Watch @ThorBjornsson_ , famously known as The Mountain from Game of Thrones, achieve an unforgettable feat at SC24—“deadlift of data”!
Missed the action live? Relive the excitement now, and see why VDURA is redefining the future of data performance and durability.
#VDURAStrong #DataDeadlift #SC24
Storage decides how much revenue your GPUs actually earn. Our CEO Ken Claffey told @vmblog why it’s a revenue driver for #AI factories and Neoclouds and not a line item.
Read the full Q&A. https://t.co/mLGQw4SGiX
Going to #SC26? @VDURAinc CEO Ken Claffey explains why storage is a GPU revenue driver, not a line item, and what V12 means for AI factories and neoclouds.
Read the @VMblog Q&A before you go: https://t.co/gWRmzysvvB
#HPC#AI#Storage
KV caches now outgrow GPU memory, and every eviction means recomputing context you already paid for. Our #SC26 Exhibitor Forum talk covers how to make persisted context a first-class storage tier.
📍Nov 18, 11:30am CST, Room W192b
Book time with us https://t.co/f7hSNO8ZI7
Every GPU you buy is a bet that your storage can keep it busy. Win that bet at @Supercomputing#SC26.
📍 Booth 1423, McCormick Place
📅 November 17-19
Book a meeting with our team and we’ll map out where your storage dollars should actually go.
https://t.co/f7hSNO9xxF
Six women shaping the future of #HPC and #AI, one table, our @Supercomputing booth 1423.
Where Innovation Meets Execution is a Women in HPC panel on the innovations they’re proudest of, the challenges they’ve faced and how they pay mentorship forward.
Nov 17, 2 PM CT. #SC26
@SemiAnalysis_ showed taller HBM stacks add capacity, not bandwidth. Our CEO says the same goes for the flash under your GPUs.
A 61 TB NVMe drive is no faster than a 7.68 TB one, at 14x the cost of disk. Buy bandwidth, not TBs.
@blocksandfiles https://t.co/Tby4RzRddL
More GPUs per storage dollar. See how at @Supercomputing#SC26.
📍 Booth 1423, McCormick Place, Chicago
📅 November 17-19
Book a meeting with our team and we’ll dig into your architecture, your budget and where the GPU math changes. https://t.co/f7hSNO8ZI7
GPUs don’t deliver value sitting idle. They deliver it when the data keeps flowing.
@AMD just published a case study and short video on how VDURA brings compute, networking and storage together into one data platform that keeps accelerators busy. #AI
https://t.co/AXHvy6u8iz
In an exclusive interview with Tech Revolt, @VDURAinc@mswalley2539 says storage is the next bottleneck for AI factories, with idle GPUs driving up costs. https://t.co/QYi880nOaz
@Supercomputing#SC26 is three weeks out and we can’t wait for Chicago.
📍 Booth 1423, McCormick Place, Chicago
Book a meeting ahead of time and let’s talk about your next #AI cluster. 💪
https://t.co/f7hSNO8ZI7
Is your #data layer an enabler or a detractor of #GPU capital, power and availability?
We tackled that question in our AI Labs Tech Talk at @AiEverythingAD. Full talk is up now👇
#AI
https://t.co/qzh2cCvhbQ
Two years ago the question was how do I get GPUs.
Now it’s how do you keep them busy and paying for themselves.
Our Senior Director @mswalley2539 talks with @khaleejtimes about why the #AI race isn’t slowing down and where storage fits in: https://t.co/B41lRytusC
Mountain vs. machine at @AiEverythingAD 🤖🥊
@ThorBjornsson_ squared up with a humanoid robot outside our booth and the whole concourse stopped to watch. We're calling it a draw.
He's back at CC-A10 tomorrow. 🦾 #AiEverything
AI’s free ride is ending. The per-token price you’re paying today doesn’t reflect what it actually costs to serve, and the subsidy covering the gap is running out.
Our CEO’s take in @CIOonline on the #AI pricing reset and what to do before it hits: https://t.co/eaajTWUqUD
The UAE isn’t just buying GPUs. It’s building the infrastructure to keep them working.
Our own @mswalley2539 talks with The Gulf Time about why high performance storage is becoming the deciding factor as AI infrastructure scales across the region.
https://t.co/GOPK3cqnNC
"Every minute a GPU waits on a checkpoint, a model load, or a cold read is inventory that expires unsold."
Our CEO on why the data layer decides how much #GPU capacity turns into billable output. New interview with @KnowledgeNile https://t.co/EaDYI4SqIY
Thor found a new sparring partner on the show floor at @AiEverythingAD, and this one came with dance moves. 🤖
Catch @ThorBjornsson_ at CC-A10 for the rest of the day. #AiEverything
Mountain vs. machine at @AiEverythingAD 🤖🥊
@ThorBjornsson_ squared up with a humanoid robot outside our booth and the whole concourse stopped to watch. We're calling it a draw.
He's back at CC-A10 tomorrow. 🦾 #AiEverything
Every GPU-hour spent waiting on data is revenue an operator never sees.
See how we partner with @AMD to keep GPUs fed, with an end-to-end reference architecture across Instinct, Pensando and EPYC that delivers up to 3x performance per watt on off-the-shelf hardware.