Artificial Intelligence - Page 56

It was only last month that OpenAI revealed its upcoming text-to-video generator platform named Sora, and the general reaction to the new AI-powered tool was impressive yet concerning.

The upcoming AI-powered tool works exactly the same way as OpenAI's extremely popular ChatGPT, but instead of the chatbot responding to user prompts with text its capable of producing high-quality video, even to the point of photorealism. OpenAI took to its YouTube channel to share the above video showcasing Sora's capabilities and at first glance it appears some of the examples shown are shot with a real-life camera.

However, upon closer inspection of the examples tell-tale signs of AI-generated content begin to stand out, such as physics-based movements like people walking, hand movements, and more. OpenAI is currently "red-teaming" Sora to iron out these issues before its released to the public, which means people are pushing the AI model to its brink to bring these vulnerabilities to light so they can be fixed.

NVIDIA has had an absolute record-breaking last 12 months or so, but that momentum isn't slowing down... it's only ramping up... to a huge predicted $130 billion in revenue once we get to 2026.

In a new report from Bloomberg, they predict NVIDIA revenue will swell to a huge $130 billion in 2026, a gargantuan $100 billion increase from 2021. The crazy numbers are fueled by the insatiable AI GPU demand, which NVIDIA is absolutely dominating in... and that's just with current-gen H100 AI GPU offerings, let alone its soon-to-be-released H200 AI GPU, and its next-gen Blackwell B100 AI GPU both right around the corner.

We already heard last year that NVIDIA was expected to generate $300 billion in AI-powered sales by 2027, so the leap from $130 billion to $300 billion in a single year -- 2026 to 2027 -- is absolutely mammoth. We've got market researchers like Omdia, predicting NVIDIA to make $87 billion this year from its data center GPUs, and with next-gen AI GPUs right around the corner... well, NVIDIA is really just getting started.

We know that AI is big business, and that is why companies like Microsoft, Meta, Google, and Amazon are investing mind-boggling amounts of money in creating new infrastructure and AI-focused data centers. As per Meta's latest post regarding its "GenAI Infrastructure," the company has announced two "24,576 GPU data center scale clusters" to support current and next-gen AI models, research, and development.

That's over 24,000 NVIDIA Tensor Core H100 GPUs, with Meta adding that its AI infrastructure and data centers will house 350,000 NVIDIA H100 GPUs by the end of 2024. There's only one response to seeing that many GPUs: a comically long and cartoonish whistle or a Neo-style "Woah." Meta is going all in on AI, a market in which it wants to be the leader.

"To lead in developing AI means leading investments in hardware infrastructure," the pot writes. "Meta's long-term vision is to build artificial general intelligence (AGI) that is open and built responsibly so that it can be widely available for everyone to benefit from."

Samsung is reportedly using MUF technology for its next-gen HBM chip production, with the South Korean giant reportedly issuing purchasing orders for MUF tools.

The company says that the "rumors" it will use MUF technology are "not true," according to Reuters, which is reporting the news. HBM makers like SK hynix, Micron, and Samsung are all fighting for the future of HBM technology and future-gen AI GPUs, and it seems Samsung has its tail between its legs now.

One reason Samsung is falling behind is that it has stuck with its chip-making technology, non-conductive film (NCF), which has caused production issues. Meanwhile, HBM competitor and South Korean rival SK Hynix has switched to mass reflow molded underfill (MR-MUF) to work through NCF's weakness, "according to analysts and industry watchers," reports Reuters.

HBM3E memory is about to be unleashed with NVIDIA's upcoming beefed-up H200 AI GPU, but now JEDEC has reportedly relaxed the rules for HBM4 memory configurations.

JEDEC has reportedly reduced the package thickness of HBM4 down to 775 micrometers for both 12-layer and 16-layer HBM4 stacks, as it gets more complex at higher thickness levels, making it easier... especially as HBM makers fly in the face of insatiable demand for AI GPUs (now, and into the future with HBM4-powered chips).

