a core strategic imperative for growth, innovation, and competitiveness, viewing it simply as a technological tool rather than a catalyst for business model transformation. Risk avoidance over responsible risk-taking: Concerns around ethics, trust, and governance, while valid, can lead to a paralysis of innovation if not balanced with a strategy of embracing innovation with built-in guardrails and responsible AI practices from day one. Lack of clarity on ROI: A significant barrier is establishing ROI on identified opportunities and making a business case for scaling initiatives, which are often perceived as more challenging than technical limitations Can you walk us through what makes this programme distinct from other executive education offerings — particularly in how it blends Stanford's academic insights with real-world application? The programme includes easily digestible, bite-sized modules, industry spotlights, case studies, and reflection activities, ensuring participants not only grasp complex concepts like technical fundamentals, foundation models, and prompt engineering but also develop the generative AI strategy and technology know-how for real-world application. This approach aims to spark reinvention agendas that can profoundly transform businesses, enabling leaders to drive innovation and navigate the digital economy effectively. Why was the Middle East chosen as the next region for the rollout of this programme, and how has the regional business landscape influenced its evolution or delivery? The Middle East, specifically Saudi Arabia and the UAE have emerged as prime focus for the rollout of programmes like the Generative AI Scholars Program due to the country visions and their unparalleled ambition and strategic commitment to becoming global leaders in AI. This region is not merely adopting AI; it's actively leading its development and integration into national visions. Both nations are making substantial financial commitments to AI infrastructure, research centers, and digital ecosystems. This includes significant government-backed AI R&D funds, free zones offering incentives for AI businesses, and partnerships with global cloud technology organisations. The region is not just investing in technology but also in building a modern digital core, which Accenture sees as essential for continuous reinvention and for organisations to rapidly seize every opportunity presented by AI. Recognisng that technology adoption requires human capital, these nations are heavily investing in developing AI fluency across their workforces, from top leadership to technical teams. We are working with MCIT Saudi Arabia in an Accenture artificial intelligence training programme organised by the Accenture LearnVantage Academy. These programmes directly address human capital requirements by equipping senior officials and business leaders with the mindset and skills needed to lead with AI responsibly. As the programme moves from theory to action, what kind of real-world business outcomes are participants expected to achieve? Can you share any early examples from previous rollouts globally? The expected real-world business outcomes are centered on driving measurable value and competitive advantage. Accenture identifies these outcomes as broader and more strategic than just cost savings. Participants are expected to achieve: Enhanced decision-making: AI tools help leadership teams make faster, more informed decisions, leading to improved strategic agility. Increased employee productivity and empowerment: By leveraging AI, leaders can free up time for their teams to focus on high-impact work rather than being buried in manual analysis or reporting. This contributes to a positive human-AI relationship, which Accenture believes is a key priority for leaders. Tangible business impact: Strategies informed by AI insights are expected to translate into winning new business, improving customer satisfaction, and accelerating the launch of initiatives. Accenture's AI Achievers report indicates that 63 per cent of high-performing companies say they've already achieved measurable ROI from their AI investments within three years. Accelerated organistional agility: AI helps organisations react faster to market changes, or spot opportunities we would've missed before, fostering a state of continuous reinvention. Business model transformation: Rather than just incremental gains, organizations are expected to achieve step-change improvements in revenue, efficiency, and customer experience by integrating AI effectively, as leaders rethink how digital systems are designed, how people work, and how they create products and interact with customers. Innovation at scale: The programme aims to unlock creativity and accelerate progress, sector by sector, by empowering people to reimagine what's possible with AI.
