Skip to content
World clockEU--:--UK--:--USA--:--CN--:--PLDEFRIT中文EN

portal about AI and technologyevents · analysis · interviews · technical background

Search
LIVE
AnalysisDr Reddy's signs exclusive deal to distribute Takeda's Qdenga in India's private market06.11NewsDoubles players sign open letter opposing ATP's 2028 cuts to draws and pay06.11ExplainerTshisekedi asks Russia to back DR Congo mediation and UN demands06.11ExplainerGoncourt winner Philippe Jaenada says he used ChatGPT three times in his novel06.11ExplainerEuropean Commission plans €525m chip package for young companies06.11ExplainerArmenia proposes 12 November 2026 session on Trump Route company legislation06.11AnalysisUS midterm losses raise doubts over Trump's support for Milei before 2027 vote06.11AnalysisEni signs up to $1.8bn Cronos contracts for Cyprus's first gas project06.11AnalysisBanksy's Girl and Balloon on Found Landscape sells for US$17.94m at Fair Warning06.11ExplainerNHTSA closed its Jeep fire probe after FCA US recalled 1,076,999 Wranglers and Gladiators06.11
›

Oracle cancels $7bn Tencent chip lease as its AMD bet becomes the second-source test

Oracle cancelled a $7bn chip lease with Tencent after US officials raised concerns, Bloomberg reported on 4 November, citing the Financial Times. Oracle is now the clearest test of AMD, Google and inference startups selling hyperscalers an alternative to Nvidia.

Grace OkonkwoGrace OkonkwoAI, models and technology
TechnologyAnalysisGrace OkonkwoPublished: 4 November 20263 min readSources 24
Oracle cancels $7bn Tencent chip lease as its AMD bet becomes the second-source test

Oracle drops the Tencent chip lease

The FT reported last month that Oracle had signed a five-year agreement to lease 100,000 chips to Tencent. It was the Chinese firm's largest ever overseas deal. Bloomberg's 4 November report says Oracle has now cancelled the $7bn lease.

US officials told Oracle of their concerns about the deal, the FT said, citing two people familiar with the matter. Tencent did not immediately respond to a request for comment. The report did not say which chips the lease covered.

The cancellation is bigger than one contract. Export controls have so far focused on the sale of hardware. Performance thresholds decide what can ship. A lease of compute is a different kind of transaction. Officials appear to treat it the same way. Any cloud operator that rents capacity to a Chinese customer now carries political risk. That risk did not appear in the spreadsheet when the deal was signed.

Oracle is the biggest test for AMD

Oracle is also the biggest bet a hyperscaler has placed on AMD silicon. Startup Fortune reported that OCI committed to deploying 50,000 AMD Instinct MI450 GPUs starting in 2026. That is the largest single hyperscaler commitment to AMD to date. The same report described it as the first time a major cloud provider has offered a publicly accessible AI supercomputer built on AMD chips rather than Nvidia's. Oracle has said it plans to expand those deployments in 2027.

It is not an exclusive relationship. On 27 October Oracle said it will become the first hyperscaler to offer Google TPUs outside Google Cloud, with availability in 2027. OCI will also deploy Cerebras systems in 2027. Positron AI's Atlas inference chip is generally available now. Oracle will offer OpenAI's Jalapeño and SambaNova's SN50 chips.

Read together, the two moves describe a company buying options on the supply side. On the demand side, it faces pushback.

AMD's data centre number behind the orders

AMD's 3 November results explain why the Oracle order got the attention it did. Revenue was $13.4bn, up 45% from a year earlier. Data centre revenue nearly doubled to $8.5bn, which is 64% of everything AMD sells. Non-GAAP earnings per share came in at $2.09, up 74% year over year.

Nvidia's share is the comparison that keeps AMD honest. Nvidia still controls more than 90% of the AI accelerator market, according to industry estimates that Startup Fortune cited. Chief executive Lisa Su told investors AMD expects total revenue to grow more than 80% in 2027. She said data centre would be above 80% of sales by then.

