C++ Secret vendor and open source libraries potentially used for HFT environments thanks to this Barclays Bank presentation video

(Last Updated On: December 30, 2014)

C++ Secret vendor and open source libraries potentially used for HFT environments thanks to this Barclays Bank presentation video
Please visit these helpful links with other videos which are part of these listings below:

Programming Language Lessons learning about Barclays HFT use cases including Java C++ Linux FPGA and Erlang

Barclays HFT uses with vendor panel including Solarflare and Corvil video: Decent Open source solutions?

Erlang from Barclays Bank Go Big video with weaknesses including potential RDMA open source solution

NOTES: My path of tech choices is very useful with C++/C backend with Matlab/Simulink trade idea generation into an analytics trading system. NOSQL still useful. These are all part of my Elite learning service.


Join my FREE newsletter if you want to learn more on this direction

Stats on languages for HFT trading environment job wise:

C++ 70% of hft software 25 % java but heavy garbage collection tuning so it might be turn off to focus on that 5% to function programming languages

All Linux based

More hft going to fpga for enormous performance boost

Simulink never mentioned for c translation to vhdl but we are in right track

Maxeller was highlighted

Linux Jitter was mentioned:

He kept mentioning jitter in Linux is problem to get lowest latency, using Java with garbage collection does not help

C++ Libraries


QuantLib used for option pricing Greek analysis

QuantLib, the Greeks and other useful option-related values

Boost Math toolkit


Here are some PDFs from this genius Barclays VP

CodeMesh presentation (live links included)

Click to access richard-croucher.pdf

Low Latency 101

Click to access LowLatency101.pdf

TCPIP Bypass

Click to access TCP_bypass.pdf

Example for raw Ethernet network programming


OpenOnLoad Ef_vi


RDMA description


Intel C compiler optimization (best C compiler)


Intel Xeon Optimazation Tuning


Big page and memory optimization


Linux RDMA Verbs API Tutorials


GPU Accelerated Computing with C and C++ (only pricing options Greek analysis via Monte Carlo simulation)


Assist with GPU CUDA for onloading data and program to GPU from CPU


Maxeler for FPGA  options to vhdl (where does MATLAB’s Simulink fall in her due to code generation to C C++ or VHDL)


Continues below:

Other topics from:

Low Latency 101

Click to access LowLatency101.pdf

Linux Kernal of choice potentially: https://www.kernel.org/pub/linux/kernel/projects/rt/

Personally saw big latency wins when I moved from RH 5.3 using 2.6.18 to a
2.6.27 kernel
Also refer to stack diagram on pg 23 and more lessons learned there
TCP Bypass
p 28
Application Programming with OFED
Highest performance, sometimes referred to as RAW is to use the Verbs
provided by the hardware
High cost to establish but then very high B/W at low CPU overhead e.g.
2GB transfer takes a single ‘C’ call
More from a fellow Skype user:
NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!

Subscribe For Latest Updates

Sign up to best of business news, informed analysis and opinions on what matters to you.
Invalid email address
We promise not to spam you. You can unsubscribe at any time.


Check NEW site on stock forex and ETF analysis and automation

Scroll to Top