ScaleFlux

ScaleFlux builds innovative silicon solutions that eliminate data bottlenecks and help scale the AI data pipeline.

[ScaleFlux @ScaleFlux]

@theCUBE was live at the AI Data Pipeline Forum 2026.
Our CEO and Co-Founder Hao Zhong sat down with @furrier to talk about what it will take to keep scaling the AI data pipeline.

Watch here: https://t.co/lmv8nL6mhY

[ScaleFlux @ScaleFlux]

Some events are best experienced in motion. Here is the #ScaleFlux AI Data Pipeline Forum 2026, from keynote to demo floor to happy hour.

[ScaleFlux @ScaleFlux]

Scaling the AI data pipeline takes an ecosystem. Thanks to @NVIDIAAIInfra @Microsoft @VAST_Data @MemVerge @Supermicro and everyone who joined Hao Zhong at the AI Data Pipeline Forum 2026.

Full recap: https://t.co/2B5jOQW1ws

[ScaleFlux @ScaleFlux]

Every AI workload depends on moving the right data to the right place, fast. We shared how ScaleFlux makes that possible with the journalists of @ITPressTour this week.

Read more: https://t.co/fNyxXfpW80

[ScaleFlux @ScaleFlux]

Built by our partners. Driven by ScaleFlux.

Learn how partners use proven ScaleFlux technology to build the exact storage and memory products AI infrastructure needs: https://t.co/0qWzSOqBTQ

[ScaleFlux @ScaleFlux]

The 70th edition of @ITPressTour is coming to Silicon Valley, with a stop at ScaleFlux.

We’ll discuss the challenges shaping AI infrastructure and share how ScaleFlux is making data movement across storage and memory more efficient.

https://t.co/AWXAkLfjEP

#ITPT

[ScaleFlux @ScaleFlux]

What does it cost to leave a GPU waiting on storage?

FC6116 is a full-stack PCIe Gen6 SSD controller: 7M random read IOPS, 28 GB/s sequential read, 2x the perf per watt of the last gen.

See how it keeps AI compute fed
➡️https://t.co/nSs0In387I

[ScaleFlux @ScaleFlux]

Adding memory isn't the same as lowering cost.

Tong Zhang on why CXL's next phase depends on the controller, not the capacity. From FMS 2026.

Learn more: https://t.co/pJZhO4FubC

[ScaleFlux @ScaleFlux]

Introducing ScaleFlux Driven, a new program giving partners the silicon, firmware, reference designs, and flexibility to build differentiated NVMe SSD and CXL solutions for AI, data center, and enterprise infrastructure.

Learn more and become a partner➡️ https://t.co/fiJS2D5uoJ

[ScaleFlux @ScaleFlux]

Storage and memory are two halves of the same problem. ScaleFlux builds both.

The MC600 CXL memory controller and FC6116 NVMe SSD controller bring that platform into PCIe Gen6. If you are designing a Gen6 platform, let's talk.

Learn more: https://t.co/VZBmFe9vcb

[ScaleFlux @ScaleFlux]

The drives aren’t the bottleneck. The path to them is.

In his #FMS2026 session, ScaleFlux's Co-Founder and VP of ASIC Engineering, Yang Liu, explained why. Read it here in case you missed out: https://t.co/Eg76S3aCuw

Stay tuned for more FMS 2026 presentation recaps.

[ScaleFlux @ScaleFlux]

Break through the memory wall.

AI and cloud infrastructure need more memory to keep CPUs efficient. MC600 CXL 3.2, Type 3 controller supports up to 2TB DDR5 and ScaleFlux RS ECC that corrects 11 symbol errors across any two DDR5 chips.

https://t.co/Iu8wcencOl

[ScaleFlux @ScaleFlux]

We've been voted Market Leader and Innovation Leader for Computational Storage in the 2026 Enterprise AI Infrastructure Surveys from IT Brand Pulse

41.6% for Market Leader. 44.1% for Intelligence & Innovation Leader. 1st place in both. Check it out➡️https://t.co/nzAcnTnTgN

[ScaleFlux @ScaleFlux]

Day 3 at #FMS2026 and the energy hasn’t slowed down! #ScaleFlux Booth 519 is still buzzing with great conversations and demos.

We also had two great sessions from Prof. Tong Zhang earlier today and one session from Ben Reese still ahead.

We hope to see you there!

[ScaleFlux @ScaleFlux]

Great talk by Tong Zhang this morning at #FMS2026 on breaking the HBM cost wall for AI inference. Smarter memory tiering, KV cache placement, and MoE compression can lower HBM cost per token without sacrificing performance. #AI #HBM #ScaleFlux

[ScaleFlux @ScaleFlux]

Hao Zhong, CEO and Co-founder of @ScaleFlux delivered a terrific keynote, packed with announcements today at #FMS2026.

A special thank you to @NVIDIA for sharing the stage with us.

[ScaleFlux @ScaleFlux]

Don’t miss @ScaleFlux CEO Hao Zhong keynote with NVIDIA today at 1:20 PM at #FMS2026!

[ScaleFlux @ScaleFlux]

@ScaleFlux latest generation of products has arrived at #FMS2026.

Visit our Booth #519 to get a firsthand look at our newest SSD and controller innovations, designed to deliver more performance, higher endurance, and better power efficiency without changing your applications.

[ScaleFlux @ScaleFlux]

Keith McKay, @ScaleFlux presented at #FMS2026 this morning about Managing write amplification without data placement.

[ScaleFlux @ScaleFlux]

The magic has begun at #FMS2026! 🪄 Visit @ScaleFlux Booth #519 and catch our next live magic show. See you there! ✨