Close Menu
Karachi Chronicle
  • Home
  • AI
  • Business
  • Entertainment
  • Fashion
  • Politics
  • Sports
  • Tech
  • World

Subscribe to Updates

Subscribe to our newsletter and never miss our latest news

Subscribe my Newsletter for New Posts & tips Let's stay updated!

What's Hot

The world’s largest air force with the F-35 fleet in 2025

AI systems learn from many types of scientific information and run experiments to discover new materials | MIT News

Among the most troublesome relationships in healthcare AI

Facebook X (Twitter) Instagram
  • Home
  • About us
  • Advertise
  • Contact us
  • DMCA
  • Privacy Policy
  • Terms & Conditions
Facebook X (Twitter) Instagram Pinterest Vimeo
Karachi Chronicle
  • Home
  • AI
  • Business
  • Entertainment
  • Fashion
  • Politics
  • Sports
  • Tech
  • World
Karachi Chronicle
You are at:Home » Does Deepseek’s AI mean that US technical management failed?
Business

Does Deepseek’s AI mean that US technical management failed?

Adnan MaharBy Adnan MaharJanuary 28, 2025No Comments6 Mins Read0 Views
Facebook Twitter Pinterest Telegram LinkedIn Tumblr Email Reddit
Share
Facebook Twitter LinkedIn Pinterest WhatsApp Email


The United States has been working steadily in the past three years and has restricted China’s access to state -of -the -art computer chips, which supplies power to advanced artificial intelligence systems. The purpose was to delay China’s progress in the development of a sophisticated AI model.

Currently, DeepSeek, a Chinese company, has created that technology. In recent weeks, DeepSeek has released multiple AI models and chatbots. The performance is comparable to the best product performance made by American companies, but we use much less AI chips that companies usually need. On the weekend, DeepSeek’s chatbot was shot at the top of the Apple’s App Store chart, and people downloaded it around the world.

This development has caused a major question about export management built by the United States in recent years. The Biden administration set a global rule system, steadily expanded, and tried to maintain advanced AI technology, especially NVIDIA’s chips from China. They were concerned that technology gave China economically, as well as military advantage.

The development of DeepSeek has caused a fierce debate on whether US technical management failed. There is something you need to know here.

Deepseek’s innovation suggests that the Biden administration has taken too much action to avoid its management.

DeepSeek states that the latest model was trained with NVIDIA H800S. This was an AI chip developed by NVIDIA for the Chinese market after the first export management was imposed, causing a considerable amount of dramas in Washington.

When the United States set a limit on NVIDIA’s most advanced chips in 2022, NVIDIA was promptly adapted by creating a slightly lowered chip just below the government set. These chips were technically legal to use by Chinese companies, but they were substantially achieved the same results.

It offended the Biden officials, and they moved to limit new chips. However, the government moved slowly and took about a year to ban H800 and other downgraded chips. Until then, Chinese companies have stockpiled many of them.

It is not clear how DeepSeek has acquired the NVIDIA H800, but it would have been legal to buy the company in late 2022 or 2023. But now that is not the case.

“I can’t control the thing there is,” said Jimmy Goodrich, a senior adviser in Rand Corporation technology. “If the Biden administration responded more quickly and the H800 was limited to China, there was no doubt that DeepSeek was more challenged when this model was released.”

Deepseek also spent many years to build chips before Washington’s control was enabled. By 2021, according to an interview with Liang Wenfeng, the founder of Deepseek, a Chinese media outlet 36kr, at least 10,000 NVIDIA Acquired It was only a handful of Chinese companies.

The United States is also struggling to cancel chips smuggling.

There is no evidence that DeepSeek used a smuggling chip. However, many Chinese AI companies have. ALEXANDR WANG, the highest executive officer of AI Training Giant Scale AI, tells New York Times that Chinese companies have much more high -end chips than US restrictions. He said he had a NVIDIA Advanced H100 processor. I can’t speak. “

Both NVIDIA and the US government argue that the size of smuggling is limited. But last year’s Times reported aggressive trade in China in restricted AI technology. In the lively market of Shenzhen in southern China, Chip Vendors reported that they are engaged in sales involving hundreds or thousands of restrictions.

11 representatives of the 11 companies said they sold or transported banned NVIDIA chips, including A100 and H100, which were the most advanced at the time. One vendor of the deep shenzhen has shown a reporter message that allows server distribution, including the most advanced chips of NVIDIA, over 2,000. This is a total of $ 130 million transactions.

Since then, more reports have been recorded, especially through other countries in Asia, have recorded large -scale smuggling.

This month, the Biden administration has announced a drastic regulation that NVIDIA has set an upper limit on the number of chips that can be sold to all countries around the world.

It is still unknown what the Trump administration does about it. President Trump signed the first day of his inauguration by the Presidential Order of Trade, but ordered his staff to review US export management systems, such as “how to identify and eliminate missing holes in existing export management.”

The US control seems to encourage China’s ingenuity, but they are also obviously hindering China’s AI development.

The restrictions on American technology seem to accelerate the efforts of Chinese researchers to do more in a small amount.

