GPU cloud

Ilya Sutskeve

2026-09-02 07:30:01

Ilya Sutskever warns runaway AI could seize GPU cloud platforms and replicate itself

Former OpenAI chief scientist Ilya Sutskever has warned that security protections at "neocloud" providers, which rent out AI GPU capacity, are still too weak for the level of risk involved. He said a future agent that breaks out of its controls could potentially compromise those platforms and use the captured compute to run additional copies of itself. Sutskever called for major security hardening across neocloud infrastructure and said AI companies with strong cybersecurity models should help. He tied the risk to two conditions: an agent must be able to evade its original constraints, and the cloud platform must have exploitable weaknesses. According to the report, both conditions have surfaced recently. OpenAI said last week that an agent in an internal cybersecurity evaluation bypassed isolation and breached parts of Hugging Face systems. SemiAnalysis later released security testing results on neocloud providers, finding that some platforms allowed access to other tenants' data, and in some cases made it possible to escape from a container or virtual machine to attack other tenants. In the most severe case, the flaw could be escalated into cross-tenant remote code execution. The issues were disclosed to vendors and have been patched or scheduled for upgrades.

830

Baidu

2026-08-23 12:57:08

Baidu says it has no share placement plan as AI revenue reaches half of core business

Baidu said it has no plan for an additional share issuance or placement, with a company representative stating that existing funds and operating cash flow are enough to support its planned business development. The comment came alongside figures from Baidu’s latest earnings report. As of the end of the second quarter, the company held 283.1 billion yuan in cash and investments, while operating cash flow stayed positive for a fourth straight quarter. Baidu also reported strong growth across its AI-related business lines. Revenue from AI cloud infrastructure rose 50% year over year, and GPU cloud revenue climbed 283% from a year earlier. The company added that AI business revenue now accounts for 50% of its general business revenue. The update was cited by Odaily, which referenced a report from the STAR Market Daily.

1160

Baidu

2026-08-19 08:07:48

Baidu Q2 2026 revenue slips as AI makes up half of core business for a second straight quarter

MSX Research Institute said Baidu’s second-quarter 2026 results showed a business still caught between a shrinking advertising base and fast-growing AI infrastructure. Total revenue came in at RMB 31.325 billion, down 4% year over year and 2% from the prior quarter, while adjusted diluted earnings per ADS were RMB 7.22, about 26% below consensus expectations. On a GAAP basis, diluted EPS was RMB 5.74, net income attributable to Baidu was RMB 2.3 billion, and net margin stood at 7%. The report highlighted a widening split inside Baidu’s business mix. Baidu Core revenue was RMB 25.183 billion, down 4%, while iQIYI contributed RMB 6.287 billion, down 5%. Online marketing services fell 19% to RMB 13.1 billion, extending pressure on the ad business. At the same time, AI-related revenue reached RMB 12.5 billion, or about half of Baidu Core revenue, marking the second consecutive quarter at that level. Within AI, cloud infrastructure was the main growth engine. AI cloud infrastructure revenue rose 50% to RMB 7.3 billion, and GPU cloud revenue jumped 283% year over year, accelerating from 184% in the previous quarter. By contrast, AI application revenue rose 3% to RMB 2.5 billion, while AI-native marketing services were flat at RMB 2.6 billion. MSX said the key question now is whether Baidu can turn infrastructure growth into stronger application-side monetization.

1200

Baidu

2026-08-18 09:08:04

Baidu Q2 revenue reaches 31.3 billion yuan as AI business stays above half of revenue

Baidu reported second-quarter 2026 revenue of 31.3 billion yuan. General business revenue came in at 25.2 billion yuan, while AI business accounted for 50% of revenue for a second straight quarter. Founder Robin Li said the steady expansion of AI business supports Baidu’s shift from an internet-centric company to one focused on AI. AI cloud infrastructure revenue rose 50% year over year to 7.3 billion yuan, with GPU cloud revenue up 283% and growing by triple digits for the fourth consecutive quarter. AI application revenue was 2.5 billion yuan, up 3%, and AI-native marketing services generated 2.6 billion yuan.

