About
Articles by Mark
-
Why I'm Joining the PyTorch Foundation`
Why I'm Joining the PyTorch Foundation`
I want to start by thanking Matt White for everything he has built over the past two years. The growth of the PyTorch…
33 Comments -
DeepSeek Proves Powering Open Source AI is All in the Infrastructure
DeepSeek Proves Powering Open Source AI is All in the Infrastructure
I'm sure you've seen DeepSeek, the whale heard around the world? Last month, the folks at DeepSeek revealed the secrets…
2 Comments
Activity
-
Adding to the "ROOST.tools is growing" category, we're ready to bring on a Director of Engineering. We're looking for someone who can develop and…
Adding to the "ROOST.tools is growing" category, we're ready to bring on a Director of Engineering. We're looking for someone who can develop and…
Liked by Mark Collier
-
Most local document extraction guides stop at "here's how Docling works." I wanted to know what happens after . How do you turn extracted PDF…
Most local document extraction guides stop at "here's how Docling works." I wanted to know what happens after . How do you turn extracted PDF…
Liked by Mark Collier
Experience & Education
-
PyTorch
View Mark’s full experience
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Recommendations received
1 person has recommended Mark
Join now to viewMore activity by Mark
-
I’m excited to share my next chapter: I’ve joined Fireworks AI. From AMD, Apple, Uber, Meta, Google, and most recently Snowflake, I’ve worked on many…
I’m excited to share my next chapter: I’ve joined Fireworks AI. From AMD, Apple, Uber, Meta, Google, and most recently Snowflake, I’ve worked on many…
Liked by Mark Collier
-
🎇Excited to share that I'll be speaking as a PyTorch Technical Ambassador (India) at AMD Dev Day 2026 in Bengaluru on 8th August. 🎤 Session…
🎇Excited to share that I'll be speaking as a PyTorch Technical Ambassador (India) at AMD Dev Day 2026 in Bengaluru on 8th August. 🎤 Session…
Liked by Mark Collier
-
[Reflections on KCD & OpenInfra Days Vietnam 2026: Passion, Community, and the AI Era] 🇻🇳✨ Returning to Vietnam for KCD & OpenInfra Days Vietnam…
[Reflections on KCD & OpenInfra Days Vietnam 2026: Passion, Community, and the AI Era] 🇻🇳✨ Returning to Vietnam for KCD & OpenInfra Days Vietnam…
Liked by Mark Collier
-
Yesterday, my colleague presented our journey at KubeCon Japan in building the GPU over APN Testbed (GoAT), a nationwide Kubernetes-based GPU…
Yesterday, my colleague presented our journey at KubeCon Japan in building the GPU over APN Testbed (GoAT), a nationwide Kubernetes-based GPU…
Liked by Mark Collier
-
Kimi K3 has first class support for vLLM on NVIDIA! vLLM serves K3 across Grace Blackwell, Blackwell, Hopper, and NVL72 systems, also integrating…
Kimi K3 has first class support for vLLM on NVIDIA! vLLM serves K3 across Grace Blackwell, Blackwell, Hopper, and NVL72 systems, also integrating…
Liked by Mark Collier
-
Milvus 3.0 is here! As an LF AI & Data graduated project, Milvus, created by Zilliz has reached its biggest architectural milestone yet. Milvus 3.0…
Milvus 3.0 is here! As an LF AI & Data graduated project, Milvus, created by Zilliz has reached its biggest architectural milestone yet. Milvus 3.0…
Liked by Mark Collier
-
NVIDIA is searching for a highly motivated, creative engineer with direct experience in Linux kernel-mode drivers to help us deliver on the future of…
NVIDIA is searching for a highly motivated, creative engineer with direct experience in Linux kernel-mode drivers to help us deliver on the future of…
Liked by Mark Collier
-
I’ll be in Vietnam, Friday and Saturday at the OpenInfra days. If you want to know more about HAMi, dynamia.ai‘s Enterprise version, DRA or generally…
I’ll be in Vietnam, Friday and Saturday at the OpenInfra days. If you want to know more about HAMi, dynamia.ai‘s Enterprise version, DRA or generally…
Liked by Mark Collier
-
#CoreWeave is hosting its inaugural AI cloud user conference, Fully Connected 2026, at the Moscone in SF from Sept 29th to Oct 1st. Curated sessions…
#CoreWeave is hosting its inaugural AI cloud user conference, Fully Connected 2026, at the Moscone in SF from Sept 29th to Oct 1st. Curated sessions…
Liked by Mark Collier
-
Many thanks to 张小珺 for the invitation! Kimi K3 hadn’t been released by Kimi (Moonshot AI) when we recorded the podcast, and I had no idea it would be…
Many thanks to 张小珺 for the invitation! Kimi K3 hadn’t been released by Kimi (Moonshot AI) when we recorded the podcast, and I had no idea it would be…
Liked by Mark Collier
-
Feeling beyond welcome at #KubeCon & #CloudNativeCon Japan! Really happy to meet the community here and see so much innovation in his region. Thank…
Feeling beyond welcome at #KubeCon & #CloudNativeCon Japan! Really happy to meet the community here and see so much innovation in his region. Thank…
Liked by Mark Collier
-
OpenStack is at KubeCon + CloudNativeCon Japan 2026 in Yokohama! A big thank you to the Japan OpenInfra User Group for volunteering. Visit us at…
