Skip to main content

Want to shove 30 GPUs in a computer system? Here's an AI solution that will work as long as you are using Dell — Liqid allows one R760 server to connect to a whopping 30 Nvidia GPUs for now with AMD and Intel likely soon

Web Hosting & Remote IT Support

Composable infrastructure firm Liqid has created a cost-effective and scalable solution for organizations looking to deploy high-performance inference capabilities in their data centers. 

Unveiled at Dell Technologies World 2024, the UltraStack 30 crams a whopping thirty Nvidia L40S GPUs into a single Dell PowerEdge R760 server. 

Liqid previously released a technical whitepaper for its UltraStack server at Nvidia GTC 2024 and at that time it had only managed to jam up to 20 Nvidia L40S into a PowerEdge server, so that's quite a leap in a short amount of time.

AMD and Intel support coming

Nvidia L40S GPUs have fewer cores and are generally less powerful than Nvidia’s H100 and A100 accelerators, as well as the newer Blackwell GB200 and B200 systems. However, they are well-suited for a variety of high-performance tasks, including AI, data analytics, and 3D rendering. This makes them a versatile choice for modern data centers and professional workloads, and they are cheaper too. L40S typically cost $7,000 to $8,000, compared to H100s which can go for $25,000 to $30,000. 

The UltraStack 30 offers enhanced enterprise-level GPU infrastructure capabilities and builds on previous solutions, with improvements in operations per second and total cost of ownership compared to similar GPU density solutions. It also boasts an energy efficient design capable of delivering high-performance while reducing power consumption and operational costs.

This is just the start, and the company has big plans for the UltraStack 30. Sumit Puri, Liqid co-founder and chief strategy officer, noted, “As we move forward, we recognize that optimal memory bandwidth will be crucial for maximizing performance. This is why composableCXL (Compute Express Link) is on the horizon at Liqid, promising to further enhance our capabilities in delivering high-performance, low-latency solutions.”

Blocks & Files adds, “Our understanding is that Liqid can support GPUs and other accelerators from suppliers such as AMD and Intel. It is also aware of the possibility of marketing its SmartStack and UltraStack technologies to MSPs that wish to supply GPUs for rent.”

More from TechRadar Pro



via Hosting & Support

Comments

Popular posts from this blog

Microsoft, Google, and Meta have borrowed EV tech for the next big thing in data centers: 1MW watercooled racks

Web Hosting & Remote IT Support Liquid cooling isn't optional anymore, it's the only way to survive AI's thermal onslaught The jump to 400VDC borrows heavily from electric vehicle supply chains and design logic Google’s TPU supercomputers now run at gigawatt scale with 99.999% uptime As demand for artificial intelligence workloads intensifies, the physical infrastructure of data centers is undergoing rapid and radical transformation. The likes of Google, Microsoft, and Meta are now drawing on technologies initially developed for electric vehicles (EVs), particularly 400VDC systems, to address the dual challenges of high-density power delivery and thermal management. The emerging vision is of data center racks capable of delivering up to 1 megawatt of power, paired with liquid cooling systems engineered to manage the resulting heat. Borrowing EV technology for data center evolution The shift to 400VDC power distribution marks a decisive break from legacy sy...

Google’s AI Mode can explain what you’re seeing even if you can’t

Web Hosting & Remote IT Support Google’s AI Mode now lets users upload images and photos to go with text queries The feature combines Google Gemini and Lens AI Mode can understand entire scenes, not just objects Google is adding a new dimension to its experimental AI Mode by connecting Google Lens's visual abilities with Gemini . AI Mode is a part of Google Search that can break down complex topics, compare options, and suggest follow-ups. Now, that search includes uploaded images and photos taken on your smartphone. The result is a way to search through images the way you would text but with much more complex and detailed answers than just putting a picture into reverse image search. You can literally snap a photo of a weird-looking kitchen tool and ask, “What is this, and how do I use it?” and get a helpful answer, complete with shopping links and YouTube demos. AI Eyes If you take a picture of a bookshelf, a plate of food, or the chaotic interior of your junk...

Passing the torch to a new era of open source technology

Web Hosting & Remote IT Support The practice of developing publicly accessible technologies and preventing monopolies of privately-owned, closed-source infrastructure was a pivotal technological movement in the 1990s and 2000s. The open source software movement was viewed at the time as a form of ‘digital civil duty’, democratizing access to technology. However, while the movement's ethos underpins much of today’s technological landscape, its evolution has proven to be a challenge for its pioneers. Hurdles Facing Young Developers Open source models successfully paved a path for the development of a multitude of technologies, cultivating a culture of knowledge sharing, collaboration , and community along the way. Unfortunately, monetizing such projects has always been a challenge, and ensuring contributors are compensated for their contributions working on them, even more so. On the other hand, closed-source projects offer greater control, security, and competitive advant...