New 2018: Pictures Of Your Trading Stations

post some of your trades as they unfold, i have

You probably do that to boost your management consultantancy activities. That's why you mention you are a management consultant. That's irrelevant information when it comes to trading. I don't need that. I only trade own money and have no other commercial activities. I don't want them and I don't need them.
 
  • Like
Reactions: d08
y'all are a bunch of plebs. Get on my level
4790034.jpg
 
Looks all a bit chaotic but here goes:

1) Workstation : Intel i9-7900X 10 core, 128GB DDR4, Titan RTX for AI compute, Nvidia NVS 810 to drive 6 of the 8 monitors, 100Gb Mellanox ConnectX-4 network card to connect to other machines via 100Gb network

2) Supermicro Fileserver : Xeon W-2123, 32GB DDR4 ECC, 2 4TB mirrored backup HDDs, 4x 1TB striped Samsung 970 Evo for main high performance storage, 100Gb Mellanox ConnectX4 network card.

3) a) Cisco RV-325 1Gb router for management ports, internet, b) Mellanox SN2700 switch with 32x 100Gb Qsfp28 ports, c) Asus ESC8000 HPC GPU Compute Server with 2x Intel Xeon Gold 6154, 196 GB DDR4 ECC, 4x Titan RTX GPU cards, 100Gb Mellanox ConnectX-4 network card.

Horrific cabling, I know, I do not even have a server rack yet due to space constraints and because I will be moving soon anyway. Total overkill for most but I spend a lot of time on algorithmic trading model development and training deep learning models hence the specific hardware.

IMG_20190322_182123.jpg
IMG_20190322_182201.jpg
IMG_20190322_182414.jpg
 
Looks all a bit chaotic but here goes:

1) Workstation : Intel i9-7900X 10 core, 128GB DDR4, Titan RTX for AI compute, Nvidia NVS 810 to drive 6 of the 8 monitors, 100Gb Mellanox ConnectX-4 network card to connect to other machines via 100Gb network

2) Supermicro Fileserver : Xeon W-2123, 32GB DDR4 ECC, 2 4TB mirrored backup HDDs, 4x 1TB striped Samsung 970 Evo for main high performance storage, 100Gb Mellanox ConnectX4 network card.

3) a) Cisco RV-325 1Gb router for management ports, internet, b) Mellanox SN2700 switch with 32x 100Gb Qsfp28 ports, c) Asus ESC8000 HPC GPU Compute Server with 2x Intel Xeon Gold 6154, 196 GB DDR4 ECC, 4x Titan RTX GPU cards, 100Gb Mellanox ConnectX-4 network card.

Horrific cabling, I know, I do not even have a server rack yet due to space constraints and because I will be moving soon anyway. Total overkill for most but I spend a lot of time on algorithmic trading model development and training deep learning models hence the specific hardware.

View attachment 199369 View attachment 199370 View attachment 199371

The codes are trading related?

Are they in Java or C#?
 
Full time for 13 years now of which 6 at banks and hedge funds and 7 independent.

Hardware wise my recommendation is to only get what you truly need, some of those components depreciate very fast. Yet, the gpu compute server is a fraction of the cost of what I would have to pay Google or Amazon, the breakeven to own given the utilization was about 4 months.

Cool setup. Are you a full time trader?

I would like to have a similar set up someday when I go full time. Currently not enough capital to go full time.
 
Looks all a bit chaotic but here goes:

1) Workstation : Intel i9-7900X 10 core, 128GB DDR4, Titan RTX for AI compute, Nvidia NVS 810 to drive 6 of the 8 monitors, 100Gb Mellanox ConnectX-4 network card to connect to other machines via 100Gb network

2) Supermicro Fileserver : Xeon W-2123, 32GB DDR4 ECC, 2 4TB mirrored backup HDDs, 4x 1TB striped Samsung 970 Evo for main high performance storage, 100Gb Mellanox ConnectX4 network card.

3) a) Cisco RV-325 1Gb router for management ports, internet, b) Mellanox SN2700 switch with 32x 100Gb Qsfp28 ports, c) Asus ESC8000 HPC GPU Compute Server with 2x Intel Xeon Gold 6154, 196 GB DDR4 ECC, 4x Titan RTX GPU cards, 100Gb Mellanox ConnectX-4 network card.

Horrific cabling, I know, I do not even have a server rack yet due to space constraints and because I will be moving soon anyway. Total overkill for most but I spend a lot of time on algorithmic trading model development and training deep learning models hence the specific hardware.

View attachment 199369 View attachment 199370 View attachment 199371

Why do you have so much hardware at home instead of the cloud? Isn't latency an issue in SE Asia as it seems you're sending orders from home.
 
I don't trade live from any machines at home. I colocate at 2 ECNs and proximity locate with 2 other brokers. Above setup is for research and development only, renting gpu compute in the cloud is prohibitively expensive.

Why do you have so much hardware at home instead of the cloud? Isn't latency an issue in SE Asia as it seems you're sending orders from home.
 
  • Like
Reactions: d08
Back
Top