C++ Fix Engine

AIM Systems - Fix Solution

FIX Engine: Latency & Performance

Performance and full FIX compliance are distinguishing features.  Following are detailed performance analysis with descriptions. We have tried to provide maximum details for performance / latency statistics and would be glad to answer any questions.


Performance Results (Time is in Nano Seconds)
Message Size (Bytes)

Tags #

(Header, Body,Trailer)



Latency / tag


Decoding 100 10(5+4+1) 203 20 4,926,108
Decoding 160 15(6+8+1) 302 20 3,311,258
Encoding 100 10(5+4+1) 320 32 3,125,000
Encoding 160 15(6+8+1) 487 32 2,053,388



100 10(5+4+1) 309 * 3,236,246



160 15(6+8+1) 443 * 2,257,336

Message Count for each Iteration: 10,000,000

For Gateway mode, 4 Fields were check for Risk. Please refer code for deatils.

**Gateway Flag enabled: All checkes are performaned in read-only mode, without any enrichment.

De-Serialization - From byte array to FIX Object

  • Sample shows performance to convert std::string or byte array to FIX Message Object
  • Time includes enriching message with all Field Tags, with numeric ID and value
  • No delayed parsing or ID conversion
  • All field are populated in respective Header, Body or Trailer

Serialization - From FIX Object to byte array

  • Rebuild String object from all Field.
  • Checksum can be optionally enabled which takes under 50 nano-seconds, disabled in above test


      Operating System: CentOS Linux X64, version 6.4
      Memory: 32 GB
      CPU: 3.2 Ghz, I7-3930   
      Overclocking - NONE
      GCC/ G++  version: 4.8.2
      GLIBC:  x.x


Sample Messages used in performance testing - NewOrderSingle

    100 Byte Message
    160 Byte Message


Performance comparison with other Open-source implementations


Latency AIMS FIX


Latency Quick FIX


Decoding Time Per Message (98 bytes) 203 4570
Encoding Time Per Message ( 98 bytes) 320 13,650

Performance Code

Sample Performance Code used for testing


  • Serialization / deserialization per tag is consistent irrespective of message size.
  • aimsFIX library outperform known open-source by huge magnitude
  • Under Gateway mode, it gives best performance, under 300 nano-sec