truly unique chipsworking at mapper    |   中文   |   English   |   

what is mapper?

Mapper designs and manufactures machines that are used for the production of computer chips (integrated circuits or ICs). The production of ICs comprises multiple successive processes for which different machines are needed. Mapper’s machines are specifically intended to draw the line structures on the ICs. The drawing of these structures is a complex process particularly because, through the years, ICs have become ever smaller, roughly the size of a peppercorn. The lines that have to be drawn on these ICs are very much smaller still, nearly a thousand times thinner than a human hair.

the limits of light

Generally, light is used to draw these structures. This process is called photolithography. To apply the structures to an IC, firstly a template is made, known as a mask. The mask is placed between the light source and the IC, and light is then shone on to the IC via the mask. This causes the line structure that has been applied to the mask to appear on the IC, scaled down many times. This process needs to be repeated a number of times. To make a fully-working IC, multiple layers need to be applied, with each layer requiring a separate mask. Complete chips often have more than sixty layers and so equally many masks are needed. The making of a mask is a technically complicated and costly process. It is therefore not surprising that lithography is the most expensive step in the IC production sequence.

The photolithography technique works very rapidly, but light unfortunately has a disadvantage. The resolution (level of detail) of the pattern is ultimately limited by the wavelength of the light, and this limit has already been reached. Attempts are nonetheless being made to continue this scaling-down process still further using ultra-advanced techniques, but these are extremely expensive to apply. The machines and masks that are used for this are very expensive. The result is that the latest generations of ICs only remain economically feasible if extremely large numbers of the same design are produced. Only manufacturers who can produce these numbers for their customers can afford such machines. This results in the paradoxical situation where, while the technique continues to make ever more things possible, these possibilities are only within the reach of an ever smaller number of IC manufacturers. The market leaders in chip-making have to concentrate on products with extremely large volumes, series of hundreds of millions of identical chips. Anyone who ‘only’ needs ten million chips, maybe with one extra function, cannot be served. To produce average or small volumes, IC manufacturers therefore have to continue to work with the less-advanced machines, and to accept their limitations. This means that the IC manufacturers’ customers (who incorporate these ICs into consumer electronics) and the consumers themselves (who use the products) are also confronted with these limitations.

mapper: volume production with electrons

Mapper’s approach is radically different. To draw the structures, Mapper employs not light but electron beams. We call this technique electron lithography. The advantages of electron beams are widely known: their resolution, depth of focus and flexibility are unequalled. The primary characteristic of the Mapper machine is that it can direct the electron beams individually so that masks are no longer needed. The structure can be designed on the computer screen to be transferred on to the IC at nano-scale. Because no mask is needed, the startup costs of a new chip design are much lower.

However, electron lithography does have an associated disadvantage. The speed of the moving electron beams is much lower than that of light, so that the IC production process is much slower. This is why electron beams are currently seldom used in IC production. Mapper is solving this problem by vastly scaling up the number of electron beams. A Mapper machine has not one electron beam but several hundred thousand. This makes the advantages of the electron technique suitable for producing larger volumes too. Mapper’s compact machines are significantly less expensive than photolithography machines. So an investment in a Mapper machine rapidly becomes attractive.

mapper’s complementarity

What makes it more interesting is that the two techniques can very well be used in parallel as they complement each other. This opens up new applications, even in existing production processes. With Mapper, the chip makers can create new applications within existing processes: for example, improve the resolution of one critical layer, increase the capacity of a small piece of integrated memory, or increase the depth of focus in a complicated layer: this simplifies the integration of different types of function, or can even make entire layers redundant. And all without expensive masks. A Mapper solution is cost-effective for both small and large production runs, because there are no startup costs for a new design. This makes small series affordable, and allows very small production runs that previously would have been impossible. In the extreme case, every chip can be made-to-measure. This opens the door to giving every chip its own physical hallmark. This hallmark could be used to protect bank cards, cars and equipment that is connected to the cloud.

For most applications, it is sufficient to employ Mapper for only a few process steps which increase the functionality of the chip significantly. Because existing processes can continue to be used for all the other steps, existing infrastructure can be re-used.

key facts and figures

Founded in May 2000 out of Delft University

268 employees as of January 1st, 2017 in three countries

300 million euros invested in technology and product development

5 kVacceleration voltage

17 µA total current on wafer level

200 mm and 300 mm wafer size

888 optical fibers connected to a single chip

1,352 electrostatic micro-lenses

66,248 parallel electron beams

3,200,000,000,000 bits per second streaming rate

26x33 mm² field size

28 nm node compatible imaging

3 nm stage positioning over full 300 mm range

2 nm beam position stability

½ nm alignment repeatability

management team

Bert Jan Kampherbeek

bert jan kampherbeek

Bert Jan Kampherbeek (CEO and co-founder) is one of the three founders of Mapper. He is responsible for the market development and positioning of Mapper in the industry. Bert Jan holds a Master's degree in Applied Physics from Delft University of Technology.

Marco Wieland

marco wieland

Marco Wieland (CTO and co-founder) is one of the three founders of Mapper. He is the key technologist of the company and an expert in electron optics and holds 28 patents. Marco is responsible for the development roadmap of the company. Marco has a Master's degree in Applied Physics from Delft University of Technology.

Guido de Boer

guido de boer

Guido de Boer (COO) joined Mapper in January 2005. Before that date he was for more than 8 years at ASML, where he held various management and systems engineering positions. Guido holds a Master's of Applied Physics from Delft University of Technology.