# Together AI Boosts AI Inference with IBM Cloud and NVIDIA B300 GPUs > Together AI's new partnership with IBM and NVIDIA unleashes cutting-edge AI inference capabilities, marking a pivotal shift towards open AI models. This collaboration promises unprecedented performance, enabling enterprises to harness the future of AI confidently. **Source**: together.ai | **Published**: 2026-10-07 | **Type**: article ## Key Facts - Together AI's use of NVIDIA B300 GPUs on IBM Cloud marks a competitive edge in AI inference. - Serving hundreds of trillions of tokens monthly indicates soaring demand for open AI models. - Collaboration enhances enterprise reliability, revealing vulnerabilities in closed AI model providers. - Cost-effective open models provide financial advantages, driving enterprise adoption and growth. - Strategic shift towards open-source AI suggests a market trend favoring transparency and scalability. ## Summary Together AI has announced a significant partnership with IBM and NVIDIA to enhance enterprise-grade AI inference capabilities. This collaboration marks the launch of a dedicated cluster of NVIDIA B300 GPUs on IBM Cloud, specifically designed for high-performance inference tasks. As the first customer utilizing this infrastructure, Together AI aims to meet the surging demand for AI services, which is expected to escalate dramatically in the coming years. The partnership leverages the strengths of each company: NVIDIA supplies the advanced hardware with its B300 GPUs and Spectrum-X Ethernet networking, while IBM provides its robust cloud infrastructure, known for supporting mission-critical applications. Together AI focuses on operating the inference layer, creating a comprehensive solution that promises to deliver high-throughput performance for open AI models. This strategic alignment is crucial as enterprises increasingly seek alternatives to proprietary AI models, driven by concerns over data sovereignty and cost efficiency. The demand for open AI models is rising sharply, with Together AI reporting that it serves hundreds of trillions of tokens each month to over a million developers. This trend highlights a growing preference among enterprises and AI-native companies for open-source solutions that allow them to maintain control over their data while benefiting from cutting-edge performance at a lower cost. The collaboration between Together AI, IBM, and NVIDIA is poised to capitalize on this shift, offering a scalable and reliable platform for AI deployment. The implications of this partnership extend beyond immediate operational capabilities. It signals a broader trend in the AI market where open-source solutions are gaining traction against traditional closed models. As enterprises increasingly prioritize flexibility and control, the demand for infrastructure that supports open AI models is likely to grow. This shift could challenge established players in the AI space, compelling them to adapt their offerings or risk losing market share. Moreover, the collaboration emphasizes the importance of reliability and security in AI deployments. As organizations integrate AI into their workflows, the need for robust guardrails becomes paramount. The combination of NVIDIA's cutting-edge hardware and IBM's experience in enterprise infrastructure positions Together AI to meet these demands effectively. This focus on security and performance will be a critical differentiator in a competitive landscape where trust and reliability are essential for enterprise adoption. Looking ahead, the partnership may set a precedent for future collaborations in the AI sector. As companies recognize the advantages of combining specialized capabilities—such as hardware, cloud infrastructure, and application development—more alliances could emerge. This trend may lead to the establishment of new standards for AI deployment, making it imperative for competitors to innovate continuously. The trajectory of AI development is shifting towards open models that can operate at scale and speed, matching the capabilities of closed systems. Together AI's initiative signals a pivotal moment in the market, where the integration of advanced technology and strategic partnerships will define the future of enterprise AI. As demand for these solutions continues to grow, the ability to deliver reliable, scalable, and secure AI inference will be a key determinant of success in the evolving landscape. ## Entities - **Companies**: Together AI, IBM, NVIDIA - **Products**: NVIDIA B300 GPUs, Spectrum-X Ethernet ## Key Concepts enterprise-grade AI inference, open models, NVIDIA B300 GPUs, IBM Cloud, collaboration, high-throughput inference, sovereign data, production grade tokens ## Definitions - **enterprise-grade AI inference**: A robust AI inference capability designed to meet the demands of large-scale enterprise applications. - **open models**: AI models that allow users to maintain control over their data and provide cost-effective performance. - **NVIDIA B300 GPUs**: High-performance graphics processing units designed for AI inference tasks. - **Spectrum-X Ethernet**: A networking technology developed by NVIDIA to support high-throughput data transfer for AI applications. - **production grade tokens**: Units of processing or data that meet the reliability and performance standards required for enterprise use. ## Use Cases - scaling enterprise-grade AI inference - running open models in production - serving tokens to developers - supporting high-throughput inference - ensuring data sovereignty - enhancing AI workload performance ## Frequently Asked Questions **What is the significance of the collaboration between Together AI, IBM, and NVIDIA?** This collaboration aims to scale enterprise-grade AI inference by leveraging NVIDIA's advanced GPU technology and IBM's cloud infrastructure. It represents a significant step towards making open-source AI models more accessible and efficient for enterprises. **How does the NVIDIA B300 GPU enhance AI inference?** The NVIDIA B300 GPU is engineered specifically for high-throughput inference, allowing for faster processing and improved performance in AI applications. This capability is crucial for handling the increasing demand for AI services. **What are open models, and why are they important?** Open models are AI frameworks that allow organizations to retain control over their data while benefiting from advanced AI capabilities. They are important because they provide enterprises with cost-effective solutions without compromising data sovereignty. **What role does IBM play in this collaboration?** IBM provides the cloud infrastructure necessary for running the AI inference operations. Their experience in managing mission-critical systems for large enterprises ensures reliability and scalability in the deployment of AI solutions. **What future plans does Together AI have regarding AI inference?** Together AI plans to expand its capabilities as token demand increases, aiming to enhance the efficiency and performance of open models in production environments. This includes scaling their infrastructure to meet growing enterprise needs. ## Links - [Read on Welcome.AI](https://welcome.ai/content/together-ai-boosts-ai-inference-with-ibm-cloud-and-nvidia-b300-gpus) - [Original source](https://www.together.ai/blog/expanding-our-enterprise-inference-capacity-with-ibm-cloud-and-nvidia) --- Source: Welcome.AI | https://welcome.ai/content/together-ai-boosts-ai-inference-with-ibm-cloud-and-nvidia-b300-gpus