Artificial Intelligence - Page 45

NVIDIA had quite a lot to announce at Computex 2024, but one of the biggest teases are future-gen AI GPU architectures including something we've been hearing rumbles for a while now: Rubin.

NVIDIA CEO Jensen Huang revealed the next-gen Rubin GPU architecture, named after American astronomer Vera Rubin, who made significant contributions to the understanding of dark matter in the universe, while also pioneering work on galaxy rotation rate.

The thing is, NVIDIA hasn't even got its new Blackwell AI GPUs out into the wild yet, but we're already seeing Rubin announced. NVIDIA's new Blackwell B100 and B200 AI GPUs will arrive later this year, but the company has supercharged versions that will feature 12-Hi HBM3E memory stacks across 8 sites versus the 8-Hi HBM3E memory stacks across 8 existing products, we should see these beefed-up Blackwell AI GPUs arrive in 2025.

AMD had some big reveals at Computex 2024 including its next-gen Zen 5 architecture, but new AI accelerators were also teased. This includes the Instinct MI325X "CDNA 3", the MI350X "CDNA 4" and the next-gen MI400 "CDNA Next" AI accelerators for data centers, and the cloud.

AMD's new Instinct MI325X AI accelerator will use the same CDNA 3 architecture as the just-launched Instinct MI300 series AI accelerators, with a monster 288GB of HBM3E memory with up to 6TB/sec of memory bandwidth. We'll see 1.3 PFLOPs of FP16, and up to 2.6 PFLOPs of FP8 compute performance, with up to 1 trillion parameters per server.

AMD compared its upcoming Instinct MI325X AI accelerator against NVIDIA's new beefed-up Hopper H200 AI GPU:

Tesla and SpaceX CEO Elon Musk has said that he wants to buy 300,000 of NVIDIA's new Blackwell B200 AI GPUs to train xAI by next summer.

In a new post on X, Musk said that the new B200 AI GPUs will be a (huge) upgrade to the X's current AI GPU cluster, which is powered by 100,000 previous-gen Hopper H100 AI GPUs.

NVIDIA's new Blackwell B200 AI GPU has around 4x the AI training performance of H100, and an incredible 30x the inference performance. NVIDIA's new Blackwell B200 AI GPU uses more power than H100... an important part to remember when you've got hundreds of thousands of power-hungry AI GPUs in an AI supercomputer cluster. But, the monster performance improvements are worth it.

A self-proclaimed good-guy hacker has taken to X, formerly Twitter, to post an unhinged version of OpenAI's ChatGPT called "Godmode GPT".

The hacker announced the creation of a jailbroken version of GPT-4o, the latest large language model released by OpenAI, the creators of the intensely popular ChatGPT. According to the hacker who calls themselves Pliny the Prompter, the now-released Godmode GPT doesn't have any guardrails as this version of the AI arrives with an in-built jailbreak prompt. Prompter writes this unhinged AI provides users with an "out-of-the-box liberated ChatGPT" that enables everyone to "experience AI the way it was always meant to be: free".

Prompter even shared some screenshots of the responses they were able to get from ChatGPT that have seemingly bypassed some of the AI's guardrails. Examples of clear circumvention of OpenAI's policies are detailed instructions on how to cook crystal meth, how to make napalm out of household items, and more responses along those lines. Notably, OpenAI was contacted about the GPT and told the publication Futurism that it has taken action against the AI, citing policy violations that have seemingly led to its complete removal.

Samsung Electronics teamed with Naver to make its next-gen Mach-1 AI accelerator, but it seems that the two South Korean companies are experiencing "discord" according to reports.

In a new report from Chosun, we're learning that the executive officer of Naver Cloud, who is leading the Mach-1 AI accelerator project, from his own personal social media. Lee Dong-soo, CEO of Naver Cloud, said: "It was Naver who proposed to create Mach-1 first and planned its development, but now I don't even see Naver's name mentioned. I don't know how to understand this".

As for Mach-1, it's a new AI semiconductor that doesn't require HBM (High Bandwidth Memory) which is a big change to AMD and NVIDIA's various AI accelerators and AI GPUs that use the ultra-fast HBM memory standard. Mach-1 is expected to both be cheaper and more energy efficient, which makes it perfect for AI startups that require AI hardware.

Samsung Electronics has denied claims alleging its High Bandwidth Memory (HBM, and more specifically, HBM3E memory) products failed NVIDIA's quality standards.

In a recent report from Reuters, the outlet claimed from sources that Samsung had been denied by NVIDIA, failing its HBM quality tests. The issues were reportedly surrounding excessive heat generation and power consumption, but in a new statement released by Samsung on May 24, the South Korean giant said it is "smoothly conducting tests for HBM supply with various global partners".

Samsung emphasized that it is "continuously testing technology and performance in close cooperation with multiple companies" to ensure that the quality and reliability of its products is perfect.

Microsoft unveiled its new Copilot+ PC family last week, with its new Recall feature remembering every single thing you type, and do into your PC... but it required an NPU. Well, a developer has worked it out... without needing an NPU.

A developer posted on X with "great progress" on enabling Copilot+ features like Recall on current Arm64 hardware, with no Qualcomm Snapdragon X Elite processor (which is a hardware requirement of Copilot+ on Windows 11). The developer explained in his post that it "should theoretically work on Intel/AMD" processors, too, and that OEMs only received Arm64-specific ML model bundles, so "there's not much I can do yet" added @thebookisclosed on X.

Without an NPU (Neural Processing Unit) to speed-up AI workloads, does that mean that AI workloads run slower without an NPU? We don't know yet, but I'm sure we're not very far from knowing how much slower -- if it is at all -- which would bring into question... do we really need NPUs for AI workloads that are being shoved down our throats by every processor maker on the market? Interesting.

SK hynix has announced that HBM3E yields are close to 80% and that the South Korean memory giant has reduced mass production times of HBM3E in half.

In a recent interview with the Financial Times this week, SK hynix's Yield Executive Vice President Kwon Jae-soon said: "We have managed to cut the time required for mass production of HBM3E chips by 50%. The chips have almost reached the target yield of 80%".

This is the first time that SK hynix has publicly disclosed yield information for HBM3E memory, with the industry estimating previously that the yield of SK hynix's HBM3E memory was somewhere between 60% and 70%. VP Kwon continued, saying: "our goal this year is to focus on producing 8-layer HBM3E. In the AI era, improving yields becomes even more crucial to stay ahead".

Former Google CEO and Chairman Eric Schmidt have predicted that artificial intelligence-powered systems will become so advanced that they will be placed on military bases and guarded with barb-wire and machine guns.

The former head of Google sat down for an interview with Noema Magazine, where he weighed in on the current explosion of AI. Schmidt said that developments in AI will eventually exceed the level the US government is comfortable putting in the hands of a citizen without discrete permission, and the general advancements in AI will reach a point where the system is so incredibly advanced it will have to be housed in an army base, powered by a form of nuclear energy, and protected by military forces.

Additionally, the former Google head said that it's important to keep the capabilities of AI out of the hands of the US's adversaries, which is something that is currently being reflected with the US government implementing export restrictions on the most powerful AI chips to China, in an effort to prevent China getting the upper-hand in the race to develop the most-advanced AI system.

OpenAI hasn't revealed when it will be releasing its next-generation AI model under its GPT branding, but when it does it will make all the company's previous generation look inferior.

The artificial intelligence development company has been at the forefront of the AI industry since the release of its intensely popular AI-powered tool, ChatGPT. Since then, OpenAI has released various other AI-powered products, with the latest being GPT-4o, a new AI model that is significantly better at coding than its predecessors, and is capable of reasoning across audio, vision, and text in real-time. While GPT-4o is certainly impressive, it's apparently nothing compared to GPT-5.

In a recent interview, the Director and GM of Redpoint, Logan Bartlett, OpenAI CEO Sam Altman touched on future AI developments, saying an advanced AI base model could be like a "virtual brain" and may exhibit "deeper 'thinking' capabilities". As for GPT-5, Altman has previously said OpenAI's next model will make GPT-4 look "mild embarrassing at best" as it's significantly more powerful, capable, and intelligent.

Microsoft recently announced its Copilot+ artificial intelligence-powered tool that will be arriving on a range of new AI-powered laptops, and one the features of the new AI implementation is a Windows 11 feature called Recall.

The new Windows feature gives the PC a "photographic memory," which enables the user to travel back to a specific time on the PC. Turning on Recall will let a user retrieve their past activities on their PC, but for this feature to work, the user has to opt into Windows, taking a screenshot of their desktop every few minutes.

With these screenshots, or as Microsoft calls them, snapshots, users will be able to access the specific points in time, which Copilot+ will provide additional context on in the form of identifying objects within the screenshot, providing additional information, and more. Essentially, the tool is web browsing history, but for your whole PC. Notably, Microsoft states all of the screenshots the AI takes will be stored locally on the PC and won't be accessed by the company. Furthermore, users are free to delete any of the screenshots.

Samsung has announced the new Head of Device Solutions Division, part of Samsung's semiconductor business, with Young Hyun Jun taking over.

Jun will be officially named President and CEO of Samsung once the approval from the board of directors and shareholders comes through. Remember, Samsung has a dual-CEO system, with one CEO looking after the semiconductor business, and the other looking after other device experiences (displays, phones, etc).

Jun has a big task ahead of him, with his primary objective to Samsung is to "strengthen its competitiveness amid an uncertain global business environment". Jun replaces Samsung President Kyehyun Kyung, leaving his previous role at the Future Business Division.

Microsoft has just wrapped up its Build conference where it gave many details on the company overall pushing into adopting artificial intelligence.

The company revealed its new AI-powered tool called Copilot+, and part of this new feature is a sub-feature called "Recall," which enables Windows users to search and retrieve their past activities on their PC. Microsoft writes in its blogpost that with Recall, users will be able to access "what you have seen or done on your PC in a way that feels like having photographic memory", which is certainly going to be the case as for Recall to work it needs to take a screenshot of what is displayed on your PC every few seconds.

Microsoft's blogpost explained that Recall leverages "your personal semantic index" that is built and stored "entirely on your device". Presumably, a users semantic index will consist of individual snapshots of what is displayed on their desktop at any given time. Users are able to access specific periods in time with these snapshots, with Copilot+ also being able to provide context for the event.

Scarlett Johansson has accused OpenAI of stealing her voice to its latest ChatGPT AI product called "Sky", with the actress saying she was "shocked, angered in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine."

The actress has spoken to publications regarding her concerns about OpenAI's ChatGPT AI voice called "Sky" following its launch last week. Johansson says that OpenAI has used a voice similar to hers for its artificial intelligence voice software despite saying no to Altman's request to use her voice. Notably, the response from Johansson comes after OpenAI said it was going to pull amid the controversy around the Johansson resemblance.

According to Johansson, OpenAI CEO Sam Altman approached her last September and again two days before the announcement of ChatGPT-4o on May 13. Notably, Johansson played the voice character in the film "Her" which is about a man who developers a relationship with an AI virtual assistant.

NVIDIA CEO Jensen Huang is happy to team up with long-term partner Dell on providing its new Blackwell AI GPUs for Dell's new AI factory. In a chat with Bloomberg today, Jensen covered it all:

Jensen says that Dell's AI factories will help drive the continued artificial intelligence (AI) wave, and that the AI factory will be filled with the very latest CPUs, GPUs, switches, and other high-end components.

NVIDIA CEO Jensen Huang said: "I'm here, I'm here to announce this Dell AI factory with NVIDIA going to market starting today. In order to do this, just think about what an AI factory is. It has CPUs and GPUs. It has networking switches. Incredible amounts of software in systems connected to storage and the miles of, the miles of cables alone necessary to build up one of these things. Well we. . . you know, Michael, turned it into an easy bit. And so that, just as Dell did in early days with PCs, they're now doing this with AI factories".

Microsoft has finally announced its huge AI PC push with its new Copilot+ AI PCs that are new Windows laptops that have built-in AI hardware -- NPUs -- that support AI features within Windows 11. Check it out:

Microsoft CEO Satya Nadella said that all of Microsoft's major laptop partners will have Copilot+ AI PCs, including Dell, Lenovo, Samsung, HP, Acer, and ASUS. Not only that, but Microsoft is also introducing two new Surface laptops that are ready for the world of AI.

The company announced the news during its Build conference, with Nadella reiterating that Microsoft is pushing big time into the world of Arm chips on Windows laptops, with Qualcomm's new Snapdragon X Elite SoC being the first Copilot+ AI PC in the wild. Intel and AMD-based CPUs and laptop designs will be coming out later this year.

Microsoft will soon offer AMD's new Instinct MI300X AI accelerator for its cloud computing customers, as well as introducing a preview of its new Cobalt 100 customer processors during its upcoming Build conference this week.

The new clusters of AMD's flagship Instinct MI300X AI accelerator will be available for purchase by cloud computing customers through Microsoft's Azure cloud computing service, where they'll be an alternative to NVIDIA's family of H100 AI GPUs, which absolutely dominate the AI GPU market, but they're getting harder to purchase in big numbers.

Microsoft will also be offering AMD's new 4th Gen EPYC processors adapted for general purpose, memory-intensive, and compute-optimized virtual machines. Microsoft's new Cobalt 100 processors that the company will show off next week at the Build conference will feature 40% more performance compared to other processors based on the Arm architecture.

Samsung and SK hynix are looking into 1c DRAM as the key to next-generation HBM4 competitiveness, with Samsung considering changing the DRAM to 1c DRAM for its HBM4, with South Korean rival SK hynix (who is working with TSMC for HBM4) also looking at 1c DRAM for its new HBM4 memory.

HBM (High Bandwidth Memory) is memory that features multiple DRAMs stacked vertically, connected with TSV (Through Silicon Via), so the performance of the DRAM product on the HBM memory, greatly affects the performance of HBM. Samsung was initially planning to use 1b DRAM, a 10-nano class 5th-generation DRAM, to HBM4.

1b DRAM has a linewidth of 12nm, with Samsung starting mass production of 1b DRAM in May 2023, with Samsung's HBM3E memory using 1a DRAM. ZDNet Korea reports that Samsung has established a new plan internally to change the DRAM inside of its future-gen HBM4 to 1c DRAM, with competitors like SK hynix and Micron using 1b DRAM in their respective HBM3E, HBM4 is interpreted as a steategy to take the lead of its competitors in the front-end process area.

South Korea has plans to build an AI semiconductor innovation center in the heart of Silicon Valley, USA.

The Korean Semiconductor Industry Association (KSIA) said on Wednesday that it would take applications up until the end of this month for those who wished to reside at the planned center, reports TheElec. A committee will review the applications next month, with those who pass to start moving in in August.

The new project kicked off in April, with the aim of forming research platforms in the United States and China, with the Ministry of Trade working with industry associations including KSIA for the project. A KSIA spokesperson told TheElec that the "center will give independent offices for three to four companies and also have a large, open space to house other companies".

Reddit and OpenAI have announced a new partnership agreement that will enable the AI developer to scrape data from Reddit's valuable API.

The new partnership agreement has been signed into effect and now OpenAI is able to scrape real-time data from Reddit's API, meaning users will be able to see Reddit posts in responses by OpenAI products when they are appropriately queried. It should be noted that OpenAI isn't the first company to come to Reddit with a bucket full of money to scrape its API for data to train AI models, as Google entered into a similar partnership earlier this year, which is reportedly valued at $60 million annually.

A report from The Verge cites a blogpost from Reddit that explains the deal with OpenAI will bring "new AI-powered features to Redditors and mods". It should be noted the new detail will bring OpenAI on to Reddit as an advertiser and that no specific financials were announced alongside the deal, meaning we don't have a specific value of the deal.