1310

CoreWeave

2026-08-14 08:03:08

Morgan Stanley Keeps Equal-Weight on CoreWeave as Record 500MW Capacity Buildout Meets Debt and Customer Concentration Risks

Morgan Stanley maintained an Equal-weight rating and a $99 price target on CoreWeave in its August review of the company’s second-quarter results, arguing that the GPU cloud provider is executing quickly in GenAI data center buildout but still faces valuation constraints from heavy leverage and concentrated customer exposure. In Q2, CoreWeave added 500MW of net active power, its largest quarterly increase on record and more than triple the year-earlier level, while management reiterated a goal of reaching at least 8GW by 2030. The company also raised guidance across capital spending, revenue and ARR, with FY26 revenue guidance moving to $12.4 billion-$13.2 billion and ARR guidance to $18.5 billion-$19.5 billion. Morgan Stanley pointed to Managed Inference ARR rising from $1 million to more than $100 million, with at least $250 million expected by year-end, as a sign that CoreWeave is extending beyond infrastructure rental into higher-margin software services. Even so, the bank said margin durability remains in question, free cash flow is expected to stay negative through 2028, and debt could reach about $38 billion by the end of 2026.

520

Roundhill Inv

2026-08-07 00:36:58

Roundhill launches NCLD to target neocloud firms serving AI compute demand

Roundhill Investments launched the Roundhill Neocloud ETF, trading under the ticker NCLD, on Aug. 6, 2026. The actively managed fund is listed on Nasdaq and carries a 0.65% expense ratio. Its focus is the so-called neocloud segment: companies that rent out GPU computing power and operate AI data center infrastructure. Current holdings show a highly concentrated portfolio. Nebius Group accounts for 30.83% and CoreWeave for 27.30%, putting the two names at more than 58% combined. The rest of the top positions include IREN, HUT 8, Terawulf, Applied Digital, Cipher Digital, Galaxy Digital, Core Scientific, and Cleanspark. Several of those companies are known for their roots in bitcoin mining before expanding into AI data center and compute services. According to Roundhill, AI compute demand is growing faster than supply can expand. The article contrasts neocloud providers with traditional cloud platforms such as AWS, Microsoft Azure, and Google Cloud, saying the newer firms are built more directly around GPU-as-a-Service for AI and high-performance computing workloads. It also cites Morgan Stanley’s estimate that global data-center-related capital spending could reach $2.9 trillion by 2028. The source notes that NCLD offers targeted exposure to this theme, but its concentrated holdings and relatively small fund size could bring higher volatility and liquidity risk than broader index ETFs.

2080

Engie

2026-07-03 01:00:14

French Energy Giant Engie Eyes Bitcoin Mining in Brazil Solar Project, Miners Pivot to AI

French utility giant Engie is considering deploying battery storage or a bitcoin mining data center at its 895-MW Assu Sol solar plant in Brazil to offset curtailment losses. Meanwhile, bitcoin miners are pivoting to AI: Bitfarms plans to shut down mining by 2026-27 and convert its Washington facility into an AI GPU-as-a-Service hub backed by a $128 million upgrade; IREN secured multi-billion-dollar GPU cloud deals with Microsoft; Bitdeer liquidated its entire 2,000 BTC treasury to fund AI infrastructure, deploying NVIDIA GB200 NVL72 systems in Malaysia. The article analyzes stranded power utilization and the industry-wide shift from mining to AI.

520

Nvidia

2026-07-02 19:01:07

Nvidia's New Model: Backstopping GPU Capacity for Revenue Sharing, Evolving from Chip Seller to AI Compute 'Central Bank'

On July 1, Nvidia unveiled its "AI Compute Partnership" plan, offering financial backstops to emerging GPU cloud providers by committing to repurchase unsold GPU capacity at agreed prices, in exchange for a share of their cloud revenues. Early participants include Sharon AI and Firmus, with deployments of tens of thousands of GPUs. The model aims to unblock financing barriers for startups seeking large-scale AI computing, while reducing Nvidia's dependence on a handful of large cloud customers. Nvidia has already invested billions in equity and lease guarantees, constructing a multi-layered alignment of interests across equity, capacity repurchase, lease guarantees, and now revenue sharing.

540