Alibaba Group Holding Ltd (阿里巴巴) cofounder Jack Ma (馬雲)-backed Ant Group Co (螞蟻集團) used Chinese-made semiconductors to develop techniques for training artificial intelligence (AI) models that would cut costs by 20 percent, people familiar with the matter said.
Ant used domestic chips, including from Alibaba and Huawei Technologies Co (華為), to train models using the so-called “mixture of experts” machine learning approach, the people said.
It got results similar to those from Nvidia Corp chips, such as the H800, they said.
Photo: AFP
Hangzhou-based Ant is still using Nvidia for AI development, but is now relying mostly on alternatives, including from Advanced Micro Devices Inc and Chinese chips for its latest models, one of the people said.
The models mark Ant’s entry into a race between Chinese and US companies that has accelerated since DeepSeek (深度求索) demonstrated how capable models can be trained for far less than the billions invested by OpenAI and Alphabet Inc’s Google.
It underscores how Chinese companies are trying to use local alternatives to the most advanced Nvidia semiconductors. While not the most advanced, the H800 is a relatively powerful processor and is barred by the US from China.
The company published a research paper this month that said its models at times outperformed Meta Platforms Inc in certain benchmarks, which has not been independently verified.
However, if they work as advertised, Ant’s platforms could mark another step forward for Chinese AI development by slashing the cost of inferencing or supporting AI services.
Ant said it cost about 6.35 million yuan (US$875,952) to train 1 trillion tokens using high-performance hardware, but its optimized approach would cut that down to 5.1 million yuan using lower-specification hardware.
Tokens are the fundamental units of text — such as words, characters or parts of words — that a language model breaks down and analyzes to understand context, meaning and structure.
In essence, they are the building blocks that enable the model to interpret human language and produce intelligent output.
The company plans to leverage the recent breakthrough in the large language models it has developed, Ling-Plus and Ling-Lite, for industrial AI solutions including healthcare and finance, the people said.
On English-language understanding, Ant in its paper said that the Ling-Lite model did better in a key benchmark compared with one of Meta’s Llama models.
Ling-Lite and Ling-Plus models outperformed DeepSeek’s equivalents on Chinese-language benchmarks.
Ant has made the Ling models open-source. Ling-Lite contains 16.8 billion parameters, which are adjustable settings that work like knobs and dials to direct the model’s performance.
Ling-Plus has 290 billion parameters, which is considered relatively large in the realm of language models. For comparison, experts estimate that ChatGPT’s GPT-4.5 has 1.8 trillion parameters, MIT Technology Review said. DeepSeek-R1 has 671 billion.
The company faced challenges in some areas of the training, including stability.
Even small changes in the hardware or the model’s structure led to problems, including jumps in the models’ error rate, it said in the paper.
Additonal reporting by staff writer
ELECTRONICS BOOST: A predicted surge in exports would likely be driven by ICT products, exports of which have soared 84.7 percent from a year earlier, DBS said DBS Bank Ltd (星展銀行) yesterday raised its GDP growth forecast for Taiwan this year to 4 percent from 3 percent, citing robust demand for artificial intelligence (AI)-related exports and accelerated shipment activity, which are expected to offset potential headwinds from US tariffs. “Our GDP growth forecast for 2025 is revised up to 4 percent from 3 percent to reflect front-loaded exports and strong AI demand,” Singapore-based DBS senior economist Ma Tieying (馬鐵英) said in an online briefing. Taiwan’s second-quarter performance beat expectations, with GDP growth likely surpassing 5 percent, driven by a 34.1 percent year-on-year increase in exports, Ma said, citing government
‘REMARKABLE SHOWING’: The economy likely grew 5 percent in the first half of the year, although it would likely taper off significantly, TIER economist Gordon Sun said The Taiwan Institute of Economic Research (TIER) yesterday raised Taiwan’s GDP growth forecast for this year to 3.02 percent, citing robust export-driven expansion in the first half that is likely to give way to a notable slowdown later in the year as the front-loading of global shipments fades. The revised projection marks an upward adjustment of 0.11 percentage points from April’s estimate, driven by a surge in exports and corporate inventory buildup ahead of possible US tariff hikes, TIER economist Gordon Sun (孫明德) told a news conference in Taipei. Taiwan’s economy likely grew more than 5 percent in the first six months
SMART MANUFACTURING: The company aims to have its production close to the market end, but attracting investment is still a challenge, the firm’s president said Delta Electronics Inc (台達電) yesterday said its long-term global production plan would stay unchanged amid geopolitical and tariff policy uncertainties, citing its diversified global deployment. With operations in Taiwan, Thailand, China, India, Europe and the US, Delta follows a “produce at the market end” strategy and bases its production on customer demand, with major site plans unchanged, Delta president Simon Chang (張訓海) said on the sidelines of a company event yesterday. Thailand would remain Delta’s second headquarters, as stated in its first-quarter earnings conference, with its plant there adopting a full smart manufacturing system, Chang said. Thailand is the firm’s second-largest overseas
SUPPLY RESILIENCE: The extra expense would be worth it, as the US firm is diversifying chip sourcing to avert disruptions similar to the one during the pandemic, the CEO said Advanced Micro Devices Inc (AMD) chief executive officer Lisa Su (蘇姿丰) on Wednesday said that the chips her company gets from supplier Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) would cost more when they are produced in TSMC’s Arizona facilities. Compared with similar parts from factories in Taiwan, the US chips would be “more than 5 percent, but less than 20 percent” in terms of higher costs, she said at an artificial intelligence (AI) event in Washington. AMD expects its first chips from TSMC’s Arizona facilities by the end of the year, Su said. The extra expense is worth it, because the company is