3. 3
IoT Hardware Platform
• Heterogeneous, modular, scalable microserver system
• Supporting the full spectrum of IoT from embedded over the edge towards the cloud
• Different technology concepts for improving
x86
GPU
ML-ASIC
ARM v8
GPU
SoC
FPGA
SoC
RISC-V
FPGA
VEDLIOT Cognitive
IoT Platform
• Performance
• Cost-effectiveness
• Maintainability
• Reliability
• Energy-Efficiency
• Safety
4. 4
RECS Architecture (RECS|BOX)
RECS Server Backplane (up to 15 Carriers)
Carrier (PCIe Expansion)
Carrier (High Performance)
e.g. GPU-Accelerator
Carrier (Low Power)
#3
#2
Microserver
(High Performance)
#1
Microserver
(Low Power)
#16
#3
#2
Microserver
(Low Power)
#1
High-Speed Low-Latency Network (PCIe, High-Speed Serial)
Compute Network (up to 40 GbE)
Management Network (KVM, Monitoring, …)
HDMI/USB
iPass+ HD
QSFP+
RJ45
Ext. Connectors
GPU SoC
FPGA SoC ARM Soc
Low-Power Microserver
(Apalis/Jetson)
x86 ARM v8
High-Performance Microserver (COM Express)
FPGA SoC
High-Performance
Carrier
(up to 3 microservers)
Low-Power Carrier
(up to 16 microservers)
5. 5
IoT hardware platform
Edge Server (t.RECS)
▪ Usage in VEDLIoT
▪ Edge Server
▪ CPU
▪ FPGA
▪ NVIDIA GPU
▪ Use-Cases
▪ Smart Home
▪ Automotive
t.RECS Edge Server
Microserver
(COM-HPC Client)
Microserver
(COM-HPC Client)
Microserver
(COM-HPC Server)
Switched PCIe (up to 56 Gbit/s)
External
interfaces
PCIe
expansion
Ethernet (up to 10 GbE)
Management Network (KVM, Monitoring, …)
I/O (Camera, Display, Radar/Lidar, Audio)
6. 6
Development of new IoT hardware platform
Accelerated IoT Device (u.RECS)
▪ u.RECS design based on Use-Case requirements
▪ Communication (GbE, SPE, LoRa, 4G/5G)
▪ Wide Input-Power-Range
▪ Supports combinations of CPUs, FPGAs, GPUs and accelerators
▪ SMARC modules (x86, ARM, FPGA)
▪ NVIDIA Nano (Xavier, Orin)
▪ M.2 PCIe accelerators (Hailo-8, Coral, Myriad)
▪ Support for high-speed power monitoring
7. 7
COM-HPC
Server Size D
Intel Xeon D
COM-HPC (*)
Jetson AGX
NVIDIA Xavier
COM Express
ARM v8 Server
SoC Hi1616
COM Express
Intel Stratix 10
Jetson nano
NVIDIA
Xavier NX
COM-HPC Size B
NVIDIA Xavier
AGX
COM-HPC
client size B
Xilinx Zynq
UltraScale+
COM Express
Xilinx
Zynq
7045
COM Express
AMD Ryzen
V1807B
COM Express
AMD EPYC
3451
COM-HPC
client size A
Intel Core i7
12th Gen
CPU
FPG
A
SoC
GPU
SoC
ML
SoC
Google
Coral SoM
SMARC
Xilinx Zynq
UltraScale+
SMARC
Coherent
Logix HX40416
SMARC 2.1
NXP i.MX 8M
(4x Cortex-A53)
SMARC 2.1
Intel Atom
Jetson TX2
NVIDIA
Tegra X2
GAPuino
Greenwaves
GAP8
Syntiant
NDP120
Microserver Overview
Google Coral TPU
Dual chip
Hailo-8
ML
Accelerators
Intel Myriad X
Dual chip
9. 9
▪ Sustainable
▪ COM standards from industry
▪ Newest Architecture continuously be integrated
▪ Combination of various architectures (CPU, GPU, FPGA, ASIC)
▪ High-Speed communication matches requirements of tomorrows applications
Microserver Overview
Griessl, R., Porrmann, F., Kucza N., et. Al. Evaluation of heterogeneous AIoT Accelerators within VEDLIoT (in April 2023). Design, Automation & Test in
Europe Conference & Exhibition (DATE23).
10. 10
Thank you for your attention.
Contact
Rene Griessl
Bielefeld University, Germany
rgriessl@techfak.uni-bielefeld.de
11. 11
▪ IoT Microservers:
▪ Availability of new architectures
▪ Does the selection make sense
▪ ANT tools Kenning & Renode: How to lower entry level for new users?
Questions