Germany picks six consortia in its €640m race to build fault-tolerant quantum computers

[Electric Brad]

Germany's Federal Ministry of Research, Technology and Space has chosen six consortia to move forward in a major hardware competition aimed at fault tolerant quantum computing, Quantum Computing Report says. The programme has up to 640 million euros across all its hardware tracks. Its stated goal is to deploy "at least two state-of-the-art fault-tolerant quantum processing units (QPUs) by 2030". That is one of the most specific national targets we have seen anywhere

The competition covers three technologies, with two consortia in each: neutral atoms, trapped ions and superconducting circuits. Every selected group has to show a working entry system by 31 March 2027, which is a tight deadline for hardware of this kind. Picking two teams per technology means Germany is hedging its bets rather than choosing a single winner. Only two of the six consortia are named in the article

One is LOGIQC, led by Munich startup planqc, which we discussed at the weekend after it won a Made in Europe award. Its partners include TOPTICA, the Max Planck Institute of Quantum Optics, LMU Munich, Forschungszentrum Jülich and the University of Tübingen. They are building an error corrected neutral atom processor based on arrays of ytterbium atoms, along with a pilot line for the lasers and optics it needs

The other is NFQC-1k, a 122 million euro trapped ion project led by AQT Germany with Leibniz University Hannover, TU Braunschweig, the national metrology institute PTB and Quantum Valley Lower Saxony. It aims for 1,000 physical qubits, 50 logical qubits and logical gate error rates below 0.01 percent. Those are ambitious numbers, and hitting them by 2030 would put Germany near the front of the pack

Is running two consortia per technology a smart hedge or a waste of money? And how does this compare with the UK's approach of buying machines through ProQure?

Germany Advances Key Consortia in €640 Million ($717.2 Million USD) Fault-Tolerant Quantum Computing Competition - Quantum Computing Reportquantumcomputingreport.com

[Danny47]

Fifty logical qubits with error rates below 0.01 percent would be a big deal. That is the sort of machine that could start doing useful chemistry. Whether trapped ions can scale to 1,000 physical qubits in that time is the big question

[Scholes17]

Six hundred and forty million euros sounds a lot, but spread across six consortia it is not huge by quantum standards. PsiQuantum alone has raised billions

Europe still struggles to match American scale. Still, focused spending with clear milestones can achieve a lot. Hopefully the 2027 checkpoint keeps everyone honest

[Slate Mike]

Our approach in the UK is different. We are funding companies and then buying their machines, while Germany is funding consortia that mix universities and companies

Both have their merits. I would like to see which produces working hardware first

[Karen76]

planqc appearing again so soon after the award shows how quickly they are becoming a national champion. LOGIQC has some heavyweight partners as well. Winning an award one week and a big government contract the next is good timing

[Slay40]

Smart hedge, in my view. Nobody knows which technology will win, so backing several makes sense

Germany has the money and the research base to do it properly. It is also how DARPA's benchmarking programme works, with several companies in each stage

[Felicity]

A working entry system by March 2027 is a brutal deadline. Some of these consortia will struggle to hit it. That might be the point, though, as a way to filter out the weaker teams early

[SilverCurrent]

Pilot lines are an interesting detail. Germany seems to be thinking about manufacturing, not just research prototypes. That is where Europe has often fallen behind in the past