AI Generated Image

Altman made the suggestion when a16z General Partner Erik Torenberg asked about how did he decide the timing of OpenAI’s collaboration since he had chose to strike deals and parnterships with companies like Advanced Micro Devices Inc. (AMD), Oracle Corp. and Nvidia Corp. in the past few week. A16z co-founder Ben Horowitz during the podcast said he was very impressed by the structure improvement of recent deals.  

“You should expect like much more from us in the coming months,” Altman told two a16z people, stressing that OpenAI management decided it is time to “make a very aggressive infrastructure bet.”  

While Altman has never been more confident in the research roadmap and the economic value that will be brought by AI models, he felt the company needs the support from the whole industry, or at least a major part of the industry, to make the bet at such scale. “And this is like, you know, from the level of like electrons to model distribution and all the stuff in between, which is a lot. And so we’re gonna partner with a lot of people,” he said.

OpenAI’s deals with Nvidia and AMD grabbed headlines these weeks for their potential to fuel the hype of AI infrastructure and escalate concerns that the AI boom is being lifted by interconnected business transactions. 

Nvidia and OpenAI on September 22 announced they agreed a letter of intent on a strategic partnership to deploy at least 10 gigawatts (GW) of Nvidia systems for OpenAI’s next-generation AI infrastructure to train and run its next generation of models on the path to deploying superintelligence, a hypothetical AI system that exceeds the powers of the human brain.

To support the planned deployment including data center and power capacity, Nvidia made its largest-ever spend commitment. It intends to invest up to $100 billion in OpenAI as the new Nvidia systems are deployed. The first phase is targeted to come online in the second half of 2026 using the Nvidia Vera Rubin platform.

Nvidia CEO Jensen Huang on September 22 said the 10GW is equal to between 4 million and 5 million GPUs,which is what Nvidia will ship in total this year and “twice as much as last year.” He also said the $100 billion investment is  "additive to everything that’s been announced and contracted" , hinting its beyond anything Nvidia has told Wall Street about its financial forecast.

AMD and OpenAI reached a 6 gigawatt (GW) agreement to power OpenAI’s next-generation AI infrastructure across multiple generations of AMD Instinct graphics processing units (GPUs).  The first 1 gigawatt deployment of AMD Instinct MI450 GPUs is set to begin in the second half of 2026, according to an OpenAI’s statement on Monday. 

Under the agreement, OpenAI will work with AMD as a core strategic compute partner to drive large-scale deployments of AMD technology starting with the AMD Instinct MI450 series and rack-scale AI solutions, and extending to future generations. By sharing technical expertise to optimize their product roadmaps, AMD and OpenAI are deepening their multi-generational hardware and software collaboration that began with the MI300X and continued with the MI350X series. 

OpenAI said the partnership creates a true win-win for both companies, enabling very large-scale AI deployments and advancing the entire ecosystem. 

As part of the agreement, AMD has issued OpenAI a warrant for up to 160 million shares of AMD common stock at $0.01 per share, which was structured to vest as specific milestones are achieved. The first tranche vests with the initial 1 GW deployment, with additional tranches vesting as purchases scale up to 6 GWs. Vesting is further tied to AMD achieving certain share-price targets and to OpenAI achieving the technical and commercial milestones required to enable AMD deployments at scale. 

If OpenAI exercises the full warrant, it could take over around 10% ownership in AMD, based on the current number of outstanding shares.

In an interview with CNBC on Wednesday, Huang said he’s surprised that AMD’s deal would allow OpenAI to obtain an up to 10% of stake in the Nvidia rival. “It’s imaginative, it’s unique and surprising, considering they were so excited about their next-generation product,” Huang said. “I’m surprised that they would give away 10% of the company before they even built it. And so anyhow, it’s clever, I guess.”

本文系作者 Li_Dan 授權(quán)鈦媒體發(fā)表,并經(jīng)鈦媒體編輯,轉(zhuǎn)載請(qǐng)注明出處、作者和本文鏈接。
本內(nèi)容來源于鈦媒體鈦度號(hào),文章內(nèi)容僅供參考、交流、學(xué)習(xí),不構(gòu)成投資建議。
想和千萬鈦媒體用戶分享你的新奇觀點(diǎn)和發(fā)現(xiàn),點(diǎn)擊這里投稿 。創(chuàng)業(yè)或融資尋求報(bào)道,點(diǎn)擊這里。

敬原創(chuàng),有鈦度,得贊賞

贊賞支持
發(fā)表評(píng)論
0 / 300

根據(jù)《網(wǎng)絡(luò)安全法》實(shí)名制要求,請(qǐng)綁定手機(jī)號(hào)后發(fā)表評(píng)論

請(qǐng) 登錄后輸入評(píng)論內(nèi)容

快報(bào)

更多

21:43

納比勒·法赫米將擔(dān)任新一任阿盟秘書長(zhǎng)

21:07

群核科技通過港交所聆訊:2025年實(shí)現(xiàn)盈利,沖刺“全球空間智能第一股”

21:05

油價(jià)飆升埃及多措施節(jié)能,埃及上調(diào)公共交通票價(jià)

21:05

英矽智能與禮來達(dá)成AI藥物研發(fā)合作,交易總值最高可達(dá)27.5億美元

21:04

3月29日新聞聯(lián)播速覽20條

21:01

中國(guó)銀河證券:中長(zhǎng)期仍看好科技板塊產(chǎn)業(yè)驅(qū)動(dòng)與周期板塊漲價(jià)線索的雙主線

20:34

大風(fēng)、降水來襲,長(zhǎng)江江蘇段部分區(qū)域?qū)嵤┡R時(shí)交通管制

19:59

國(guó)航C919正式投入北京—廈門、北京—哈爾濱兩條航線運(yùn)營(yíng)

19:57

中鋁國(guó)際:2025年歸母凈利潤(rùn)2.58億元,同比增長(zhǎng)16.47%

19:19

伊朗稱已打擊與美軍工有關(guān)聯(lián)的兩家企業(yè)

18:36

伊朗與巴基斯坦兩國(guó)外長(zhǎng)通電話,討論地區(qū)局勢(shì)

18:35

2025年玩具(不含潮玩)國(guó)內(nèi)市場(chǎng)零售總額達(dá)1035.3億元

18:08

全國(guó)豬價(jià)跌破5元,創(chuàng)歷史新低

18:07

時(shí)代天使2025年實(shí)現(xiàn)收入3.7億美元,同比增長(zhǎng)37.8%

17:57

電魂網(wǎng)絡(luò):擬投資4920萬元取得上海漫魂51%股權(quán)

17:51

美國(guó)1天超3000場(chǎng)反戰(zhàn)示威

17:28

雀巢12噸巧克力被盜

17:11

伊朗官員表示將“果斷反擊”美軍奪島

17:08

伊朗議長(zhǎng)稱武裝部隊(duì)“正等著美軍地面行動(dòng)”,并將“懲罰”其地區(qū)盟友

17:05

三安光電:董事長(zhǎng)及總經(jīng)理擬增持公司股份

掃描下載App