DeepSeek Hub

AMD's new Ryzen AI Max+ 395 'Strix Halo' APU is 3x faster in DeepSeek R1 AI bench than RTX 5080

AMD's new Ryzen AI Max 395 'Strix Halo' APU gets benchmarked with DeepSeek R1 AI models: over 3x faster than NVIDIA's new GeForce RTX 5080.

Apple Mac Studio cluster: 4 x M3 Ultra systems use half the power of a network switch at idle

Apple launched its new M3 Ultra chip inside of the new Mac Studio, with a new Mac Cluster involving 4 systems using around 33W at idle, less than a switch.

Apple's new M3 Ultra runs DeepSeek R1 with 671B parameters: uses 448GB of RAM, only 200W power

Apple's new M3 Ultra processor slices and dices DeepSeek R1 models: uses 448GB of unified RAM, only 200W of power... no multi-GPU setup necessary.

RTX 5090 scalped for up to $5000 on Chinese black markets, goes for up to $10K on eBay

Chinese AI firms running DeepSeek now prefer getting NVIDIA consumer-level GeForce RTX graphics cards, RTX 5090 sells for $5000 on Chinese black market.

Huawei goes all-in with its first AI laptop built entirely in-house

An industry analyst has revealed that Huawei plans to release its first AI laptop, built entirely from in-house components, in April.

NVIDIA AI servers used in Singapore allegedly using banned AI GPUs smuggled through Malaysia

Dell is investigating whether Dell and Super Micro servers shipped to Malaysia had NVIDIA AI GPUs that were banned from China, new scandal.

New underwater data center capsule in China packs the power of 30,000 high-end gaming PCs

The new capsule linked to an existing underwater data center complex can handle up to 7,000 conversations per second for DeepSeek-powered AI agents.

NVIDIA stocks have almost fully recovered, 1 month after DeepSeek panic sales

Just a month after DeepSeek devastated NVIDIA's market cap and stock price - the AI chipmakers financials have almost fully recovered.

DeepSeek is rumored to be working on in-house AI chips, would compete with NVIDIA

DeepSeek is reportedly working on a semiconductor venture, rurmored to be working on developing its own in-house AI chips that would compete with NVIDIA.

Sam Altman backpedals on release of o3, shortly after de-hyping 'superintelligence' claims

Sam Altman took to X to detail OpenAI's new plans for their upcoming models. Including the cancellation of o3, and the push to release GPT-5.

Trump administration's stance heading into Paris AI summit: 'America first'

Heading into the Paris AI Action Summit, the Trump administration are focused on maintaining US dominance in the field of artificial intelligence.

Perplexity update outperforms ChatGPT's paid plan (and it's free)

In a social media post, Perplexity announced that they've increased the context window of their file and image processing tools for free users.

China's humanoid robots slated for mass production by end of 2025

Recent reports highlight how Chinese robotics manufacturer, UBTech, are pushing forward with the release of AI-enabled humanoid robots this year.

China's AI expansion continues, DeepSeek now in PCs, Robotics, EVs and more

Recent reports reveal that Chinese firms are rapidly deploying DeepSeek R1 AI across multiple industries, integrating it into PCs, robotics, EVs, and more.

Apple stays cautious on AI as DeepSeek's budget model shakes up the industry

As DeepSeek disrupts AI with low-cost innovation, and tech giants battle for users - Apple sticks to its slow-and-steady strategy. Will it pay off?

ChatGPT makes previously paid feature available to everyone - no login required

OpenAI took to X on Wednesday to inform users that ChatGPT's AI-powered search feature is now available for all users without needing an account.

US researchers built a DeepSeek competitor for less than a tank of gas - and it's actually good

A recent paper, published by researchers from Stanford and the University of Washington, highlights a notable development in the creation of low-cost AI.

Google releases Gemini 2.0 models: 'We kicked off the agentic era'

In an announcement on Wednesday morning, Google has officially released the Gemini 2.0 series of AI models, including Gemini 2.0 Flash and Gemini 2.0 Pro.

AMD downplays DeepSeek threat: '[Innovation is] good for AI adoption'

In a recent Q4 earnings call, AMD's CEO Lisa Su provided some comments on hardware sales, demand for AI, and DeepSeek's effect on the broader industry.

Australian government bans DeepSeek due to national security risks

The Australian Government has responded to the security concerns presented by DeepSeek, issuing a federal ban that has come into immediate effect.

DeepSeek AI begins to steal customers away from ChatGPT

DeepSeek is beginning to take a top global position in the AI chatbot rankings, with customers now appearing to move away from OpenAI's ChatGPT.

Singapore responds to accusations of smuggling illegal NVIDIA GPUs to China

Questions have been raised for Singapore's reported involvement in being an intermediary for China to acquire NVIDIA GPUs the US government has banned.

DeepSeek AI found to be stunningly vulnerable to jailbreaking

Researchers have pitted DeepSeek's R1 model against several harmful prompts and found it's particularly susceptible to jailbreaking.

