There will be no Intel SUPER FAB in Spain to create chips. The government’s proposal of 11,000 million euros (almost a quarter of the EU budget for this purpose) has had no effect on Pat Gelsinger’s, but it did make its CEO look to our country to invest in development and I +D For this reason and despite the aforementioned initial failure, Barcelona will be the headquarters of the new Intel TOP laboratory worldwide.
Intel’s investment amounts to 400 million in Spain, is it enough?
Well, given the numbers that are handled in Germany, France and Italy, probably not. But we are late here and we have to start for something, so we will be positive in this regard.
The laboratory to be created with funding from both the blue company and the government is set for a period of 10 years for a value of 400 million and will represent the bastion in a sector where Intel wants to exploit: supercomputers for Zettascale.
Do not confuse Exascale with Zettascale, here we talk about the design and R&D of what would be the world’s first first-generation supercomputer. The Barcelona Supercomputing Center or BSCwill be in charge of bringing it to life after leaving the laboratory and will have the peculiarity that this supercomputer will be based on the ISA RISC-V architecture, therefore, Intel wants to investigate something totally different.
Target 2027: The world’s first Zettascale platform
The investment will be for 10 years, but Intel cannot wait that long. For this reason, it estimates that in 2027 the platform should be ready to start working on this supercomputer, something that it already had in its plans and that the BSC will make a reality by that date. Jeff McVeigh, Branch Vice President and General Manager Super Compute Group of Intel clearly expresses what is to come in our country:
High-performance computing is the key to solving the world’s most challenging problems, and we at Intel have an ambitious goal to quickly move into the Zettascale era for HPC. Barcelona Supercomputing Center shares our vision for this goal, with equal emphasis on sustainability and an open approach. We are excited to partner with them to embark on this journey.
To give a little more context to everything we have discussed, you have to understand that a ZettaFLOPS class supercomputer like the one you want to build means that you have to increase performance around 1000 times compared to the current ones, all in approximately 5 yearsthat is to say, it is a major challenge in a very short time that has a deadline.
To give even more shape to what has been said: new architectures are needed, new types of memory, new I/O interfaces, new connection systems and chip packaging, in short, a whole new generation of processors and platforms that will be the basis of a future where Intel intends to change the rules of the game in HPC. Will those be enough? 400 million that Intel is going to invest in Spain together with the government to give life to this mammoth supercomputer?