HBM manufacturers, including SK hynix, Micron, and Samsung, were poised to use hybrid bonding with the process, a newer packaging technology, and more to reduce the package thickness of HBM4, which uses direct bonding with the onboard chip and wafer. However, HBM4, being a new technology, sees that hybrid bonding would increase pricing, making HBM4-powered AI GPUs of the future even more expensive.

Cerebras Systems just unveiled its new WSE-3 AI chip with 4 trillion transistors and 900,000 AI-optimized cores... as well as its new CS-3 AI supercomputer.

The new CS-3 AI supercomputer has enough power to train models that are 10x larger than GPT-4 and Gemini, which is thanks to its gigantic memory pool. Cerebras Systems' new CS-3 AI supercomputer has been designed for enterprise and hyperscale users, delivering huge performance efficiency gains over current AI GPUs.

The new Condor Galaxy 3 supercomputer features 64 x CS-3 AI systems, packing 8 Exaflops of AI compute performance, which is double the performance of the previous system, but at the same power... and the same cost.

Cerebras Systems has just revealed its third-generation wafer-scale engine (WSE) chip, WSE-3, which packs 4 trillion transistors and 900,000 AI-optimized cores.

The company hasn't stopped on its journey of AI processor releases, with some truly crazy specifications for Cerebras' new WSE-3 chip. We have 4 trillion transistors, 900,000 AI-optimized cores, 125 petaflops of peak AI performance, and 44GB of on-chip SRAM made on the 5nm process node at TSMC.

WSE-3 also features either 1.5TB, 12TB, or 1.2PB of external memory -- yeah, 1.2 petabytes of memory -- capable of training AI models with up to 24 trillion parameters. Cerebras says its new WSE-3 has a die size of 46,225mm2, which is an insane 57x larger than NVIDIA's current H100 AI GPU, which measures 826mm2.

Microsoft's Copilot just got a lot better for users of the free version of the AI with the introduction of the GPT-4 Turbo model.

Previously, GPT-4 Turbo was only available to paying subscribers - those on Copilot Pro. However, as Mikhail Parakhin, head of Advertising and Web Services at Microsoft, made clear on X (formerly Twitter), the more advanced model is now available to those on the freebie version of Copilot.

Free users were on vanilla GPT-4 up until now, with the Turbo version - which provides faster responses, and better, more accurate ones, to boot - effectively paywalled.

A new paper penned by researchers from Google DeepMind, ETH Zurich, University of Washington, OpenAI, and McGill University reveals that OpenAI and Google's AI models have been cracked open.

The new paper reveals that thirteen computer scientists from the aforementioned locations were able to launch an attack on OpenAI and Google's closed AI services, and this attack resulted in the revealing of a significant hidden portion of the underlying transformer models. More specifically, the attack revealed the embedding projection layer of a transformer model through API queries. Notably, the attack technique was originally proposed back in 2016 and has since been built upon to achieve the breaking of OpenAI and Google's AI models.

The team made Google and OpenAI aware of their infiltration, which both companies responded to by implementing mitigation techniques for that specific type of attack. Furthermore, the team decided not to publish their findings online, which would have been the exact size of OpenAI's GPT-3.5-turbo model, as this information was deemed harmful to the product since bad actors would learn aspects of the model, such as total parameter count, weight, size, etc.

Micron was the first to announce mass production of its new ultra-fast HBM3E memory in February 2024, seeing the company ahead of HBM rivals in SK hynix and Samsung... leaving its HBM competitors shocked.

The US memory company announced it would provide HBM3E memory chips for NVIDIA's upcoming beefed-up H200 AI GPU, which will feature HBM3E memory, unlike its predecessor with the H100 AI GPU, which featured HBM3 memory.

Micron will make its new HBM3E memory chips on its 1b nanometer DRAM chips, comparable to 12nm nodes, which HBM leader SK Hynix is using on its HBM. According to Korea JoongAng Daily, Micron is "technically ahead" of HBM competitor Samsung, which is still using 1a nanometer technology, which is the equivalent of 14nm technology.

