As an environment for creativity, you can't beat IBM's Almaden Research Laboratory in the golden, oak-covered hills above San Jose, California.
The lab buildings are calm and encouraging, with wood-panelled halls and wisteria-covered patios where barbeques or informal al-fresco lunches are held. The setting is also spectacular, located in the middle of a 690-acre former cattle ranch that cascades over hillsides and looks towards California's redwood-forested coastal mountains.
In an unusual deal with local authorities, IBM bought the ranch in the 1980s, but was only permitted to develop 10 acres for its own use.
The rest - inhabited by cattle, deer, wild turkey and circling hawks and eagles, coyote and bobcats (signs on the fences also warn against rattlesnakes) - is partly maintained as a public access area for hiking and partly as a nature reserve. IBM describes it as a sophisticated summer camp for "some of the world's most creative minds".
But it's not all big country - big science gets done here, too, with the 500-person lab at the centre of IBM research into nanoscience (science at the atomic and molecular level); human-computer interfaces; document and information management, databases and experiments with new materials for storage systems and microprocessors.
The lab prides itself on having been the place where relational databases were first described in a research paper; where the computing language Fortran was invented, and where researcher Don Eigler first moved individual atoms, proving in theory that tiny nanostructures built of atoms could be created (famously, he lined up xenon atoms to form a tiny IBM logo, imaged on a microscope).
The lab does "world class science to improve mind share, market share and ultimately the profit for IBM products," says researcher Gian Luca Bona. "But such advancementsusually don't happen in months; it can take years. That's the advantageof doing long-term research."
That means some projects at Almaden seem esoteric because they are so distanced from a real-world use or a specific product to go on a shelf. Researchers who recently took some time out to describe their projects to laypeople hada difficult task, because the work can be so complex and borders on the edge of the believable.
Almaden is a place where the mind can easily boggle. For example, spend 30 minutes with IBM's Dr Stuart Parkin and every cliché of the brainbox mad scientist is confirmed. The energetic and amiable Parkin, originally from Cambridge University,speaks in a rapid manner and makes excited gestures. To top it all off, he works in a lab filled with an extraordinary three-part contraption of chambers, tubes, wires and mechanics, which test new materials.
Given his enthusiasm for his subject, Parkin interests visitors in his research area, "spintronics", which he sums up as the study of "whether we can create new devices by taking advantage of the spin of electrons".
The spin produces magnetic qualities and if a current of electricity is passed through the magnetic field, the direction of the spin changes. This enables the construction of magnetic disk drives for storage, and spintronics, he says, is the science underlying the explosion in capacity of storage drives in recent times as well as the drop in price - and, therefore, has helped develop the internet.
His group is exploring whether spintronics can also be used for Ram (random access memory), the main component in the brains of a computer. Unusually for a research lab, this research is already on its way to commercialisation and magnetised Ram will go into limited production next year, Parkin says.
Down the hallway, other teams are using specially-built microscopes to examine the structures of new materials at the atomic level, or to build tiny working computer circuits atom by atom.
But many projects are more accessible. For example, work on databases, security systemsand intelligent document managementhas IBM championing a method of handling health records that it hopes will cut down on patient deaths due to mix-ups over treatment and make national health programmes more manageable.
There's research into inputting information on plain paper forms so that it can be automatically stored in a database. A new system for writing on handheld devices, called Shark, looks like it will be going commercial soon, and insights into how people read webpages and absorb information is changing the way human-computer interfaces are designed.
"Almaden is a very good environment to do things in and have an impact," says Parkin, leaning back in a chair in his office with its lovely hillside view.