What happens next

AMD's MI450 deployments at Oracle start in 2026, with expansion planned for 2027. Google TPUs arrive on OCI in 2027, with adoption ramping over a multi-year period.

None of those dates settles the price. Oracle has not put a value on its AMD commitment.

Comments 0

Sources

24
  1. 01Oracle Scraps $7 Billion Deal to Lease Chips to Tencent, FT SaysEN
  2. 02AMD's Data Center Revenue Nearly Doubles as AI Chip Demand AcceleratesEN
  3. 03Oracle Cloud to Offer Google TPUs and New AI ChipsEN
  4. 04Apple warns of ‘unprecedented’ rise in memory chip pricesEN
  5. 05Bosch gives chips their own division, eyeing €5bn in sales by 2030EN
  6. 06 AI chip smugglers exploit trade control loophole to bring banned HBM from South Korea into China EN
  7. 07UK’s most polluting power plant could power country’s largest AI datacentreEN
  8. 08Huawei tells customers to find their own AI chip memory, effectively outsourcing procurement to the black marketEN
  9. 09Chip Industry Technical Paper Roundup: Nov. 3EN
  10. 10Is There Room for Small-Scale OSATs in India’s Chip Packaging Boom?EN
  11. 11Samsung, SK Hynix poised for upside as Qualcomm advances LPDDR-based HBC AI chipEN
  12. 12Marvell Snaps Photonics Chipmaker Lumentum to Bolster AI ConnectivityEN
  13. 13Global AI Accelerator Market - 2026 Edition: Size, Share, Trends, and Forecast to 2030 | Generative AI Infrastructure ExpandsEN
  14. 14Scientists confirm global warming is accelerating faster than ever recordedEN
  15. 15Li Auto launches 2026 i6 with unchanged starting price, upgraded in-house chip and batteryEN
  16. 16China’s CXMT defends bet on risky new DRAM architecture to circumvent US chip curbsEN
  17. 17Chinese AI chipmaker Iluvatar CoreX says new GPU matches Nvidia's H200EN
  18. 18AI Chip Burn-in Test Market: Reliability Testing for AI Accelerators, GPUs, and High-Performance ChipsEN
  19. 19As AI cyber risks grow, Chinese chip expert urges shared ethics standards with US rivalsEN
  20. 20AI, fuelled by advanced chips, set to power semiconductor manufacturing: industry veteranEN
  21. 21FuriosaAI raises $620 mn in Series D at $2.9 bn valuation to scale AI chips - BeinsureEN
  22. 22Huawei’s Tau Scaling Law roll-out reaches base stations in blow to US chip curbsEN
  23. 23ByteDance accelerates robotics commercialisation with new recruitment driveEN
  24. 24South Korea urged to guard chip profitability as China’s CXMT and YMTC gain market shareEN

All figures and quotations in this text come from the sources listed below.

Content prepared by the editorial team with AI assistance.

Grace Okonkwo

Grace Okonkwo

AI, models and technology

Grace Okonkwo covers AI, models and technology for FLASH24, working from primary sources such as model cards, API documentation and benchmark papers rather than vendor summaries. She checks training data provenance, evaluation conditions and reported scores against the underlying datasets before any figure reaches print. She interviews researchers and engineers directly, tracks release calendars from major labs, and compares successive model versions on the same tests. Her own self-hosting, home-network and documentation-reading habits feed straight into that desk, since she tests tools on her own hardware first. She does not publish benchmark claims without a reproducible method.

Newsroom →

Comments

0
  1. No comments yet — be the first.

Write a comment

Comments are public. We do not publish abuse, spam or advertising.

Previous articlemSecure tells users to reset passkeys as synced-key attacks surfaceNext articleAdvanced packaging will cap AI chip growth in 2027, TechInsights says