Reuters has RDF-D which is distributed. bloomberg has B-PIPE. Both are not as good as using an overpriced FPGA (redline, nova spark, etc) and parsing the data yourself.