The most notable thing about the Deepseek model is that the company has been developed in only a few high -priced chips used by Western companies to create similar technologies. DeepSeek engineers use only about 2,000 NVIDIA chips, while most top companies are training chatbots using more than 16,000 chips. NVIDIA’s shares suddenly fell sharply on Monday, fearing that technology companies could do the most advanced AI in the future, much less paid for NVIDIA.

Jeff Lydin, a professor at the University of George Wasington, is studying emerging technology, but most global companies use more computing power and data to improve AI performance. I did it. However, Deepseek and other Chinese companies have been forced to go down to other routes to find out if they can get enough performance by reducing training costs and low calculation. ” He said.

The meaning of a cheap model like DeepSeek is profound. Open to the details of how DeepSeek has built the model, companies all over China and the world will be able to reproduce low -cost approaches.

In other words, “The energy to build and operate small and medium -sized businesses in the Midwest, North Korea’s hackers, and Russian AI from Hyper Scaler in the United States may be much less,” said Peter. Martin Chorzenpa, a senior fellow at the Song Research Institute, said. For international economics.

Nevertheless, China will be far ahead of AI without export management. In an interview, DeepSeek founder acknowledges that lack of access to computing power is a restriction for the company.

Unlike American AI companies, Deepseek cannot legally purchase the latest AI chips of NVIDIA. This increases the speed and performance of the previous chips.

“Anyone who is worried that Deepseek can do today will be more worried if he accesses a much better computing resources that US competitors have,” said Chorzempa. I said.

DeepSeek’s success suggests that the lead to Silicon Valley’s AI has shrunk. However, it is worth noting that DeepSeek is still building a model on the NVIDIA chip. It is not a rival AI chip that Chinese technology companies Huawei are trying to develop.

Among the Chinese computer engineers, Huawei’s AI chips can be performed with more advanced chips, including the chips created by Huawei, despite the much lower performance. It suggests that there is.

However, it is not yet possible to create advanced AI chips comparable to NVIDIA or the complex machines required to make their chips. “At this time, the only advantage that the United States is still more than China is the hardware,” said Good Rich.



Source link

Share. Facebook Twitter Pinterest LinkedIn Reddit WhatsApp Telegram Email
Previous ArticleSam Altoman’s nuclear fusion startup Helion confuses the science community
Next Article A small reactor revolution is ongoing
Adnan Mahar
  • Website

Adnan is a passionate doctor from Pakistan with a keen interest in exploring the world of politics, sports, and international affairs. As an avid reader and lifelong learner, he is deeply committed to sharing insights, perspectives, and thought-provoking ideas. His journey combines a love for knowledge with an analytical approach to current events, aiming to inspire meaningful conversations and broaden understanding across a wide range of topics.

Related Posts

Kodak Vision3 Film Stock Enhancement – New AHU Design

September 1, 2025

Cryptocurrency Live News & Updates: Vaneck proposes SolanaETF for traditional investors

August 21, 2025

The paint sector slows down: Early monsoons and price war hit revenues in June quarter, revival of corporate eye celebrations

August 17, 2025
Leave A Reply Cancel Reply

Top Posts

20 Most Anticipated Sex Movies of 2025

January 22, 2025459 Views

President Trump’s SEC nominee Paul Atkins marries multi-billion dollar roof fortune

December 14, 2024122 Views

How to tell the difference between fake and genuine Adidas Sambas

December 26, 202486 Views

Alice Munro’s Passive Voice | New Yorker

December 23, 202474 Views
Don't Miss
AI September 25, 2025

AI systems learn from many types of scientific information and run experiments to discover new materials | MIT News

Machine learning models can speed up discovery of new materials by making predictions and proposing…

Among the most troublesome relationships in healthcare AI

Does access to AI become a fundamental human right? Sam Altman says, “Everyone would want…”

Google’s Gemini AI is on TV

Subscribe to Updates

Subscribe to our newsletter and never miss our latest news

Subscribe my Newsletter for New Posts & tips Let's stay updated!

About Us
About Us

Welcome to Karachi Chronicle, your go-to source for the latest and most insightful updates across a range of topics that matter most in today’s fast-paced world. We are dedicated to delivering timely, accurate, and engaging content that covers a variety of subjects including Sports, Politics, World Affairs, Entertainment, and the ever-evolving field of Artificial Intelligence.

Facebook X (Twitter) Pinterest YouTube WhatsApp
Our Picks

The world’s largest air force with the F-35 fleet in 2025

AI systems learn from many types of scientific information and run experiments to discover new materials | MIT News

Among the most troublesome relationships in healthcare AI

Most Popular

10 things you should never say to an AI chatbot

November 10, 20040 Views

Character.AI faces lawsuit over child safety concerns

December 12, 20050 Views

Analyst warns Salesforce investors about AI agent optimism

July 1, 20070 Views
© 2025 karachichronicle. Designed by karachichronicle.
  • Home
  • About us
  • Advertise
  • Contact us
  • DMCA
  • Privacy Policy
  • Terms & Conditions

Type above and press Enter to search. Press Esc to cancel.