OpenStack is at KubeCon + CloudNativeCon Japan 2026 in Yokohama! A big thank you to the Japan OpenInfra User Group for volunteering. Visit us at…
Liked by Mark Collier
-
I was in Sherman, Texas to celebrate the groundbreaking for Coherent’s new factory – where indium phosphide lasers will be built to connect hundreds…
I was in Sherman, Texas to celebrate the groundbreaking for Coherent’s new factory – where indium phosphide lasers will be built to connect hundreds…
Liked by Mark Collier
-
Richard Sikang Bian, Product & Operations Director for AntLing, will join AGNTCon + MCPCon China as a keynote speaker in Shanghai on September…
Richard Sikang Bian, Product & Operations Director for AntLing, will join AGNTCon + MCPCon China as a keynote speaker in Shanghai on September…
Liked by Mark Collier
-
Have I ever told you I love my job? The secure container runtime projects I've worked on at CoreWeave have now unlocked new business opportunities…
Have I ever told you I love my job? The secure container runtime projects I've worked on at CoreWeave have now unlocked new business opportunities…
Liked by Mark Collier
Other similar profiles
Explore more posts
-
Kit Yu
NetApp should be well positioned to participate in this enterprise-level AI spend as it materializes because of its: (1) deep enterprise relationship for traditional business applications; (2) its differentiated hybrid multi-cloud capabilities; and (3) its AI-tailored hardware and software, including the AFX system and AI Data Engine (AIDE). First, NTAP has deeply entrenched relationships with some of the largest enterprises in the world with 85% of Fortune 500 using its product. We believe NetApp's position as a highly reliable incumbent with a long track record makes for an easier cross-sell motion as enterprises look to make investments to ready themselves for future AI workloads. Second, NTAP's ONTAP platform offers a unified storage infrastructure for hybrid cloud/multi cloud environments. In addition to being a widely adopted, mature on-premise storage platform from traditional business applications, ONTAP is also uniquely runs as a native, first-party cloud service in all three major public clouds. ONTAP allows customers to simplify its management of a data set across multiple environments, which will be increasingly important as enterprise implement AI solutions in the most cost effective and secure way. Third, NTAP recently launched a new AFX storage system that supports the independent scaling or compute and storage resources. This disaggregated storage architecture is well suited for the performance and the capacity demands of AI workflows.
-
Nate Nead
👉️ https://lnkd.in/g-kUShWB 🖥️ Is That Really Your Device? Hardware-Fingerprinting Says Yes—Until It Doesn’t. In a world of remote work, cloned VMs, and increasingly stealthy adversaries, verifying device identity is more critical than ever. Enter hardware-fingerprinting: a way to bind trust to the physical machine—not just the login credentials. ✅ Stronger than software-based IDs ✅ Low user friction ✅ Real-time policy enforcement But… it’s far from bulletproof: ⚠️ Hardware swaps trigger false negatives ⚠️ Privacy regs like GDPR and CCPA loom large ⚠️ Advanced attackers are already finding ways around it This blog explores the promise and pitfalls of using low-level hardware signals (like MACs, TPMs, and SSD serials) to enforce endpoint integrity. It’s a powerful tool—but only when paired with re-enrollment plans, privacy safeguards, and a broader security stack. 💬 Are you using hardware-fingerprinting today? Is it delivering value—or just more help-desk tickets? Let’s discuss 👇 #Cybersecurity #EndpointSecurity #ZeroTrust #ITOps #SecurityArchitecture #PrivacyByDesign #HardwareFingerprinting #DeviceTrust #InfoSec #ITCompliance
-
Simone Morellato
The Kubernetes layer often limits GPU cloud providers in ClusterMAX rankings, not the hardware. AI teams expect managed Kubernetes like EKS or GKE, but delivering isolated clusters per customer traditionally means operating dozens of separate physical clusters. Leading GPU cloud providers are solving this by virtualizing the Kubernetes control plane and giving each customer full isolation while workloads run efficiently on the GPU infrastructure. At vCluster we have seen our customers delivering: - Managed K8s services in months, not years - GPU workloads with zero performance overhead - Full compatibility with GPU Operator, NCCL, and distributed training - Strong multi-tenancy without managing dozens of clusters Our latest article maps this architecture to every ClusterMAX Kubernetes criterion. Read the full breakdown below. #Kubernetes #DevOps #GPUCompute #AI
-
Jeff Robertson
Looking forward to being at #APCOWRC2026 and spending time with so many agencies leading the way in NG9-1-1. Arizona has made real progress building the foundation – now it’s about fully using it. I’ll be sharing a few practical thoughts on how we move past “portal overload” and start getting more out of the ESInet. Hope to see you there.
1 Comment
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content