Saudi Arabia robotics company QSS unveiled its humanoid robot called Mohammad at a premiere described as the "meeting place for the global artificial intelligence ecosystem." During the unveiling, Mohammad appeared to reach out and try to grab a female reporter's backside.

The DeepFest event in Riyadh was held last week and during the event a female reporter for Al Arabiya named Rawya Kassem, was standing in front of the humanoid robot talking to the audience. The above video shows the robot reaching its hand out with the goal of what appears to be touching the backside of Kassem. The reporter quickly moves back away from Mohammad raising her palm towards it before she continues to address the crowd.

It wasn't long before users on X began to accuse the humanoid robot of attempting to grope the reporter, but QSS has responded to the claims, telling Metro that Mohammad is built to help out in hazardous situations and may have been attempting to encourage Kassem to step further back on the stage to prevent falling all its edge. Additionally, QSS stated it conducted a thorough review of the footage and the circumstances surrounding the incident and found there were "no deviations from expected behavior".

Microsoft has revealed that it's holding an online event on March 21 where it'll announce some new AI features for Windows and Copilot, with new (unspecified) Surface devices to boot.

Given that the event is entitled a 'New Era of Work' that certainly hints at some pretty major introductions in the pipeline, and as Windows Latest reports, one of the AI features that's supposedly going to be teased is for Windows 11's Paint app.

As previously rumored, this might be a 'LiveCanvas' feature that's akin to the capability seen in Leonardo.Ai. This allows you to sketch a very rough picture, and an AI will fully realize it (and you can use text prompts to direct the AI further in its image composition).

SK hynix is reportedly increasing its spending on advanced chip packaging, where it wants to maintain its leadership in AI development with High Bandwidth Memory, or HBM.

In a new report from Bloomberg, the South Korean giant is investing more than $1 billion in South Korea this year to expand and improve the final steps of its chip manufacturing technology. Lee Kang-Wook, who leads up packaging development at SK hynix -- and was a former Samsung engineer -- said during an interview recently: "The first 50 years of the semiconductor industry has been about the front-end. But the next 50 years is going to be all about the back-end," or packaging.

Lee specializes in advanced ways of combining and connecting semiconductors, which have been the bedrock of the AI industry for the last few years. The executive started out in 2000, earning his PhD in 3D integration technology for micro-systems from Japan's Tohoku University under Mitsumasa Koyanagi, who was the man responsible for inventing stacker capacitor DRAM used in smartphones.

Analyst firm Omdia predicts NVIDIA could make $87 billion from its data center GPUs alone in 2024, calling the GeForce giant the "kingmaker."

NVIDIA has already lit up the stock market, bursting through the $2 trillion market capitalization milestone, and this year will only see them push harder into the AI market. NVIDIA has been leading the AI market with an estimated 90%+ market share with its AI GPU hardware, but next-generation AI GPUs are around the corner, and so are next-generation Blackwell-based GeForce RTX 50 series GPUs. "Kingmaker" sounds right.

Omdia's Cloud and Data Center Market Snapshot report for February is healthy for NVIDIA. Server and data center revenues in Q4 2023 were up 21.5% compared to Q3 2023 and 12.7% higher than they were in Q4 2022. In Q1 of 2023, the company had a server BOM of 15%, but that rocketed up to 44% in Q4 2023. This means that data centers are pushing more of their budgets into NVIDIA GPUs than ever before, and they're buying them quickly.

AMD has just announced its own localized and GPT-based LLM-powered AI chatbot, capable of running on Ryzen AI processors and Radeon RX 7000 series GPUs.

AMD's new LLM-based GPT chatbot can run on a bunch of different Ryzen AI platforms, including Ryzen 7000 and Ryzen 8000 series APUs that feature AMD's new XDNA NPUs, as well as Radeon RX 7000 series GPUs that pack AI accelerator cores.

The company published a new blog that helps you through the setup, so you can run your own localized chatbot powered by GPT-based LLMs (Large Language Models). If you've got a Ryzen AI processor, you'll need the standard LM Studio copy for Windows, while if you've got an RDNA 3-based Radeon RX 7000 series GPU, you'll need the ROCm Technical Preview.

