AIToday
AI Business & IndustrySiliconANGLE AIPublished: Nov 22, 2026, 10:01 JST

HammerHAI chief: sovereignty starts with data, not chips

Send on LINE
HammerHAI chief: sovereignty starts with data, not chips

3 Key Points

  1. What happened

    Bastian Koller of the High-Performance Computing Center Stuttgart said HammerHAI, built on Hewlett Packard Enterprise Co. and Nvidia Corp. Grace Blackwell systems, treats sovereignty as data control, since full independence down to European chips is 'not realistic in the near term.'

  2. Why it matters

    The center designs its services so workloads can move to a hyperscaler and back, meaning customers get a performance boost without being locked in, according to Koller.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

HammerHAI is one of several publicly funded AI factories across Europe racing to give startups and small manufacturers access to supercomputer-class GPUs. The Stuttgart center has already worked with customers on existing systems while it waits for the new hardware, and several have produced success stories. Koller's pragmatic stance reflects a trade-off: rather than waiting for a sovereign processor or GPU, the center focuses on data control and portable workloads. The concierge model, which pairs each customer with a consistent contact, is presented as a way to learn how users work and avoid building systems nobody uses. The center runs an ISO-certified environment with TISAX Level 3 certification for automotive work, signaling attention to security and access demands. Its efficiency story—doubling performance at about 700 kilowatts versus a 3.2-megawatt predecessor—suggests the new system offers significant power savings, though the body does not quantify the ratio.

FAQ
What hardware does HammerHAI use?
It is taking shape around a system from Hewlett Packard Enterprise Co. using Nvidia Corp. Grace Blackwell chips.
How does HammerHAI prevent lock-in?
It designs its services so workloads can move from the center to a hyperscaler or back, according to Koller.
What efficiency gains does the new system offer?
The new system doubles the performance of its 3.2-megawatt predecessor at about 700 kilowatts, and the center reuses its waste heat.
SiliconANGLE AIRead Original Article
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articletaiziii survey: 56.5% of managers have abandoned AI tools