T 1072/11 (Matching unit comprising two computer entities directly connected to a shared memory storing pre-calculated values/NASDAQ) 18-06-2019

Introductory remarks

1. The application was refused based on lack of inventive step in view of D1 (US 6098051 A1), a publication introduced during the examination proceedings with the summons for oral proceedings before the first instance. However, no complete search for pertinent prior art was carried out before the first instance. The European Patent Office acting as International Searching Authority issued a declaration of non-establishment of the International Search Report under Article 17(2)(a) PCT and Rule 13ter.1(c) and 39 PCT, because the claims on file at that date related to subject-matter that did not require an international search under the PCT provisions. The applicant entered into the European phase before the European Patent Office on the basis of amended claims. No supplementary European Search Report was established. According to Article 153(6) EPC the declaration replacing the International Search Report took the place of the European Search Report in analogy to Rule 63 EPC.

The present main request corresponds to a set of claims submitted during oral proceedings before the first instance as an auxiliary request adding further features. The Board has no indication that a systematic search for pertinent prior art has been carried out for the claimed subject-matter according to the present main request.

Main request

2. Article 56 EPC - Inventive step

The Board does not agree with the decision under appeal that the subject-matter of independent claims 1, 6, 11 and 12 lacks an inventive step over the disclosure of D1.

2.1 The independent claims are directed to a mix of technical and non-technical features. The Board does not dispute that they appear in a technical context. Even the method can be considered to be performed by technical means, because it involves a first and second computer processor as well as a memory, i.e. means for processing and storing data and, therefore, has technical character. Accordingly, the claimed subject-matter is an invention in the sense of Article 52(1) EPC (see T 258/03 "Auction method/HITACHI").

2.2 However, the question of inventive step requires an assessment of whether the invention makes a technical contribution over the prior art. Features which do not make such a contribution cannot support the presence of an inventive step (see T 641/00 "Two identities/COMVIK", Headnote I).

2.3 The Board is not convinced by the appellant's argument that the matching of combination orders implied certain technical constraints that combined with the other technical features had to be accounted for by the skilled person when designing the claimed system (see point 3.2.7 of the statement setting out the grounds of appeal).

First of all, D1 deals with combination orders (see e.g. column 3, lines 33 to 36 and column 4, lines 36 and 37). The alleged differences in the concept of how to match such combination orders in comparison to the claimed subject-matter are in the non-technical domain. The concept of matching combination orders involves constraints originating from the financial domain such as that they have to be matched simultaneously (see D1, column 4, lines 50 to 57). Those constraints are provided by the financial expert to the technically skilled person, a computer expert provided with the complete description of the non-technical abstract financial matching concept, for implementation.

2.4 D1 discloses the following technical features of claim 1:

an automated exchange system for receiving orders to be matched, comprising - a first computer entity (see figure 1, central matching controller CMC) associated with a first computer processor (see column 6, lines 48-51; figure 1), wherein received combination orders (see column 3, lines 33 to 36 and column 4, lines 36 and 37; see also column 4, lines 49 to 56 baskets of securities with corresponding profiles to be linked together) are matched; - a second computer entity (see figure 1, trader CPU 10, 12 or 14) associated with a second computer processor (see column 6, line 64, computers or workstations) wherein data values (satisfaction density profile) required in the matching are provided (column 4, lines 49-61); - a memory (see figure 1, storage device 4) directly connected to the first computer entity.| 2.5 D1 does not disclose the following features of claim 1: - the second entity is located within the matching unit,- a calculation of matching data values required in the matching process by the second computer entity,- a shared memory directly connected to the second computer entity and- storing calculated matching data values calculated by the second computer entity in the shared memory. 2.6 The underlying objective technical problem is considered to be to implement the financial matching concept in a way which improves real-time and latency constraints in the matching unit. 2.7 The technical effect achieved by the claimed solution is that data values required by the first computer entity when performing matching operations are pre-calculated by the second computer entity and are stored for potential use. According to D1, however, there are structural and functional differences. Regarding the structure of the claimed solution D1 discloses neither incorporating the second computer entity in the matching unit, nor a shared memory directly connected to both computer entities. According to D1, a memory (database 4 in figure 1) is only connected to the matching unit CMC. The trader terminals 10, 12 and 14 are separate from the CMC and each comprise a CPU since they are personal computers or workstations (see D1, column 6, line 64). Regarding the function in D1, the trader CPUs are for entering satisfaction density profiles (see D1, e.g. column 11, line 10). Such satisfaction density profiles cannot be considered to correspond to matching data. There are no calculations with regard to performing matching operations. Rather, satisfaction density profiles in D1 correspond to orders (sell or buy) which are merely entered and transmitted to the CMC where all the matching operations are performed. Hence, no pre-calculated matching data values are provided by the trader CPUs. Furthermore, the memory in D1 is only for storing satisfaction density profiles, i.e. the orders, but not for storing pre-calculated matching data values. With regard to the objective technical problem posed, D1 handles complex combination orders by creating a connection matrix which is used for performing matrix manipulations on each of the different sub-orders/legs of the combination order before being able to actually match such orders (see D1, column 4, lines 37 onwards, calculation of mutual satisfaction cross product profiles for every buy/sell profile which are matched in order and matching trades are aggregated by the CMC). Regarding the complexity of operations and real time constraints, D1 merely suggests using fast computers with a single processor (see D1, column 6, line 52, supercomputer). The Board therefore concurs with the appellant's arguments (see e.g. point 5.2.1.5 of the statement setting out the grounds of appeal). 2.8 Even if one were to follow the argumentation in the contested decision that it was obvious for the skilled person in view of D1 and the common general knowledge to make use of a multiprocessor system with a centralised architecture (see point 10.5 of the decision), the functional differences identified above would still remain. In the Board's view, the subject-matter of claim 1 is therefore not rendered obvious by the disclosure of D1 alone. Similar arguments apply, mutatis mutandis, for corresponding independent claims 6, 11 and 12. 3. The Board is aware of computer processors comprising a coprocessor as well as first and second level cache memory, which have been used for special tasks already before the priority date of the present application, including improving processing power and dealing with real time constraints. However, the Board is not convinced that this has been notorious knowledge of the skilled person. Therefore a look into the prior art is necessary. 3.1 As has already been outlined above (see point 1), no Search Report was established for the application and D1 was introduced during substantive examination with the summons for oral proceedings before the first instance. Since a corresponding set of claims for the subject-matter of the present main request has been presented only at a later stage during oral proceedings before the first instance, i.e. at the end of the first instance proceedings, there is no indication that a systematic search for pertinent prior art has been carried out for the present set of claims. 3.2 However, according to the jurisprudence of the Boards an additional search for pertinent prior art may be dispensed with only if the technical features of the claims are considered to be "notorious", i.e. generic and so well known that they cannot reasonably be refuted (see T 1411/08 of 6 June 2011, point 4; T 2299/10 of 31 March 2014, point 4.5). In the Board's judgment, the aforementioned distinguishing features (see points 2.5 and 2.8) go beyond the mere notorious knowledge (for example the commonly encountered use of a computer, a network, an electronic database) and cannot be considered "notorious". In particular, it is not considered to be notorious knowledge to use a second computer entity within a first computer entity both directly connected to a shared memory which stores pre-calculated data. 3.3 Thus, present claim 1 cannot be definitively assessed for inventive step without knowledge of the relevant documented prior art. Thus, the main request requires a search for relevant prior art. Hence the matter must be remitted for an additional search and further examination. 4. According to Article 111(1) EPC the Board may exercise any power within the competence of the examining division (which was responsible for the decision under appeal) or remit the case to that department for further prosecution. It is thus at the Board's discretion whether it examines and decides the case or whether it remits the case to the department of first instance. As it appears necessary for a further search to be carried out, the Board considers that in the present case remittal is the more appropriate course of action. Since the decision can be taken on the basis of the main request, the Board does not have to deal with the first to third auxiliary requests. |