LugInvest (@LugInvest) header
LugInvest (@LugInvest) profile picture

Investing | Distressed stocks | Value & Small Cap Investor | Stock Markets | Economics | Banking | Emerging Markets | IT & Technology

Munich
Joined February 2014
Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)Photo by LugInvest (@LugInvest)
AI inference is scaling. Memory needs to keep up. Marvell and @WeAre_Penguin, the AI Factory Platform Company, are collaborating to address one of AI infrastructure's growing challenges: memory capacity, bandwidth and latency. The collaboration brings together Marvell® Structera™ X memory-expansion controllers, designed to extend server memory capacity by reusing existing DDR4 memory, and Marvell® Photonic Fabric™ optical interconnect technology, designed to enable high-bandwidth, low-latency memory sharing across multiple racks. Together, these complementary technologies offer a roadmap from direct-attach memory expansion to disaggregated, multi-rack sharing, helping data centers scale memory as AI inference demands evolve. Discover how Marvell and Penguin Solutions are advancing memory expansion from the server to the pod: mrvl.co/4eH5iN3
4
10
4
63
4,413
Announcing the availability of our new ComputeAI NML72-VR200: a pre-validated, liquid-cooled rack-scale platform built on the @NVIDIA Vera Rubin NVL72 architecture. Designed as a deployment-ready foundation for agentic AI, inference-time scaling, and gigascale model development, this fully integrated system brings together 72 NVIDIA Rubin GPUs, 36 NVIDIA Vera CPUs, sixth-generation NVIDIA NVLink, ConnectX-9 networking, and BlueField-4 DPUs. As large-scale AI workloads push beyond conventional multi-GPU server capabilities, our validated configuration: 🔹 reduces integration risk 🔹 accelerates time to production 🔹 simplifies operations for enterprises, AI cloud providers, and sovereign AI initiatives. Stop by booth #1723 at #SC26 to see how we’re expanding our Full-Stack AI Factory Platform to power the next generation of AI applications. ir.penguinsolutions.com/news… #PenguinSC26 @Supercomputing
2
5
1
24
891
🚨 MAJOR VALIDATION: $OSS & Leonardo DRS $DRS Co-Brand Integrated Edge AI System Live at SC26! 🛡️❄️ Don't scroll past this trade show post! This is a massive milestone for @_OneStopSystems. At Supercomputing 2026, OSS and Tier-1 defense giant Leonardo DRS are showcasing an integrated, co-branded rugged compute solution live at Booth #2824. What was previously behind-the-scenes supply chain work is now an officially demonstrated technology partnership. 1. The Technical Division of Labor: 🔹 Leonardo DRS Infrastructure (IcePiercer®): DRS provides the IcePiercer4 ®, a hermetically sealed, military-hardened single-phase liquid immersion cooling tank delivering up to 35 kW of cooling capacity in a compact 9U footprint. Filled with 26 gallons of dielectric fluid, it is specifically engineered for warships, submarines, armored ground vehicles, and mobile command posts. 🔹 The $OSS Compute Core (Oceanside 1U SDS-IC): DRS doesn't build the high-performance AI servers inside. $OSS supplies the custom-engineered 1U Short-Depth Server (SDS-IC)—built fanless, specially sealed, and designed to be fully submerged directly inside DRS’s dielectric fluid. 2. Why This Is a Home Run for $OSS Investors: 🔹 Official Tier-1 Co-Branding: Seeing the Leonardo DRS logo right next to the OSS logo on official promotional graphics proves deep architectural integration between both engineering teams. 🔹 Solving Edge AI's Biggest Bottleneck: Modern NVIDIA GPUs generate too much heat for traditional air cooling inside submarines, ships, or tactical vehicles (where salt fog, dust, heat, and acoustic signatures kill standard servers). Combining the OSS 1U SDS-IC with DRS IcePiercer® boosts compute density by up to 8x, slashes acoustic noise by -40 dBA (critical for submarine stealth), and doubles system reliability (MTBF). 🔹 Direct Q3 Earnings Validation: Management repeatedly highlighted custom Short-Depth Server (SDS) deliveries for a major naval/maritime customer. The Oceanside 1U SDS-IC built for IcePiercer® fits that exact profile down to the spec. $OSS is no longer operating as an anonymous sub-tier supplier in the shadows, but they are standing shoulder-to-shoulder with one of the Pentagon's premier defense primes. For multi-year rollout potential across both the U.S. Navy and U.S. Army, this visible integration with Leonardo DRS is the ultimate stamp of approval! 💪🏻
As AI, sensor processing and other compute-intensive workloads move to the edge, managing heat becomes increasingly important when maximizing GPU performance. OSS and Leonardo DRS are showcasing a new approach to rugged computing and thermal management at SC26. Together, the 1U SDS-IC and IcePiercer® significantly increase the compute capacity that challenges most data centers and optimize energy consumption. See both systems live at SC26, Booth #2824. #SC26 #ImmersionCooling #EdgeComputing #RuggedComputing #HPC
1
3
32
1,596
LugInvest retweeted
NEWS | Embraer surpasses its all-time high for a third quarter, with US$2.3 billion in revenues. Read full news: embraer.com/media-center/en/…
3
14
3
203
22,727
Zwei Werte aus dem Wikifolio „Distressed Value“, zwei Wendegeschichten: Ceragon wächst wieder und liefert Gewinn und Cashflow, Alpha & Omega verschiebt sein Geschäft vom schwachen PC-Markt zur KI-Infrastruktur. Die Q3-Zahlen und meine Einordnung gibt’s im neuen Beitrag. 👇
1
2
74