Due to the US embargo, China, which has developed its own products in many areas, especially mobile processors, has accelerated its artificial intelligence studies. Intellifusion introduced its artificial intelligence card called DeepEyes, which can reach 48 TOPS. Here are the details…
Intellifusion DeepEyes artificial intelligence card features
The first “Deep Eyes” artificial intelligence card of 2024 leverages the DeepEdge10Max processor for 48 TOPS in int8 training performance. The 2024 H2 DeepEyes card, which is stated to be introduced in the second half of the year, will use the DeepEdge10Pro processor supporting 24 TOPS.
Described as the most powerful model of the series, 2025 H1 DeepEyes will be powered by the DeepEdge10Ultra processor, which can reach 96 TOPS. Claiming that they will come up with a product that is 90 percent cheaper than the market average, the company aims to produce hardware that will meet Microsoft’s artificial intelligence-supported computer requirements.
ChatGPT compared to real doctors! The result is as expected…
A remarkable study was conducted on the artificial intelligence-supported chat bot ChatGPT. Accordingly, it was seen that the vehicle was faster than the doctors.
Microsoft announced that artificial intelligence-supported computers should have at least 40 TOPS. Considering that current artificial intelligence chips have an average value of 16 TOPS, it is possible to say that Intellifusion can dominate the industry.
All of the introduced hardware leverages Intellifusion’s custom NNP400T neural network chip. Looking at the technical specifications of the chips, the 1.8GHz 2+8-core RISC CPU and up to 800 MHz GPU in DeepEdge 10 more than meet expectations.
Making a statement on the subject, Intellifusion president Dr. Chen Ning said, “In the next three years, 80 percent of companies worldwide will use major AI models. The cost of training a large model is tens of millions of dollars, and the average price of all-in-one training machines for general use is one million yuan. “Most companies cannot afford such costs.” He stated: