1. A process for isomerising paraffins in a feed comprising hydrocarbons containing 5 to 8 carbon atoms per molecule, characterized in that said feed is separated into at least two fractions, a fraction A mainly comprising hydrocarbons containing 5 or 6 carbon atoms, and a fraction B mainly comprising hydrocarbons containing 7 or 8 carbon atoms, and in which said fractions A and B are then separately treated under specific conditions in distinct isomerisation reaction zones wherein said reaction zones contain chemically different isomerisation catalysts in which the isomerisation activity of the catalyst treating fraction A is greater than the isomerising activity of the catalyst treating fraction B, in which at least one of the isomerisation reactions is carried out in the presence of an excess of hydrogen such that the ratio R of the number of moles of hydrogen to the number of moles of hydrocarbons leaving at least one reaction zone is in the range 0.06 to 0.3, and in which at least one of the isomerisation reactions is carried out in an excess of hydrogen such that the ratio R of the number of moles of hydrogen to the number of moles of hydrocarbons leaving at least one reaction zone is between 0.3 and 30.
2. A process according to claim 1, in which the isomerisation catalyst or catalysts comprise any from a group consisting of: supported catalysts containing at least one halogen and at least one group VIII metal, zeolitic catalysts containing at least one group VIII metal, Friedel-Crafts catalysts, superacid HPA on zirconia catalysts, tungsten oxides, on zirconia catalysts, and sulphated zirconias.
3. A process according to claim 1, in which the total pressure in the isomerisation reaction zones is about 0.1 to 10 MPa relative, the hourly space velocity being about 0.2 to 10 \u22121.
4. A process according to claim 1, in which the isomerates from isomerising fractions A and B are then mixed.
5. A process according to claim 4, in which said mixture is separated into two effluents, the first of said effluents comprising the major portion of the isopentane and paraffins with at least two branches, the second of said effluents comprising the major portion of the normal-paraffins and paraffins containing at least six carbon atoms and a single branch, said second effluent being recycled and mixed with the feed or with fraction B.
6. A process according to claim 1, in which isomerate B derived from isomerising fraction B is separated into two effluents, the first of said effluents comprising the major portion of the isopentane and paraffins with at least two branches, the second of said effluents comprising the major portion of the normal-paraffins and paraffins containing at least six carbon atoms and a single branch, said second effluent being recycled and mixed with the feed or with fraction B.
7. A process according to claim 6, in which the first effluent is mixed with the isomerate derived from isomerising fraction A.
8. A process according to claim 6, in which isomerate A derived from isomerising fraction A is separated into two effluents, the first of said effluents comprising the major portion of the isopentane and paraffins with at least one branch, the second of said effluents comprising the major portion of the normal-paraffins, said second effluent being recycled and mixed with the feed or with fraction A.
9. A process according to claim 8, in which said first effluent from isomerate B is mixed with said first effluent from isomerate,A.
10. A process according to claim 1, wherein the catalyst for treating fraction A is a chlorinated alumina-based catalyst and the catalyst used for treating fraction B is a less acidic zeolite PtH-beta catalyst.
11. A process according to claim 1, wherein at least one catalyst is other than a Friedel-Crafts catalyst and is selected from the group consisting of supported catalysts containing at least one halogen and at least one group VIII metal, zeolitic catalysts containing at least one group VIII metal, superacid HPA on zirconia catalysts, tungsten oxides, on zirconia catalysts, and sulphated zirconias.
The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.
1. A method of determining adapter affinity in a high performance computing (HPC) environment of the type including a plurality of distributed computing components defining a plurality of locations and a plurality of inputoutput (IO) adapters, with each distributed computing component including at least one processing element, and with each IO adapter coupled to a distributed computing component among the plurality of distributed computing components, the method comprising:
for each of a plurality of candidate mappings that speculatively map at least one IO adapter to at least one location among the plurality of locations, running a performance test for a task executed by a processing element in a distributed computing component at a first location among the plurality of locations, wherein the plurality of candidate mappings includes first and second candidate mappings, wherein the first candidate mapping maps a first IO adapter among the plurality of IO adapters to the first location, and the second candidate mapping maps a second IO adapter among the plurality of IO adapters to the first location, and wherein running the performance test respectively generates first and second test results for the first and second candidate mappings; and
generating speculative adapter affinity information that assigns at least one IO adapter to at least one location among the plurality of locations based upon the performance test run for each of the plurality of candidate mappings, including assigning the first IO adapter to the first location based upon a comparison of the first and second test results for the first and second candidate mappings.
2. The method of claim 1, wherein the HPC environment is of a type where preconfigured adapter affinity information is unsupported.
3. The method of claim 1, wherein the task is a first task, the method further comprising:
storing the speculative adapter affinity information; and
during initialization of a second task to which is allocated the first IO adapter:
binding the second task to at least one processing element in the distributed computing component at the first location;
determining that the at least one processing element is at the first location;
determining that the first IO adapter is assigned to the first location based upon the stored speculative adapter affinity information; and
selecting the first IO adapter as a primary adapter for the second task based upon determining that the first location is common to the at least one processing element and the first IO adapter.
4. The method of claim 1, wherein the plurality of distributed computing components includes a plurality of multi-chip modules (MCM’s), wherein each IO adapter is coupled to an MCM among the plurality of MCM’s, and wherein assigning the first IO adapter to the first location indicates that the first IO adapter is coupled to a first MCM associated with the first location.
5. The method of claim 1, wherein the task is a first task, the distributed computing component is a first distributed computing component and the processing element is a first processing element, wherein running the performance test for the first task includes communicating data from the first task to a second task executed by a second processing element in a second distributed computing component at a second location among the plurality of locations, wherein the first candidate mapping maps the second IO adapter to the second location, and the second candidate mapping maps the first IO adapter to the second location.
6. The method of claim 5, further comprising, for each of the plurality of candidate mappings, running a performance test for the second task by communicating data from the second task to the first task using each of the first and second candidate mappings.
7. The method of claim 6, wherein generating the speculative adapter affinity information assigns the second IO adapter to the second location based upon a comparison of test results generated from the performance test for the second task.
8. The method of claim 6, further comprising fixing processing element bindings for the first and second tasks such that the performance test performed for each candidate mapping and for each of the first and second tasks is performed using fixed processing element bindings.
9.-20. (canceled)