Chris Wright
@kernelcdub
Followers
7K
Following
4K
Media
90
Statuses
2K
Red Hat CTO. Tezos Foundation council member. Passion for open source SW innovation. Father and husband. Cyclist. Human.
Boston, MA
Joined May 2012
Tushar and I discuss critical infrastructure, open standards like MCP, and building that platform on the new Technically Speaking. Check it out!
Building for the "Year of the Agent"? You’ll need more than just a model. From Llama Stack to MCP, Tushar Katarki joins @kernelcdub to break down the open source stack powering the next wave of #AI. Listen here: https://t.co/V7hbfouDf1
0
0
5
Big breakthroughs rarely happen in a silo. The latest Technically Speaking episode dives into the importance of open science and shared benchmarks—the unsung heroes that power massive discoveries. Listen in as Red Hat CTO @kernelcdub and Dr. Jianying Hu from @IBM discuss
0
5
15
We're accelerating engineering with AI, not replacing core practices. It's about empowering choice & ensuring AI is grounded in open source principles (and yes, code review is non-negotiable): https://t.co/kylY6TO1y8.
redhat.com
Explore Red Hat's adoption of AI within their engineering organization, focusing on culture, choice, and the acceleration of open source innovation. Learn about the internal rollout, productivity...
0
5
9
Cool to see this from the RHEL team. Smaller images for minimal attack surface along with rapid rebuild/updates so images are free of known vulnerabilities. Also freely available and redistributable, like UBI.
redhat.com
Project Hummingbird is an early access program for Red Hat subscription customers, providing a catalog of minimal, hardened container images.
0
0
3
On the heels of the recent news that Red Hat will offer Confirmed Sovereign Support ( https://t.co/qyevNLNd0e), we are also joining a coalition of organizations to ensure Europe’s technological future is open, trusted and its own: https://t.co/DYgNeuduLL.
1
3
16
"An average household uses about 200 cubic feet of natural gas a day...the volume of about 10 fridges." https://t.co/QhH4QP686H
tristateinfrastructurenews.com
"An average household uses about 200 cubic feet of natural gas a day...the volume of about 10 fridges."
1
2
11
Dr. Jianying Hu from @IBMResearch and I dive into how open science and domain-specific models are driving real discoveries, from speeding up drug development to creating a "virtual cell." This is AI innovation at the intersection of technology and humanity!
What if your healthcare wasn't one-size-fits-all? AI foundation models are helping researchers understand "patient stratification", or why different people respond differently to the same disease or therapy. Hear about this new frontier on the Technically Speaking podcast:
0
2
8
#AI demands smarter #networks, but how do we scale them for the future? 🔮 Our CTO @erikekudden and @RedHat CTO Chris Wright tackle the challenge in the latest #CTOFocusSessions episode. 👉 Watch the full session for their insights: https://t.co/MozH3FNFLr
1
5
16
The AI model gets all the press, but I’m here to talk about intelligent infrastructure. The real work (and the real business value) is in efficiently running inference across your hybrid cloud. Read my take on why the AI system is the next frontier: https://t.co/xbhB1o61lR.
redhat.com
Learn about llm-d, an open source control plane that enhances Kubernetes for AI workloads. Discover features like semantic routing, workload disaggregation, and support for advanced architectures....
0
5
9
What does it take to make mobile networks ready for AI at scale? Check out my discussion with @Ericsson CTO Erik Ekudden to see why it all comes back to open source and a strong hybrid cloud foundation: https://t.co/aMOeVVLGVe.
ericsson.com
The latest strategic discussions on the 5G ecosystem across industries hosted by Ericsson’s CTO, Erik Ekudden.
0
0
1
Thinking about how to manage both privacy and integrity for AI models during inference? Here's some work we're exploring in this space, bridging confidential compute and LLMs.
next.redhat.com
Red Hat’s Office of the CTO is collaborating in the upstream project Tinfoil community to explore pioneering a complete, cloud-native solution for Confidential AI. The community is focused on solving...
0
1
1
Plenty of discussion about the @thinkymachines paper on Defeating Nondeterminism in LLM Inference. Great to see this work in #vLLM!
🚀 Excited to share our work on batch-invariant inference in vLLM! Now you can get identical results regardless of batch size with just one flag: VLLM_BATCH_INVARIANT=1 No more subtle differences between bs=1 and bs=N (including prefill!). Let's dive into how we built this 🧵👇
0
0
1
Navigating the legal issues with open source and AI-assisted development. Let’s dive into what’s being actively debated inside open source communities right now: https://t.co/tXA1UGpiuS
redhat.com
Explore the legal and cultural issues surrounding AI-assisted development in open source communities. Learn about attribution, copyright, licensing, and the DCO in this comprehensive guide.
0
1
0
Enjoyed spending time in London yesterday with top industry analysts and other @RedHat leaders. It was great to discuss how our open hybrid cloud strategy is helping customers navigate a complex IT landscape. Our open approach for AI is a critical part of that journey.
1
4
23
AI is changing how we do everything. Lots of discussion about AI in the sw development But not enough about how open source sw development will evolve with AI. Let's have that discussion.
redhat.com
Explore Red Hat's approach to AI in open source, focusing on empowering developers and ensuring appropriate safeguards. Learn about the potential of AI-assisted development and Red Hat's commitment...
0
1
4
Using light to transmit data rather than relying on electronic components could slash latency
itpro.com
Using light to transmit data rather than relying on electronic components could slash latency
0
1
0
vLLM day zero support for Qwen3-Next...Hybrid attention, sparse MoE, and multi-token prediction...keep pushing fwd!
Qwen3-Next dropped yesterday and you can run it with Red Hat AI today. ✅ Day-zero support in vLLM ✅ Day-one deployment with Red Hat AI Step-by-step guide: https://t.co/ZjLJyfmMJm The future of AI is open.
0
0
5
Check out my new Technically Speaking episode with Bernd Greifeneder from @Dynatrace. We're talking about taming AI agents with observability and why being able to trust these systems is so critical. Give it a listen.
1
5
9
It's awesome to see how we're making the OS feel faster and more secure for our customers with things like Lightspeed, Insights, image mode, and post-quantum security. This is what it’s all about—giving people the tools to move at the speed of innovation. https://t.co/VmxrMUl5Qk
1
5
14
Tired of tangled fur and endless shedding? The 2-in-1 Pet Dematting Comb tackles stubborn mats and gently removes loose hair, keeping your pet’s coat smooth, shiny, and healthy—all from the comfort of home.
3
9
142
.@kernelcdub and Carlos Costa explain how a "shared vision" allows an entire industry, from hardware providers to cloud companies, to solve common pain points together and create a "growing pie" for everyone. Hear more about the power of open source collaboration in the full
0
7
28
Ever wonder what the 'v' in vLLM stands for? 💡 @kernelcdub and @nickhill33 explain how "virtual" memory and PagedAttention make AI inference more efficient by solving GPU memory fragmentation. Tune into the full Technically Speaking episode to learn more about optimizing LLMs:
1
10
40