• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Launches Ryzen Embedded 8000 Series Processors with Integrated NPUs for Industrial AI

GFreeman

News Editor
Staff member
Joined
Mar 6, 2023
Messages
1,099 (2.70/day)
AMD has introduced the Ryzen Embedded 8000 Series processors, the first AMD embedded devices to combine NPUs based on the AMD XDNA architecture with traditional CPU and GPU elements, optimized for workload versatility and adaptability targeting industrial AI applications. Embedded solution engineers and developers can harness the processing power and leadership features for a variety of industrial AI applications including machine vision, robotics, and industrial automation. AI is widely used in machine vision applications today to enhance quality control and inspection processes.

AI can also help robots make real-time, route-planning decisions and adapt to dynamic environments. In industrial automation, AI processing helps intelligent edge devices perform complex analysis and decision-making without relying on cloud connectivity. This allows for real-time monitoring, predictive maintenance, and autonomous control of industrial processes, enhancing operational efficiency and reducing downtime.



AMD RYZEN EMBEDDED 8000 PROCESSORS IN DETAIL
AMD embedded products have traditionally leveraged AMD Radeon graphics processing coupled with AMD CPUs, where the GPU has been be used for AI and workload acceleration. However, for next-generation embedded designs, the XDNA architecture and its powerful AI engines are leveraged on the new AMD Ryzen Embedded 8000 Series processors to augment AI performance.

With these resources in play, the full processor performance can scale up to 39 TOPS (trillion operations per second), with up to 16 TOPS offered by the NPU - a powerful profile targeted for demanding, diverse industrial computing workloads. The increased TOPS is achieved by the combination of AMD architectures, AMD XDNA in the NPU, AMD RDNA 3 in the GPU, and "Zen 4" in the CPU, that work together to optimize overall system performance.

AMD Ryzen Embedded 8000 Series processors harness the next-generation AMD "Zen 4" 4 nm core CPU architecture, providing powerful x86 CPU compute with up to 8 cores (16 threads). Process node advancements have extended performance-per-watt metrics achieved with previous generation Ryzen Embedded Series devices, enabling AI processing in a comparable power envelope. Supporting scalable thermal design power (TDP) profiles spanning from 15 W to 54 W, AMD Ryzen Embedded 8000 Series processors deliver a major step forward in industrial compute efficiency for AI applications.

ADVANCED CAPABILITIES, OPEN ECOSYSTEM
With the help of AMD RDNA 3, a single, small-footprint AMD Ryzen Embedded 8000 Series processor can power up to four independent displays at 4K resolution, leveraging hardware decoding and encoding for popular video codecs including AV1, H.264 and H.265.

They also deliver 20 lanes of on-chip PCIe Gen 4, providing ample I/O flexibility for specialized accelerators and other add-ons. AMD Ryzen Embedded 8000 Series processors also feature the latest dual-channel memory support (DDR5 up to 5600MT/s) with support for error-correcting code (ECC) that enables efficient, reliable memory transfer.

This robust set of capabilities is complemented by an open, royalty-free software suite of libraries and inferencing models, and a software development kit (SDK) to optimize and streamline development and deployment. The solution supports the Windows operating system today. It also supports popular AI frameworks including PyTorch and TensorFlow, featuring pretrained models available on HuggingFace.

AMD is partnering with embedded ODMs, including Advantech, ASRock, and iBASE to build a diverse ecosystem of innovative platforms supporting AMD Ryzen Embedded 8000 Series processors. Additionally, AMD Ryzen Embedded Series processors offer extended product availability to help support long-term design lifecycles.

The introduction of the AMD XDNA architecture and its powerful AI engines in select AMD Ryzen Embedded products - marks an exciting new era for these applications opening up AI opportunity in the embedded sector.

View at TechPowerUp Main Site | Source
 
Joined
May 3, 2018
Messages
2,270 (1.04/day)
LOL, hard to see how the current piss weak NPU's will do much for AI in industry. GPU is far more powerful currently and I doubt power is an issue. Maybe when Strix releases with 3x the TOPS it might make sense.
 
Joined
Oct 27, 2009
Messages
1,133 (0.21/day)
Location
Republic of Texas
System Name [H]arbringer
Processor 4x 61XX ES @3.5Ghz (48cores)
Motherboard SM GL
Cooling 3x xspc rx360, rx240, 4x DT G34 snipers, D5 pump.
Memory 16x gskill DDR3 1600 cas6 2gb
Video Card(s) blah bigadv folder no gfx needed
Storage 32GB Sammy SSD
Display(s) headless
Case Xigmatek Elysium (whats left of it)
Audio Device(s) yawn
Power Supply Antec 1200w HCP
Software Ubuntu 10.10
Benchmark Scores http://valid.canardpc.com/show_oc.php?id=1780855 http://www.hwbot.org/submission/2158678 http://ww
LOL, hard to see how the current piss weak NPU's will do much for AI in industry. GPU is far more powerful currently and I doubt power is an issue. Maybe when Strix releases with 3x the TOPS it might make sense.
You don't need tons to do inference on 15fps 720p cameras. Think factory work. The ARM NPUs are more like.... 0.5-3.2tops.
 
Joined
May 3, 2018
Messages
2,270 (1.04/day)
You don't need tons to do inference on 15fps 720p cameras. Think factory work. The ARM NPUs are more like.... 0.5-3.2tops.
Like I said the cpu and gpu are far more powerful, we don't need these NPU's other than to save power in mobile situations where performance is not an issue. So I repeat what's the point?
 
Joined
Oct 27, 2009
Messages
1,133 (0.21/day)
Location
Republic of Texas
System Name [H]arbringer
Processor 4x 61XX ES @3.5Ghz (48cores)
Motherboard SM GL
Cooling 3x xspc rx360, rx240, 4x DT G34 snipers, D5 pump.
Memory 16x gskill DDR3 1600 cas6 2gb
Video Card(s) blah bigadv folder no gfx needed
Storage 32GB Sammy SSD
Display(s) headless
Case Xigmatek Elysium (whats left of it)
Audio Device(s) yawn
Power Supply Antec 1200w HCP
Software Ubuntu 10.10
Benchmark Scores http://valid.canardpc.com/show_oc.php?id=1780855 http://www.hwbot.org/submission/2158678 http://ww
Like I said the cpu and gpu are far more powerful, we don't need these NPU's other than to save power in mobile situations where performance is not an issue. So I repeat what's the point?
CPU no, GPU, yes, because you can use the NPU and the GPU at the same time.
 
Top