Late last month, reports surfaced that the US military used artificial intelligence-powered algorithms to identify airstrike targets in the Middle East, according to a defense official.

The use of AI-powered technologies in warfare publicly began around 2017 when Project Maven was implemented, which was when the Pentagon put out a call for suppliers developing identifying object recognition software designed specifically to identify objects on drone footage. That same year, Marine Corps Colonel Drew Cukor said via press release that the Pentagon hoped to integrate that new object recognition software by the "end of the calendar year".

Now we are seemingly starting to hear about the impact that new technology has had on the success of missions. According to Schuyler Moore, CTO for US Central Command, the military deployed Project Maven's systems into real campaigns shortly after the Hamas attack on Israel last year, and now the technology has been used to carry out over 85 air strikes across seven locations in Iraq and Syria.

Microsoft is set to unveil next-gen Surface Pro and Surface Laptop hardware on March 21, the company's first wave of new AI PCs, ahead of next-gen Windows 11 AI features later this year.

The news is coming from Windows Central, which reports its sources said that Microsoft's new Surface Pro 10 and Surface Laptop 6 are expected to be unveiled on March 21. Inside, they'll feature Intel's latest Core Ultra "Meteor Lake" CPUs and Qualcomm's latest Snapdragon X Elite processors with next-gen NPUs (Neural Processor Units) for boosted AI performance.

Windows Central reports that these new processors will enable "huge performance and efficiency gains" over the previous-gen Surface Pro and Surface Laptops on the market. Both devices are posted to achieve "true all-day battery life and high-end performance capabilities". We're also told to expect other upgrades like new displays, higher-speed ports, and more.

Researchers are warning that it's only a matter of time before AI-powered malware, such as the "worm" they created, is discovered in the wild.

The team behind the new malware published a paper that is yet-to-be-peer-reviewed that details the creation of a new type of malware that targets AI-powered email assistants. The researchers conducted their experiment in a closed-circuit environment and found that their malware, or "worm," was capable of targeting email assistants powered by popular language models such as OpenAI's GPT-4, Google's Gemini Pro, and LLaVA.

The result was the worm infecting these email assistants, obtaining sensitive user information, and then sending out spam emails that can infect other PCs, replicating the same process. So, how does it work? The researchers used an "adversarial self-replicating prompt" that forces the targeted AI model to create another prompt within its response. Essentially, the target AI assistant receives a prompt, generates a response, and within that response is another prompt.

AMD has been stopped by a US government roadblock, where it was trying to sell cut-down AI GPUs to China, but stopped by the Commerce Department.

Bloomberg reports that AMD "hit a US government roadblock in attempting to sell an artificial intelligence chip tailored for the Chinese market," and once again, the news is coming from "people familiar with the matter." AMD was hoping it wouldn't have issues receiving approval to sell the AI GPU to Chinese customers from the Commerce Deparment, but they were denied, according to people who "asked not to be identified because the situation is private".

This purported AI GPU featured lowered performance to what AMD sells outside of China, and was designed to meet the required US export restrictions, they said. US officials told AMD that the AI GPU was still too powerful, and that the company needed to obtain a license from Commerce's Bureau of Industry and Security in order to sell it. Both AMD and the Bureau of Industry and Security declined to comment, so we don't know if AMD secured its license or not.

Meta CEO Mark Zuckerberg has said that he will be working with Samsung in the AI semiconductor sector, to lessen the reliance on using TSMC.

The news is coming from the South Korean presidential office, where Zuckerberg had a 30-minute meeting with President Yoon Suk Yeol in Seoul. Zuckerberg recognized Samsung for its status as one of the largest foundry companies in the world, with competitors like TSMC and Intel.

A senior official at Yoon's office told reporters: "Zuckerberg said Samsung's status can be the key point in their cooperation". The official added: "His remarks mean that Samsung's status can help in stabilizing Meta's heavy reliance on TSMC under the current geopolitical situation".