@HPC Podcast Archives - OrionX.net - HPC News Bytes – 20260803

Episode Date: August 3, 2026

- Open Secure AI Alliance - Open vs. closed AI - European AI gigafactories - Sovereign AI policy to architecture - Trusting results from Quantum Computers [audio mp3="https://orionx.net/wp-content/uploads/2026/08/HPCNB_20260803.mp3"][/audio] The post HPC News Bytes – 20260803 appeared first on OrionX.net.

Welcome to HPC Newsbytes, a weekly show about important news in the world of supercomputing, AI, quantum computing, and other advanced technologies. Hi, everyone. Welcome to HBC Newsbytes. I'm Doug Black, and with me is Shaheen Khan. Amidst a welter of recent AI security news, Nvidia and a group of technology, cybersecurity, cloud, enterprise software, and open-source organizations have launched the Open Secure AI Alliance. Participants include the Linux Foundation, Microsoft, IBM, Hugging Face, Cisco, Crowdstrike, Cloudflare, Red Hat, Palo Alto Networks, and HPE, among others. Early contributions address several layers of the agent stack, cryptographic workloads, identity, safer model formats, digitally-signed software patches, multi-agent vulnerability,

scanning, secure coding workflows, and tools for tracing and governing agent behavior. NVIDIA is contributing models, weights, data, and its open source NOA research framework. That acronym stands for the NVIDIA Labs object-oriented agent. Well, setting aside the question of why these capabilities are not already present in what everybody provides, the alliance is about security, but it's also very much about. the open versus closed AI debate. It basically says that cybersecurity defense is a community project. It requires transparency. It needs systems that defenders can inspect and modify and self-host. And since AI needs cybersecurity, it would need the open model. Hosted models can be opaque and may

refuse legitimate requests because, as we've discussed here before, malicious commands and genuine debugging requests can look identical. Open models give security teams greater control, although they also give attackers access to capabilities that might be used in malicious ways, but historically they have been proven to be more secure. The more difficult task is securing the complete agent stack. Model weights are only one layer. There's also the usual requirements of identity, permissions, isolation, provenance, logging, and the entire software supply chain all the way to applications and users. And in this case, the users are increasingly AI agents. We've joked about, quote, AI agents being people too kind of a thing,

but it's a pretty good way to point to the complexity as AI agents are granted or gain more autonomy. Anyway, shared standards in those areas could become foundational for AI agents and are important. Another consideration is how a coalition of so many large companies can work together productively on technologies and in this case also to act as a lobbying body to influence policy. The EuroHPC organization has launched its formal call for consortia to build and operate European AI gigafactories. These facilities are intended to sit above the existing AI factory program and scale, combining very large accelerator installations, data infrastructure, networking, software, and access mechanisms for European model developers and industrial users. The call asks prospective

consortia to address financing, construction power, operations technology sourcing, and long-term commercial viability. It also reflects Europe's preference for public-private infrastructure rather than relying entirely on foreign cloud providers. The program remains a procurement and policy framework rather than an operating system as of today. Many questions remain for this effort, including how much it will cost, accelerator supply, power availability, delivery schedules, and how many installations will be justified by European demand. Europe is marching on with its unique take on AI.

It has worked on regulating data and then AI, built large-scale systems around the European Union and is now moving towards financing the infrastructure that AI data centers need. The Gigafactory model recognizes that frontier-scale AI requires hard infrastructure like land, power, cooling, and system infrastructure, like systems and clusters, and that all of it needs capital at one end

and customers at the other. As AI development has turned into a race, everyone's challenge has been execution speed. Either that or an inability to explain why a slower path will be okay. While it has its defenders, Europe is typically seen as lacking speed even within Europe. Public-private partnerships have emerged as a way to bring in more capital and efficiency into the process. These structures can aggregate resources and support strategic autonomy, but they can also

create fragmented governance and slow procurement. The European program will be strategically useful even if Europe does not match the scale and scope of American investment. It can support regional models, scientific computing, defense applications, industrial AI, and negotiating leverage with external suppliers. Now, much of the system infrastructure comes from non-European companies, and that will continue. So a deeper question is what sovereignty means. and sovereignty cannot mean complete technological independence. The U.S. comes closest to having that, but it's not 100%. Having your own models, weights, jurisdiction, operational control, data governance, and

assured access are arguably more important considerations. Speaking of sovereignty, our next story involves the question of whether sovereign AI is not just a matter of policy, but also of architecture. A new archive paper presents a German-English Open Source AI Foundation model designed around European sovereignty and efficient deployment. It uses a mixture of experts' architecture with roughly 30 billion total parameters. It also combines transformer and mamba-style components, which the authors say keeps the inference cache comparatively stable as context length grows.

The work represents a broader shift, towards nationally controlled models optimized for specific languages, regulatory environments, and infrastructure constraints. As with any newly released model, independent evaluation will be needed to validate quality, throughput, and safety claims. Sovereign AI is often discussed as a question of where a model was trained or whose laws govern the data. This project highlights another aspect, that sovereignty is also a question of architecture, as you mentioned. For example, an AI model intended for regional deployment may prioritize language, quality, open weights, efficient inference, predictable memory use, and compatibility with locally available infrastructure

over absolute benchmark leadership. A region, or even a country, may not have the resources for a large AI training program, but it can still build useful models around domain knowledge, public services, industry, culture, and trusted deployment. This perspective is aligned with the interest in focused small language models. Open source models also lead to discussions of whether sovereignty requires complete national ownership of every layer. Countries could combine shared research with local adaptation and operational control. The market could therefore separate into a small number of global frontier models and a much larger population of efficient regional or domain models.

Sovereign AI could become less about isolation and more about maintaining credible freedom of action and meeting local requirements, or shall we say, biases. But whether architecture is a form of bias is another discussion. IBM and researchers from the University of Chicago have announced a quantum computing demonstration that they say addresses the criteria for quantum advantage, defined as computations beyond the reach of classical simulation method, while also providing trust that the computation has returned accurate results. In their new paper, the researchers reported that these two goals could be simultaneously achieved

by a novel construction of encoded quantum circuits enabling, they say, one of the largest demonstrations of logical quantum computing to date. The quantum industry has debated for years what qualifies as quantum advantage. We've talked about quantum speed, or accuracy advantage. But if you have enough speed advantage, then users must decide whether to believe an answer that classical systems cannot independently verify.

So useful quantum computing will require an entire trust stack in its own right, capabilities that may become as strategically important as qubit counts or gate fidelity. This also further shows the umbrella nature of HPC, because CPUs for scalar processing, VPUs for vectors, GPUs for matrices, and QPUs for tensors, all must cooperate and will remain essential for orchestration, partial or full simulation, decoding, error analysis, and verification. The work here can help shift the market's reference point beyond spectacular benchmark claims

to sustained and defensible computation. The immediate commercial impact will be limited, but moving the scientific threshold is needed. All right, that's it for this episode. Thank you all for being with us. HPC Newsbytes is a production of OrionX. Shaheen Khan and Doug Black host the show. Every episode is posted on OrionX.net. If you like the show, please rate and review it. Thank you for listening.