WEKA

🤝 At @RaiseSummit, @scality CEO @jlecat shared why the Scality + WEKA partnership continues to grow. Together, we're helping organizations in France adopt #AIInfrastructure with local, French-language expertise and support. 👉 Learn more: weka.ly/4aNCkJC

📣 We're excited to welcome Shanta Kohli as WEKA's new Chief Marketing Officer! Shanta brings 25+ years of experience building category-leading brands and growth engines across enterprise infrastructure, cybersecurity, and cloud. 🔗 weka.ly/4bcwUIn

💭 One of the biggest takeaways from #RAISE2026? AI is no longer just about building bigger models. It's about lowering the cost per token and making production AI more efficient. @liranzvibel and Dylan Patel (@SemiAnalysis_) shared their thoughts at @RaiseSummit.

💜 Paris looks good in purple. Thank you, @RaiseSummit, for an incredible experience this week. And thank you to our customers, partners, and the entire WEKA team for making it so memorable. #RAISESummit #RAISE2026

🎙️ WEKA's Val Bercovici joined @vikramskr on the @semidoped podcast to discuss KV cache economics, token pricing, DeepSeek V4, CXL, NVLink, edge AI, and what's next for #AIInfrastructure. 👉 Watch: weka.ly/4f8bA7H

Vik and Val Bercovici (@weka) map where AI inference memory is headed. Every 100x cut in KV cache gets swallowed by ~10,000x more usage, so demand climbs. - NVLink beats the board: 128 lanes vs 32 PCIe - WEKA serves NAND-backed storage faster than DRAM over network - DeepSeek's cache reads run ~87x cheaper, China only - CXL needs a dedicated bus, WEKA pools NAND over RDMA/NVLink instead - Val's call: SaaS giants and Neoclouds must merge Chapters: 0:00 Intro and memory prices 3:00 Model routing and offloading 5:10 Network faster than motherboard 13:06 Four-tier memory hierarchy 19:40 KV cache and Jevons paradox 23:50 DeepSeek cache read pricing 31:40 Sliding window attention 34:49 NAND tiers, SLC vs QLC 43:57 High bandwidth flash use cases 49:59 CXL versus NVLink 53:44 AMD Mex and small models 1:03:14 SaaS, Neoclouds and tokenomics Get more of Austin and Vik daily, free! Sign up: daily.semidoped.com Connect with Vik and Austin: Vik's Paid Substack: viksnewsletter.com Austin's Paid Substack: chipstrat.com @austinsemis @vikramskr

🇫🇷 That's a wrap on @RaiseSummit! Two days of great conversations, packed sessions, and connecting with customers, partners, and AI leaders from around the world. Thanks to everyone who stopped by the WEKA booth. 👋 Until next time! #RAISESummit

@WhiteFiber_ 👏

Deploy without designing around bandwidth constraints. WhiteFiber's proprietary cross-data-center networking solution: built-in resilience, compliance, and a scaling path that doesn't start over. Commercial availability: Q3 2026. whitefiber.com/blog/cross-dat…

✨ Pure cinéma. After the @RaiseSummit show floor closed, WEKA and @scality brought together some of the brightest minds in #AIInfrastructure for an unforgettable evening in Paris. Merci to everyone who joined us.

🙌 What a start to @RaiseSummit! Day 1 was full of great conversations, new connections, and exciting discussions about the future of #AIInfrastructure, production AI, inference, and memory. ⚡ On to Day 2!

💸 As AI applications consume trillions of tokens, GPU memory limits and KV cache eviction are quietly driving up inference costs and making it harder to scale agentic AI efficiently. This article explores why 👇 weka.ly/44in6ss

🍹 The show floor is buzzing, our juice is made to order, and the conversations around #AIInfrastructure and production inference are flowing. If you're at #RAISESummit, stop by the WEKA booth for a fresh drink from The Bottleneck Juice Bar and a little extra hydration.

🇫🇷 Bonjour Paris, and hello @RAISESummit! The WEKA flag is flying, and we’re looking forward to a great week of conversations about AI, infrastructure, inference, and everything shaping what’s next. 🤝 If you’re here, come connect with the team on the show floor.

🇫🇷 WEKA's Val Bercovici is headed to @RAISESummit! If you're attending this week, catch his panel on the economics of production AI, and don't miss @liranzvibel's session on the AI memory wall with Dylan Patel of @SemiAnalysis_. 🔗 weka.ly/4h17hxD

☝️ Your PostgreSQL performance might have less to do with PostgreSQL than you think. We benchmarked PostgreSQL on Amazon EKS using NeuralMesh and gp3 EBS to show how storage latency impacts throughput, TPS, and application performance. weka.ly/4eQ0c1t

In this article, WEKA's Val Bercovici explores how @nebiusai is rethinking #AIInfrastructure with smarter hardware, software, cooling, and data movement to maximize performance while reducing energy waste. Take a look 👇 weka.ly/3TkBrSB

💡 Enterprise AI has moved beyond isolated training jobs. In this article, WEKA's Nilesh Patel explains why today's #AIInfrastructure must support training, fine-tuning, and inference all at once. Take a look 👇 weka.ly/3QA53e3

📣 ICYMI: On the latest Deep Geeks podcast, Dr. Serena Huang sits down with Jessica Audette of @MSKCancerCenter to explore how the right infrastructure is helping bring AI to healthcare at scale. 👉 Watch: weka.ly/4y2d4t0

🧠 AI isn't just a race for better models. It's a race for energy, memory, and trust. In a recent interview, WEKA's Val Bercovici shares why infrastructure will define the next era of enterprise AI. weka.ly/4wi1TKY

🇫🇷 Next week: @RAISESummit! Stop by the WEKA booth and catch sessions with @liranzvibel and Val Bercovici. ✈️ weka.ly/4bnKJDB