DeepSeek misled the AI market, training costs range in the billions not millions

A research firm has estimated the expenditure needed to create DeepSeek's R1 model, which caused the market to drain $1 trillion when it was unveiled.

ChatGPT creator OpenAI launches o3-mini: its new 'reasoning' AI model to compete with DeepSeek

OpenAI is launching its new o3-mini reasining model inside of ChatGPT, new model responds 24% faster than o1-mini, providing more accurate answers, more.

NVIDIA CEO Jensen Huang expected to meet with President Trump today, will talk all things AI

NVIDIA CEO Jensen Huang is expected to meet with President Trump at the White House today, will talk about the future of the AI industry for the USA.

Apple CEO Tim Cook says DeepSeek shows 'innovation that drives efficiency'

Apple CEO Tim Cook comments on Chinese AI surprise DeepSeek, says 'innovation that drives efficiency is agood thing' as it competes with ChatGPT.

US officially launches probe into DeepSeek over illegal use of NVIDIA chips

Officials in the White House and FBI have begun a probe into DeepSeek to determine whether banned NVIDIA chips were used to create the R1 model.

AMD shows you how to run DeepSeek on your Ryzen AI CPU and Radeon GPU

AMD provides instructions on how to run DeepSeek's R1 AI model on Ryzen AI processors and Radeon GPUs, running locally on your PC.

Move over OpenAI and DeepSeek, a new Chinese AI model claims supremacy over both

After the markets felt the pressure of DeepSeek's R1 model, another Chinese AI model claims to outperform both R1 and OpenAI models.

DeepSeek database contains chat history, internal secrets for anyone to see

Cloud security services provider Wiz Research discovers DeepSeek security keys, messages, and more... includings chat history, API secrets.

OpenAI says DeepSeek stole its data to train its breakthrough AI

OpenAI has alleged that DeepSeek stole data from the company to train its R1 AI model and used a special technique to mask its tracks.

OpenAI CEO Sam Altman responds to DeepSeek AI hype amid widespread stock value crash

OpenAI CEO Sam Altman has responded to the market hype of the recently unveiled DeepSeek AI, which caused tech company stocks to plummet.

NVIDIA breaks silence after biggest single-day loss of any company in history

NVIDIA's stock price tumbled 17% on Monday after Chinese start-up DeepSeek unveiled its R1 model, and now the green team has responded.

Ex-Intel CEO Pat Gelsinger buys the dip with NVIDIA stock, is using DeepSeek R1 for his startup

Ex-CEO of Intel, Pat Gelsinger, has come out praising China's new DeepSeek R1 model: also announces he's buying the dip with cheap NVIDIA shares.

Analysts say that DeepSeek is the 'Temu of AI' as it wreaks havoc on the expensive AI industry

Wedbush analysts say that DeepSeek's success will be short lived, and that China's game-changing ChatGPT competitor is the 'Temu of AI'.

DeepSeek R1 was trained on NVIDIA H800 AI GPUs, inferencing is done on Huawei 910C AI chips

DeepSeek R1 model was trained on NVIDIA H800 AI GPUs, while inferencing was done on Chinese made chips from Huawei, the new 910C AI chip.

OpenAI CEO calls DeepSeek R1 'impressive' and says it's 'invigorating to have a new competitor'

DeepSeek is making waves with its cost-effective and powerful AI models, with OpenAI CEO Sam Altman praising the Chinese AI startup's AI advancements.

Razer Blade 16 with GeForce RTX 5090 will be hard to find because of AI

The new Razer Blade 16 with AMD Ryzen AI 9 HX 370 CPU and GeForce RTX 5090 laptop GPU is coming soon, but availability will be limited.

President Trump plans tariffs on foreign semiconductors, says DeepSeek AI success 'a positive'

President Trump on DeepSeek AI, says 'should be a wake-up call for our industries that we need to be laser-focused on competing to win'.

NVIDIA statement on DeepSeek: says Chinese AI company is an 'excellent AI advancement'

NVIDIA issues a statement regarding Chinese AI company DeepSeek, after its game-changing AI model, says it's an 'excellent AI advancement'.

DeepSeek unleashes Janus-Pro 7B model: focuses on task-specific AI models, rivals DALL-E 3

Chinese AI company DeepSeek releases new Janus-Pro 7B model: multimodal AI models that are meant to outperform OpenAI's dominant DALL-E 3.

NVIDIA shares bleed $384 billion in value in a few hours after China's DeepSeek shocks AI world

NVIDIA shares are down 11% in pre-market trading, bleeding $384 BILLION in a few hours over fears of China's game-changing DeepSeek AI startup.

DeepSeek's new AI assistant app is the top free app on the Apple App Store, ahead of ChatGPT

Chinese AI company DeepSeek has huge success on the Apple App Store: its AI assistant app is the top free app, beating OpenAI's ChatGPT app.

Chinese AI firm DeepSeek has 50,000 NVIDIA H100 AI GPUs says CEO, even with US restrictions

Chinese AI company DeepSeek says its DeepSeek R1 model is as good, or better than OpenAI's new o1 says CEO: powered by 50,000 NVIDIA H100 AI GPUs.