Happy #HotFuzz day!

🧱 WEKA’s Augmented Memory Grid unveiled at #GTC25 will integrate with @NVIDIA Triton Server so you can scale the dreaded GPU memory wall like an expert rock-climber. You don't have to choose between speed, accuracy, or costs – you can optimize all 3👇 weka.ly/3Rf7CPc

😵‍💫 There’s no way around it: Shift happens. But there’s good news: WEKA can help. Visit us this week at #GTC25 to learn how you can shift with AI instead of AI shifting on you. And if you need help finding us, just keep an eye out for the purple threads 👕👇

🎬 Our Augmented Memory Grid announced at #GTC25 addresses a critical AI topic: KV Cache. Tune in to this video for the difference between KV Cache memory and KV Cache storage to avoid hitting the memory wall. Go to our blog for even more info: 👇 weka.ly/4iG7FQ4

👋 Callout to the builders, dreamers & AI doers on National Customer Appreciation Day! We're proud to help power your journey — whether you're scaling billion-param models or smashing data bottlenecks. Here’s to what we’ve built (and what’s next) 👇 weka.ly/3RlPDXz

🚀 We’re proud to share that WEKA was named a Customers’ Choice in the May 2025 Gartner Peer Insights™ “Voice of the Customer”: File and Object Storage Platforms! Read the press release to see why customers are choosing WEKA 👇 weka.ly/45v4V4j

Every thread serves a purpose. The strongest networks aren't built - they're woven with precision and intelligence. Everything connects on 06.18.25

📣 BIG NEWS! Introducing NeuralMesh™ by WEKA, a major shift in AI infra that scales with AI’s demands. AI pushes traditional stacks to the brink. NeuralMesh gets smarter, faster + more resilient as workloads grow. 🎥 Watch the video, then read the press release for more 👉

Scaling #AI with endless tweaks + rewrites? Hard pass. 🚫 NeuralMesh lets you move fast, crush bottlenecks, and scale smarter. Proof-of-concept → production, without the drama. Learn more 👇 weka.ly/4fYgjsA

AI lock-in is like quicksand: you don’t notice until you’re sinking. Build flexibility in from day one, so your AI can scale without the struggle. Learn more in this blog post 👇 weka.ly/3JEou1g

Stuck in legacy storage? 's NeuralMesh delivers microsecond latency, blazing-fast inference, and GPUs firing on all cylinders — no bottlenecks, no idle cycles, no compromises. Discover the truth: spr.ly/6013AcxgZ

🧠 Congrats, you're now part of the brain dump. #SC25 #Supercomputing25 #HPCignites

🤝 Big shout-out to our amazing partners at #SC25! From joint demos to insightful conversations and everything in between, your collaboration is helping push #AIInfrastructure forward. We’re thrilled to be building the future together. #Supercomputing25 #HPCignites

At #SC25, we caught up with NVIDIA’s Dion Harris and WEKA CTO Shimon Ben David to break down what AI factories really are — and how NVIDIA + WEKA are working together to build them. From GPU-feeding data pipelines to seamless compute, networking and storage alignment, this is the

😅 If you're chasing NVMe SSDs, you’re not alone. Months-long lead times and tight supply shouldn’t be what holds back your AI plans. NeuralMesh Axon helps teams skip the wait by unlocking the NVMe they already have. 🎁 weka.ly/4jhdncn

🖼️ Let’s zoom out for a second. Happy little GPUs only stay happy when the whole system works together. NeuralMesh brings data, compute, and AI into harmony—and gets faster as you scale. 🎨 Paint a better AI picture: weka.ly/4aEKGnt

GPU memory shortage? Don’t panic-buy. WEKA’s NAND Flash Shortage Survival Guide shows 5 ways to maximize GPU utilization, cut waste, and extend memory with what you already have. Grab it here ⬇️ spr.ly/6010h6uy2

😀 “That was awesome.” Said no one ever about 6-month NVMe lead times. When AI workloads can’t wait, chasing drives isn’t a strategy. With NeuralMesh Axon, WEKA lets you use the NVMe already inside your GPU servers—turning it into a shared, ultra-fast storage layer for AI. 👉

🎶 You should be dancin’… not waiting on NVMe. NeuralMesh Axon turns the NVMe in your GPU servers into a shared, ultra-fast storage layer for AI. ✔️ More memory ✔️ Better GPU use ✔️ No supply headaches 🕺 weka.ly/4bPOL8Y

Inference is taking center stage in AI for 2026. WEKA's Val Bercovici had a great conversation with Kyle Sosnowski from Crusoe during #AIInfraSummit about memory-bound optimization, KV cache, and self-healing infrastructure. Watch below ⬇️ spr.ly/6013htUYh