Supercomputer for Deep Learning PhD

  1. Home
  2. /
  3. Builds
  4. /
  5. Build
seriestoo
Member since
United States
Build Views
74
Build Approvals
5
Fistbumps
0
Followers
1
Posts
0

Supercomputer for Deep Learning PhD

Personal supercomputer for my deep learning research.
Faster and more powerful than my university's supercomputer cluster.
Our research will identify plant species from satellite imagery using deep learning convolutional neural networks.

16 Core 32 Thread Threadripper 2950X on a X399 ASUS ROG Zenith Extreme Motherboard, Titan V GPU, 128gb DDR4 RAM, 10G Ethernet.

Room to expand up to 4 GPUs with the AMD Threadripper platform, 64 PCI-e lanes supporting multiple GPUs for deep learning network training.
Color(s): Black Gold Silver
RGB Lighting? Yes
Cooling: AIO Cooling
Theme: Futuristic
Size: E-ATX
Type: General Build

Contests

This build participated in 1 contest.

Hardware

CPU
$ 629.49
AMD - Threadripper 2950X
Cores: 16
Socket: TR4
Motherboard
ASUS - ROG Zenith Extreme
Chipset: X399
CPU Socket: TR4
Interfaces Available: M.2 (M), PCIe x1, PCIe x16, PCIe x4, PCIe x8, SATA 6 Gb/s, U.2, USB 3.0
Memory
G.Skill - Trident Z
Capacity: 128 GB
Type: DDR4
Graphics
$ 449.10
NVIDIA - Titan
Chip: GTX Titan
Chip Manufacturer: NVIDIA
Storage
$ 85.82
Seagate - ST4000DM000
Capacity: 4 TB
Form Factor: 3.5 Inch
Interface: SATA 6 Gb/s
PSU
$ 499.94
EVGA - SuperNOVA 1600 T2
Efficiency: 80+ Titanium
Form Factor: ATX
Modularity: Fully-Modular
Case
$ 139.98
Corsair - Carbide Air 540
Motherboard Support: ATX
Side Panel: Window
Type: Mid-Tower
Cooling
$ 112.20
EVGA - EVGA CLC 280 RGB
Compatibilty: AM3, AM4, FM1, FM2, LGA 1150, LGA 1151, LGA 1155, LGA 1156, LGA 2011, LGA 2011-3, TR4
Height in mm: 196
Type: AIO
Unable to dertermine total build value (missing part value)
74
5
Approved by: