Close Menu
TechCentralTechCentral

    Subscribe to the newsletter

    Get the best South African technology news and analysis delivered to your e-mail inbox every morning.

    Facebook X (Twitter) YouTube LinkedIn
    WhatsApp Facebook X (Twitter) LinkedIn YouTube
    TechCentralTechCentral
    • News
      New tech retailer opens in South Africa, promising to undercut rivals - Amber Dendy Young

      New tech retailer opens in South Africa, promising to undercut rivals

      20 November 2026
      Meet the CIO | Wits's Stanley Mpofu on taking campus Wi-Fi to malls and libraries

      Meet the CIO | Wits’s Stanley Mpofu on taking campus Wi-Fi to malls and libraries

      20 November 2026
      Caby takes on Uber and Bolt in Cape Town - Denise Walker

      Caby takes on Uber and Bolt in Cape Town

      20 November 2026
      Tencent props up Naspers as iFood and Just Eat spending bites - Fabricio Bloisi

      Tencent props up Naspers as iFood and Just Eat spending bites

      20 November 2026
      MTN, Telkom and the price of consolidation

      MTN, Telkom and the price of consolidation

      19 November 2026
    • World
      Critical Java bug could let hackers take over servers

      Critical Java bug could let hackers take over servers

      19 November 2026
      Nvidia shuts the door on a Jensen Huang comeback

      Nvidia shuts the door on a Jensen Huang comeback

      17 November 2026
      Could Jensen Huang return to Nvidia?

      Jensen Huang weighs a comeback at Nvidia

      16 November 2026
      Intel sued by shareholders over 29% share price plunge - Intel CEO Lip-Bu Tan. Laure Andrillon/Reuters

      Intel sued by shareholders over 29% share price plunge

      16 November 2026
      Microsoft 364

      Microsoft outage hits the world

      13 November 2026
    • In-depth
      Inside the 24 hours that broke Microsoft

      Inside the 24 hours that broke Microsoft

      16 November 2026
      Ducking hell: inside Apple's hidden list of words the iPhone won't suggest

      Ducking hell: inside Apple’s hidden list of words the iPhone won’t suggest

      6 November 2026
      Why investors are souring on Netflix

      Why investors are souring on Netflix

      22 October 2026

      10 days that changed the course of AI

      21 September 2026
      Meta to the AI industry: slow down without us - Mark Zuckerberg

      Meta to the AI industry: slow down without us

      16 September 2026
    • TCS
      TCS | Piet le Roux on Sakeliga's fight with Icasa over BEE

      TCS | Piet le Roux on Sakeliga’s fight with Icasa over BEE

      13 November 2026
      W&W | The AA’s Bobby Ramagwede on a car industry running out of road

      W&W | The AA’s Bobby Ramagwede on a car industry running out of road

      12 November 2026
      TCS | Pick n Pay fights back in online grocery wars - Vince Viviers

      TCS | Pick n Pay fights back in online grocery wars

      11 November 2026
      TCS | Nomvuyiso Batyi on the battle over Sim registration - Nomvuyiso Batyi

      TCS | Nomvuyiso Batyi on the battle over Sim registration

      9 November 2026
      TCS+ | Old Mutual's Celiwe Ross on leading people through AI - Old Mutual

      TCS+ | Old Mutual’s Celiwe Ross on leading people through AI

      6 November 2026
    • Opinion
      Icasa's Sim rules need a privacy backstop - Lloyd Matthew and Heinz Vollmer

      Icasa’s Sim rules need a privacy backstop

      5 November 2026
      AI's boardroom honeymoon is over - Gary Stocks

      AI’s boardroom honeymoon is over

      22 October 2026
      Before South Africa licenses Starlink, it should look at Ukraine - Duncan McLeod

      Before South Africa licenses Starlink, it should look at Ukraine

      21 October 2026
      Nonkqubela Thathakahle Jordan-Dyani

      Why cybersecurity is now a national security issue for South Africa

      15 October 2026
      When a machine can choose, who does it become? Fanie van Rooyen

      When a machine can choose, who does it become?

      9 October 2026
    • Company News
      • 1Stream
      • Africa Data Centres
      • AfriGIS
      • Altron Digital Business
      • Altron Document Solutions
      • Altron Group
      • Arctic Wolf
      • Ascent Technology
      • AvertITD
      • BBD
      • Braintree
      • CallMiner
      • CambriLearn
      • CM.com
      • Contactable
      • CYBER1 Solutions
      • Digicloud Africa
      • Digimune
      • Domains.co.za
      • ESET
      • Euphoria Telecom
      • HOSTAFRICA
      • Incredible Business
      • iONLINE
      • IQbusiness
      • Iris Network Systems
      • Kaspersky
      • LSD Open
      • Mitel
      • NEC XON
      • Netstar
      • Network Platforms
      • Next DLP
      • Ovations
      • Paracon
      • Paratus
      • Publishared
      • Q-KON
      • SevenC
      • SkyWire
      • Solid8 Technologies
      • Telit Cinterion
      • Telviva
      • Tenable
      • Vertiv
      • Videri Digital
      • Vodacom Business
      • Vox
      • Wipro
      • Workday
      • XLink
    • Sections
      • AI and machine learning
      • Banking
      • Broadcasting and media
      • Cloud services
      • Contact centres and CX
      • Cryptocurrencies
      • Education and skills
      • Electronics and hardware
      • Energy and sustainability
      • Enterprise software
      • Fintech and payments
      • HealthTech
      • Information security
      • Internet and connectivity
      • Internet of things
      • Investment
      • IT services
      • Lifestyle
      • Policy and regulation
      • Public sector
      • Retail and e-commerce
      • Satellite communications
      • Science
      • SMEs and start-ups
      • Social media
      • Talent and leadership
      • Telecoms
      • Watts & Wheels
    • Events
    • Advertise
    TechCentralTechCentral
    Home » Sections » AI and machine learning » OpenAI chip rethink signals turning point in AI hardware market

    OpenAI chip rethink signals turning point in AI hardware market

    OpenAI is unsatisfied with some of Nvidia’s latest AI chips and has sought alternatives, sources have said.
    By Agency Staff3 February 2026
    Twitter LinkedIn Facebook WhatsApp Email Telegram Copy Link
    Get breaking news on WhatsApp

    OpenAI chip rethink signals turning point in AI hardware market - Sam Altman. Shelby Tauber/Reuters
    OpenAI CEO Sam Altman. Shelby Tauber/Reuters

    OpenAI is unsatisfied with some of Nvidia’s latest AI chips, and it has sought alternatives since last year, eight sources familiar with the matter said.

    The news potentially complicates the relationship between the two highest-profile players in the AI boom.

    The ChatGPT maker’s shift in strategy, the details of which are first reported here, is over an increasing emphasis on chips used to perform specific elements of AI inference, the process when an AI model such as the one that powers the ChatGPT app responds to customer queries and requests. Nvidia remains dominant in chips for training large AI models, while inference has become a new front in the competition.

    The deal had been expected to close within weeks. Instead, negotiations have dragged on for months

    This decision by OpenAI and others to seek out alternatives in the inference chip market marks a significant test of Nvidia’s AI dominance and comes as the two companies are in investment talks.

    In September, Nvidia said it intended to pour as much as US$100-billion into OpenAI as part of a deal that gave the chip maker a stake in the start-up and gave OpenAI the cash it needed to buy the advanced chips.

    The deal had been expected to close within weeks. Instead, negotiations have dragged on for months. During that time, OpenAI has struck deals with AMD and others for GPUs built to rival Nvidia’s. But its shifting product road map also has changed the kind of computational resources it requires and bogged down talks with Nvidia, a person familiar with the matter said.

    ‘Nonsense’

    On Saturday, Nvidia CEO Jensen Huang brushed off a report of tension with OpenAI, saying the idea was “nonsense” and that Nvidia planned a huge investment in OpenAI.

    “Customers continue to choose Nvidia for inference because we deliver the best performance and total cost of ownership at scale,” Nvidia said in a statement. A spokesman for OpenAI in a separate statement said the company relies on Nvidia to power the vast majority of its inference fleet and that Nvidia delivers the best performance per dollar for inference.

    Seven sources said that OpenAI is not satisfied with the speed at which Nvidia’s hardware can spit out answers to ChatGPT users for specific types of problems such as software development and AI communicating with other software. It needs new hardware that would eventually provide about 10% of OpenAI’s inference computing needs in the future, one of the sources said.

    Read: Bill Gates, OpenAI team up for AI health push in Africa

    The ChatGPT maker has discussed working with start-ups including Cerebras and Groq to provide chips for faster inference, two sources said. But Nvidia struck a $20-billion licensing deal with Groq that shut down OpenAI’s talks, one of the sources said.

    Nvidia’s decision to snap up key talent at Groq looked like an effort to shore up a portfolio of technology to better compete in a rapidly changing AI industry, chip industry executives said. Nvidia, in a statement, said that Groq’s intellectual property was highly complementary to Nvidia’s product road map.

    Nvidia CEO Jensen Juang
    Nvidia CEO Jensen Juang

    Nvidia’s graphics processing chips are well suited for massive data crunching necessary to train large AI models like ChatGPT that have underpinned the explosive growth of AI globally to date. But AI advancements increasingly focus on using trained models for inference and reasoning, which could be a new, bigger stage of AI, inspiring OpenAI’s efforts.

    The ChatGPT maker’s search for GPU alternatives since last year focused on companies building chips with large amounts of memory embedded in the same piece of silicon as the rest of the chip, called SRAM. Squishing as much costly SRAM as possible onto each chip can offer speed advantages for chatbots and other AI systems as they crunch requests from millions of users.

    Inference requires more memory than training because the chip needs to spend relatively more time fetching data from memory than performing mathematical operations. Nvidia and AMD GPU technology relies on external memory, which adds processing time and slows how quickly users can interact with a chatbot.

    Inside OpenAI, the issue became particularly visible in Codex, its product for creating computer code

    Inside OpenAI, the issue became particularly visible in Codex, its product for creating computer code, which the company has been aggressively marketing, one of the sources added. OpenAI staff attributed some of Codex’s weakness to Nvidia’s GPU-based hardware, one source said.

    In a 30 January call with reporters, Altman said that customers using OpenAI’s coding models will “put a big premium on speed for coding work”.

    One way OpenAI will meet that demand is through its recent deal with Cerebras, Altman said, adding that speed is less of an imperative for casual ChatGPT users.

    Competing products such as Anthropic’s Claude and Google’s Gemini benefit from deployments that rely more heavily on the chips Google made in-house, called tensor processing units, or TPUs, which are designed for the sort of calculations required for inference and can offer performance advantages over general-purpose AI chips like the Nvidia-designed GPUs.

    Groq talks

    As OpenAI made clear its reservations about Nvidia technology, Nvidia approached companies working on SRAM-heavy chips, including Cerebras and Groq, about a potential acquisition, the people said. Cerebras declined and struck a commercial deal with OpenAI, announced last month. Cerebras declined to comment.

    Groq held talks with OpenAI for a deal to provide computing power and received investor interest to fund the company at a valuation of roughly $14-billion, according to people familiar with the discussions. Groq declined to comment.

    Read: Nvidia’s next AI chips are in full production

    But by December, Nvidia moved to license Groq’s tech in a non-exclusive all-cash deal, the sources said. Although the deal would allow other companies to license Groq’s technology, the company is now focusing on selling cloud-based software, as Nvidia hired away Groq’s chip designers.  — Max A Cherney, Krystal Hu and Deepa Seetharaman, (c) 2026 Reuters

    Get breaking news from TechCentral on WhatsApp. Sign up here.

    Add TechCentral as a preferred source on GoogleFollow TechCentral on Google NewsGet breaking news on WhatsApp


    AMD ChatGPT Groq OpenAI Sam Altman
    WhatsApp YouTube
    Share. Facebook Twitter LinkedIn WhatsApp Telegram Email Copy Link
    Previous ArticleSpaceX acquires xAI in record-breaking deal
    Next Article Standard Bank branches are going cashless

    Related Posts

    Workday puts its AI agents on a leash - Aneel Bhusri

    Workday puts its AI agents on a leash

    19 November 2026
    Anthropic's $81-billion IPO is the biggest in history

    Anthropic’s $81-billion IPO is the biggest in history

    19 November 2026
    Why SA physicists think AI is far from a theory of everything

    Why SA physicists think AI is far from a theory of everything

    17 November 2026
    Company News
    2027: the year ID documents stop being enough

    2027: the year ID documents stop being enough

    19 November 2026
    Why banks' digital channels still leave customers stuck - BBD Software

    Why banks’ digital channels still leave customers stuck

    19 November 2026
    Q-Day is coming. Is your business ready? - Antony Russell, Telviva

    Q-Day is coming. Is your business ready?

    19 November 2026
    Opinion
    Icasa's Sim rules need a privacy backstop - Lloyd Matthew and Heinz Vollmer

    Icasa’s Sim rules need a privacy backstop

    5 November 2026
    AI's boardroom honeymoon is over - Gary Stocks

    AI’s boardroom honeymoon is over

    22 October 2026
    Before South Africa licenses Starlink, it should look at Ukraine - Duncan McLeod

    Before South Africa licenses Starlink, it should look at Ukraine

    21 October 2026

    Subscribe to Updates

    Get the best South African technology news and analysis delivered to your e-mail inbox every morning.

    Latest Posts
    New tech retailer opens in South Africa, promising to undercut rivals - Amber Dendy Young

    New tech retailer opens in South Africa, promising to undercut rivals

    20 November 2026
    Meet the CIO | Wits's Stanley Mpofu on taking campus Wi-Fi to malls and libraries

    Meet the CIO | Wits’s Stanley Mpofu on taking campus Wi-Fi to malls and libraries

    20 November 2026
    Caby takes on Uber and Bolt in Cape Town - Denise Walker

    Caby takes on Uber and Bolt in Cape Town

    20 November 2026
    Tencent props up Naspers as iFood and Just Eat spending bites - Fabricio Bloisi

    Tencent props up Naspers as iFood and Just Eat spending bites

    20 November 2026
    © 2009 - 2026 NewsCentral Media
    Built and maintained by Chronon
    • Cookie policy (ZA)
    • TechCentral – privacy and Popia

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

    Manage consent

    TechCentral uses cookies to enhance its offerings. Consenting to these technologies allows us to serve you better. Not consenting or withdrawing consent may adversely affect certain features and functions of the website.

    Functional Always active
    The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
    Preferences
    The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
    Statistics
    The technical storage or access that is used exclusively for statistical purposes. The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
    Marketing
    The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
    • Manage options
    • Manage services
    • Manage {vendor_count} vendors
    • Read more about these purposes
    View preferences
    • {title}
    • {title}
    • {title}
    🇿🇦 Sign up to the TechCentral newsletter