Miscellaneous News

[jwnz]

Some one bribe Jim Cramer to say that this is just a minor correction and that NVDA will shoot up to 400 dollars in no time.

On the other hand, where is the Downfall meme for the DeepSeek shock?

 

[00CuriousObserver]

View attachment 144513
View attachment 144514

It's happening again

It's only been 2 weeks... Please

---

@Temstar

 

[siegecrossbow]

On the other hand, where is the Downfall meme for the DeepSeek shock?

What’s the point when you can do one for the Trump regime before the decade is out.

 

[tankphobia]

Tbh this was bound to happen. The brute force approach to train models does not work when you have effectively already inputted all of human knowledge in text form into the model. Exact same thing has already happened to video games. A magnitude increase in processing power have not resulted in linear improvement in performance, why would AI be any different?

At some point smarter inputs will have equal if not better response than simply throwing more oomph into the machine.

This whole fiasco is just American venture capitalists falling for tech bubble once again, nothing in tech is infinitely scalable.

 

[siegecrossbow]

It's only been 2 weeks... Please

---

@Temstar

View attachment 144515

You tell the poor man/woman that if the Chinese Dream is realized, all they need to survive is to pretend to be a XJP Thought enthusiast but if Trump actually makes America great again, they’ll need extensive skin grafting from a Caucasian donor.

 

[Chevalier]

At this rate India should stop comparing itself to China; it’s embarrassing but Indians have no shame.

 

[tokenanalyst]

A lot of people in the US are becoming financial and economic illiterates. When the price of a technology goes down the investment usually goes up. Imagine these idiots in the 80s when MOS technology introduced the 6052 and sold those for 20 bucks when the rest of the market was selling much more expensive CPUs, what comes next? the personal computer revolution. This could open the door for more companies to enter the game which will mean more investment. ​

 

[plawolf]

Please, Log in or Register to view URLs content!

Most of the photos of Liang Wenfeng used by the world media are wrong. This is the boss of the same name.

Poor dude is going to get Sarah Connor’d by the CIA at this rate.

 

[daifo]

I wonder if China is very close to EUV , efficient DUV or complete semiconductor supply chain then we think they are at. Seems they are just publicly flexing hard the past month rather than hiding their strength, unafraid of likely call for even more tech sanctions from the derange us hawks.

 

[FriedButter]

Please, Log in or Register to view URLs content!

Nvidia says DeepSeek advances prove need for more of its chips​

SAN FRANCISCO, Jan 27 (Reuters) - Nvidia (NVDA.O), opens new tab on Monday said Chinese AI firm DeepSeek's advances show the usefulness of its chips for the Chinese market and that more of its chips will be needed in the future to meet demand for DeepSeek's services.

Nvidia issued a statement on Monday after its shares tumbled 17% on investor concerns that China's DeepSeek had matched rivals such as OpenAI using far fewer Nvidia chips than U.S. firms.

"DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant," Nvidia said in its statement.

One of DeepSeek's research paper's showed that it had used about 2,000 of Nvidia's H800 chips, which were designed to comply with U.S. export controls released in 2022 but which experts told Reuters would barely slow China's AI progress.

The U.S. microchip export controls were designed to freeze China's development of supercomputers used to develop nuclear weapons and artificial intelligence systems.
Jimmy Goodrich, a senior adviser to the RAND Corp for technology analysis, said there are at least a dozen major supercomputers in China with significant numbers of Nvidia chips that were legal for purchase at the time that DeepSeek used to learn how to become more efficient.

Computing efficiency has also been a major focus of U.S. AI firms.

"DeepSeek didn't come out of nowhere - they've been at model building for years," Goodrich said. "It's been long known that DeepSeek has a really good team, and if they had access to even more compute, God knows how capable they would be."
DeepSeek was struggling on Monday to accommodate an influx of new users. Servicing new users is a process that AI firms call "inference," which Nvidia said demonstrated that its chips will remain in demand.

"Inference requires significant numbers of Nvidia GPUs and high-performance networking," Nvidia said in its statement.
Nvidia is currently selling a chip called the H20 that is designed to meet the most recent export control regulations. While the restrictions limit the chip's usefulness for AI training, Goodrich said it is "probably the best chip in the world for inference."
"How long will Washington allow the best inference chip in the world to be sold to China?" he added.

"DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant," Nvidia said in its statement.

It sounds like Nvidia is saying DeepSeek didn’t use any banned GPUs. An huge shipment 5000cc of IV Copium is going to be needed for tens of thousands tech bros and China hawks.