KAYTUS Unveils Upgraded MotusAI to Accelerate LLM Deployment
SINGAPORE--(BUSINESS WIRE)-- KAYTUS, a leading provider of end-to-end AI and liquid cooling solutions, today announced the release of the latest version of its MotusAI AI DevOps Platform at ISC High Performance 2025. The upgraded MotusAI platform delivers significant enhancements in large model inference performance and offers broad compatibility with multiple open-source tools covering the full lifecycle of large models. Engineered for unified and dynamic resource scheduling, it dramatically improves resource utilization and operational efficiency in large-scale AI model development and deployment. This latest release of MotusAI is set to further accelerate AI adoption and fuel business innovation across key sectors such as education, finance, energy, automotive, and manufacturing.
As large AI models become increasingly embedded in real-world applications, enterprises are deploying them at scale, to generate tangible value across a wide range of sectors. Yet, many organizations continue to face critical challenges in AI adoption, including prolonged deployment cycles, stringent stability requirements, fragmented open-source tool management, and low compute resource utilization. To address these pain points, KAYTUS has introduced the latest version of its MotusAI AI DevOps Platform, purpose-built to streamline AI deployment, enhance system stability, and optimize AI infrastructure efficiency for large-scale model operations.
Enhanced Inference Performance to Ensure Service Quality
Deploying AI inference services is a complex undertaking that involves service deployment, management, and continuous health monitoring. These tasks require stringent standards in model and service governance, performance tuning via acceleration frameworks, and long-term service stability, all of which typically demand substantial investments in manpower, time, and technical expertise.
The upgraded MotusAI delivers robust large-model deployment capabilities that bring visibility and performance into perfect alignment. By integrating optimized frameworks such as SGLang and vLLM, MotusAI ensures high-performance, distributed inference services that enterprises can deploy quickly and with confidence. Designed to support large-parameter models, MotusAI leverages intelligent resource and network affinity scheduling to accelerate time-to-launch while maximizing hardware utilization. Its built-in monitoring capabilities span the full stack—from hardware and platforms to pods and services—offering automated fault diagnosis and rapid service recovery. MotusAI also supports dynamic scaling of inference workloads based on real-time usage and resource monitoring, delivering enhanced service stability.
Comprehensive Tool Support to Accelerate AI Adoption
As AI model technologies evolve rapidly, the supporting ecosystem of development tools continues to grow in complexity. Developers require a streamlined, universal platform to efficiently select, deploy, and operate these tools.
The upgraded MotusAI provides extensive support for a wide range of leading open-source tools, enabling enterprise users to configure and manage their model development environments on demand. With built-in tools such as LabelStudio, MotusAI accelerates data annotation and synchronization across diverse categories, improving data processing efficiency and expediting model development cycles. MotusAI also offers an integrated toolchain for the entire AI model lifecycle. This includes LabelStudio and OpenRefine for data annotation and governance, LLaMA-Factory for fine-tuning large models, Dify and Confluence for large model application development, and Stable Diffusion for text-to-image generation. Together, these tools empower users to adopt large models quickly and boost development productivity at scale.
Hybrid Training-Inference Scheduling on the Same Node to Maximize Resource Efficiency
Efficient utilization of computing resources remains a critical priority for AI startups and small to mid-sized enterprises in the early stages of AI adoption. Traditional AI clusters typically allocate compute nodes separately for training and inference tasks, limiting the flexibility and efficiency of resource scheduling across the two types of workloads.
The upgraded MotusAI overcomes traditional limitations by enabling hybrid scheduling of training and inference workloads on a single node, allowing for seamless integration and dynamic orchestration of diverse task types. Equipped with advanced GPU scheduling capabilities, MotusAI supports on-demand resource allocation, empowering users to efficiently manage GPU resources based on workload requirements. MotusAI also features multi-dimensional GPU scheduling, including fine-grained partitioning and support for Multi-Instance GPU (MIG), addressing a wide range of use cases across model development, debugging, and inference.
MotusAI's enhanced scheduler significantly outperforms community-based versions, delivering a 5× improvement in task throughput and 5× reduction in latency for large-scale POD deployments. It enables rapid startup and environment readiness for hundreds of PODs while supporting dynamic workload scaling and tidal scheduling for both training and inference. These capabilities empower seamless task orchestration across a wide range of real-world AI scenarios.
About KAYTUS
KAYTUS is a leading provider of end-to-end AI and liquid cooling solutions, delivering a diverse range of innovative, open, and eco-friendly products for cloud, AI, edge computing, and other emerging applications. With a customer-centric approach, KAYTUS is agile and responsive to user needs through its adaptable business model. Discover more at KAYTUS.com and follow us on LinkedIn and X.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Verge
29 minutes ago
- The Verge
PCIe 7.0 is coming, but not soon, and not for you
The PCIe 7.0 specification has now been released, while many of us are still waiting for PCIe 6.0 to materialize in consumer products. The PCI Special Interest Group (PCI-SIG) announced on Wednesday that PCIe 7.0 is now available to members of its organization, delivering a theoretical maximum bandwidth speed of 512GB per second in both directions, across a x16 connection. 'PCIe technology has served as the high-bandwidth, low-latency IO interconnect of choice for over two decades and we are pleased to announce the release of the PCIe 7.0 specification, which continues our long-standing tradition of doubling the IO bandwidth every three years,' PCI-SIG President Al Yanes said in the announcement. 'As artificial intelligence applications continue to scale rapidly, the next generation of PCIe technology meets the bandwidth demands of data-intensive markets deploying AI, including hyperscale data centers, high performance computing (HPC), automotive, and military/aerospace.' You may have noticed that consumer computing devices weren't included in that statement — the specification is targeting data-driven applications like cloud and quantum computing datacenters for now, and will take some time to even appear in those markets. PCI-SIG says that PCIe 7.0 will be backward compatible with previous PCI Express versions, but there's no mention of plans to bring it to everyday desktop SSDs or GPUs any time soon. That shouldn't be surprising, given the PCIe 5.0 spec that launched in 2019 only started trickling into consumer hardware two years ago, and is still fairly uncommon. Image: PCI-SIG
Yahoo
29 minutes ago
- Yahoo
Cynthia Lummis Proposes Artificial Intelligence Bill, Requiring AI Firms to Disclose Technicals
Senator Cynthia Lummis (R-WY) has introduced the Responsible Innovation and Safe Expertise (RISE) Act of 2025, a legislative proposal designed to clarify liability frameworks for artificial intelligence (AI) used by professionals. The bill could bring transparency from AI developers – stoping short of requiring models to be open source. In a press release, Lummis said the RISE Act would mean that professionals, such as physicians, attorneys, engineers, and financial advisors, remain legally responsible for the advice they provide, even when it is informed by AI systems. At the time, AI developers who create the systems can only shield themselves from civil liability when things go awry if they publicly release model cards. The proposed bill defines model cards as detailed technical documents that disclose an AI system's training data sources, intended use cases, performance metrics, known limitations, and potential failure modes. All this is intended to help help professionals assess whether the tool is appropriate for their work. "Wyoming values both innovation and accountability; the RISE Act creates predictable standards that encourage safer AI development while preserving professional autonomy,' Lummis said in a press release. 'This legislation doesn't create blanket immunity for AI," Lummis continued. However, the immunity granted under this Act has clear boundaries. The legislation excludes protection for developers in instances of recklessness, willful misconduct, fraud, knowing misrepresentation, or when actions fall outside the defined scope of professional usage. Additionally, developers face a duty of ongoing accountability under the RISE Act. AI documentation and specifications must be updated within 30 days of deploying new versions or discovering significant failure modes, reinforcing continuous transparency obligations. The RISE Act, as it's written now, stops short of mandating that AI models become fully open source. Developers can withhold proprietary information, but only if the redacted material isn't related to safety, and each omission is accompanied by a written justification explaining the trade secret exemption. In a prior interview with CoinDesk, Simon Kim, the CEO of Hashed, one of Korea's leading VC funds, spoke about the danger of centralized, closed-source AI that's effectively a black box. "OpenAI is not open, and it is controlled by very few people, so it's quite dangerous. Making this type of [closed source] foundational model is similar to making a 'god', but we don't know how it works," Kim said at the time.
Forbes
34 minutes ago
- Forbes
AI Is Giving You Back Half A Day Every Week. How To Use It Wisely.
AI Is Giving You Back Half a Day Every Week If you're using AI at work, even occasionally, you may already be gaining back a valuable resource: time. Microsoft's Copilot study found that users spent 30 minutes less on email each week and completed documents 12 percent faster. Adecco reports time savings of around five hours a week for knowledge workers. Thomson Reuters says four. These small efficiencies often amount to a hidden dividend of roughly half a working day every single week. That figure is supported directionally, if more cautiously, by the Federal Reserve Bank of St. Louis. Their analysis estimates that AI saves an average of 5.4% of their weekly work hours, which is about 2.2 hours every week. Even taking the lowest estimate, the result is still meaningful: reclaimed time, quietly reshaping the workweek. I describe the rising AI flood in earlier Forbes pieces and my book, The Human Edge. Some jobs are being submerged. But within the remaining roles, AI is leaving behind micro-efficiencies: small bursts of reclaimed time that quickly add up. This is not speculative. It's already happening. That landmark Federal Reserve survey of nearly 10,000 people found that: Even the recent ChatGPT outage caused over 500,000 Google searches in hours. That wasn't hype. This tells us something important: AI has already woven itself into the fabric of modern work. Over two decades of leadership development, I've worked with CEOs, scientists, creatives, and entrepreneurs. Different sectors, different goals, but the same underlying refrain: "I'd love to reflect, learn, or think more strategically…but I don't have time." Fair enough. Work is relentless. But AI is shifting the equation. The real question is this: Here are five high-leverage moves to make the most of your AI-liberated time. Here's my prediction. The AI age doesn't just reward those who move fast. It rewards those who use time differently.
