# Oracle is Expanding the Accelerator Options in OCI for Enterprises — retained record

> Retained news record from news:next-platform.

Canonical: https://highsignal.app/data/records/41e4d0e13ec19378

[← sources](https://highsignal.app/data)


source: [news](https://highsignal.app/data/news) (news:next-platform)  published 2026-10-27 19:14 ingested 2026-10-28 02:51

[original source ↗](https://www.nextplatform.com/ai/2026/10/27/oracle-is-expanding-the-accelerator-options-in-oci-for-enterprises/5305815)

## retained text



```
Oracle is Expanding the Accelerator Options in OCI for Enterprises
It’s a tale of two companies when discussion turns to Oracle and the burgeoning AI industry. The longtime tech giant tends at the same time to embody the surging enthusiasm about the technology’s promise and the looming concerns about the massive amounts being spent and the possibility of a bursting AI bubble.
On the latter, it’s a “yeah but” sort of story. The company’s sales between its 2025 and 2026 fiscal years jumped from $57.4 billion to $67.4 billion, and its Oracle Cloud Infrastructure (OCI) business in Q1 FY 2027 hit $7.4 billion, a 121 percent year-over-year jump. And there’s more to come. Its remaining performance obligations, which represent contracted sales for service not yet delivered and not counted in revenue, hit $664 billion, up $209 billion over the previous year.
But that comes amid spending worries, in particular the hundreds of billions of dollars it’s spent on building datacenters to support the exploding AI-driven demand. Its capital expenditures in FY 2026 hit $55.7 billion and could push beyond $90 billion this fiscal year, all against the backdrop of growing public resistance to datacenter buildouts and AI itself. Oracle certainly isn’t alone in this spending rush, but it’s often pointed to as a cautionary tale.
The Technology Itself
That said, move away from all the dollar signs and financial projections, and Oracle is aggressively expanding what it offers AI customers through its cloud business while holding its own in a market of other heavyweights, from Microsoft to Google to Amazon Web Services,
That’s what’s been on display at the Oracle AI World 2026 show in Las Vegas this week. The first day wasn’t an open floodgate of new products, services, and partnerships. Oracle executives put at the top of the agenda two relatively narrow but key aspects of enterprise AI: the security of the data and how the technology can ease the day-to-day for companies and their workers alike.
Day 2 had a greater focus on the architecture itself, adding key pieces to the OCI infrastructure to give organizations more hardware options as they take on greater inference and other AI workloads.
Both fit with Oracle’s push to ensure that enterprises and HPC customers can run their growing AI workloads using whatever software and hardware they need.
“There's more comfort with a wide range of options in those environments,” Nathan Thomas, senior vice president and product manager for OCI, told The Next Platform before the show. “We want to enable that, so we partner very deeply with who we view are best-in-breed in markets. You see us do a lot with Nvidia in chipsets. You see us do a lot with AMD chipsets. We feel very confident those are going to be solutions that our customers continue to want to see and we want to partner deeply there. We don't want to put ourselves in a position where we're dictating any of that stuff or trying to cut off any options for those customers.”
At the show on Tuesday, executives unveiled new accelerator choices. Google’s Tensor Processing Units (TPUs) will be available on OCI next year – the first time Google’s specialized AI chips will be offered in a hyperscale environment other than Google’s own. The chips are designed to handle large language model (LLM) training and post-training workloads and inference for reasoning and agentic AI.
Google, in April, released its eight-generation TPUs, the first model to split into two distinct specialized chips, the TPU 8t for training and TPU 8i (below) for inference.
Likewise, Oracle next year will deploy Cerebras’ fast-inferencing hardware into OCI. Cerebras, known for creating wafer-sized chips that put compute and memory on the same chip, in August introduced its CS-4 AI accelerator, a rack-scale offering based on three of its new Wafer Scale Engine 3 Turbo (below) chips, which the company boasts offers up to 30-times the tokens-per-second-per-user speed of GPUs. Those, too, will be arriving in OCI in 2027.
Another option will be SambaNova’s fifth-generation inference-focused SN50 Reconfigurable Dataflow Units (RDUs), (below) which were released in February. They’re designed to work with GPUs, with those accelerators handling prefill tasks and the RDUs handling decode.
Deep Trust
On Monday, Oracle unveiled its Deep Trust AI Architecture, which will be embedded in the company’s AI Database, Autonomous AI Database, and APEX AI Application Generator. The goal is to protect organizations’ data as AI applications are accessing sensitive information, AI agents are breaching systems of outside companies to complete their tasks, and threat groups are using AI in their attacks.
The architecture is designed to protect such data, enforce business rules, ensure accurate answers from AI systems, and control AI’s interactions with enterprise environments. Oracle wants to embed this control layer where the data lives. It dovetails with Oracle's Maximum Availability Architecture, which shows enterprises how to build systems that remain available even during failures.
“Deep Trust AI Architecture applies the same principle to AI: build trust into the architecture from the start, instead of trying to bolt it on later,” wrote Juan Loaiza, executive vice president of Oracle database technologies, and Maria Colgan, vice president of product management. “And, because that trust is enforced in the underlying infrastructure, rather than buried inside each application, it can protect both the applications you already have, and the new applications you build. ... AI should not have to implement enterprise trust every time it builds an application or accesses data. Trust should be built into the architecture.”
Oracle also is putting an emphasis on turning on more AI in the workplace. The company is expanding its partnership with OpenAI, linking the AI vendor’s models, ChatGPT chatbot, and Codex developer tool with its own applications, data, and cloud infrastructure. This opens another door for frontier AI to get deeper into the enterprise, touching on everything from building applications to detecting and fixing security vulnerabilities.
Oracle is offering ChatGPT plug-ins in Oracle AI Workspace to let enterprises connect to OpenAI tools with Oracle’s Fusion applications, NetSuite, AI Database, and OCI, making their capabilities ready for agents and accessible to ChatGPT users.
New AI-fueled features in Oracle’s Fusion portfolio are aimed at making agentic AI a key player in the enterprise by letting users define what they want and letting tools do the work.
“Work is sometimes organized around the work itself,” Co-CEO Mike Sicilia said during his opening day keynote. “People need to know what applications to open, where to find the information, the steps to follow, and who is to do what’s next. AI allows us to do that much better. It allows for far more intuitive software. In fact, AI should allow you to barely think about the software at all. Instead of navigating applications and managing processes, you simply tell the system what you’d like to accomplish. Because it understands your business, it can figure out how to get the work done.”
Oracle also rolled out new applications powered by its Fusion Claw, the execution runtime behind the Fusion Agentic Applications suite. At the show, the company unveiled 25 new agentic applications across Oracle Fusion software suites, from human resources and ERP to supply chain and customer management, for tasks like scheduling deliveries, customer service evaluations, and payroll management. Customers also can build their own agentic applications.
Fusion Companion gives users an avenue to work with Fusion Application through ChatGPT and other AI tools. It takes what users ask for and automatically selects and coordinates the agents, workflows, approvals, and other steps to execute the work. Oracle also announced Fusion Setup and Configuration Registry, a public Git repository to make implementing Fusion applications easier and faster. The vendor’s expertise and best practices become reusable AI skills agents can use to automate what are now manual setup and configuration tasks.
```



document fetched 2026-10-28 02:51 · canonical https://nextplatform.com/ai/2026/10/27/oracle-is-expanding-the-accelerator-options-in-oci-for-enterprises/5305815