- 68
- 110 832
Xilinx Research & Open Source Projects
เข้าร่วมเมื่อ 11 ก.พ. 2021
This channel is to share educational videos, tutorials, demo designs and recorded technical talks relating to Xilinx research projects, XACC and Xilinx open source projects including PYNQ, RFSoC PYNQ, FINN and RapidWright.
AMD HACC Tech Talks: TaPaSCo on Versal: AIE, QDMA Streaming and MRMAC
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research.
www.amd-haccs.io
Torben Kalkhof, TU Darmstadt
The Task Parallel System Composer (TaPaSCo) is an open-source framework designed to simplify the construction of heterogeneous computing systems by efficiently offloading tasks to hardware accelerators. In this presentation, we explore how TaPaSCo leverages key features of the AMD Versal architecture, including AI Engines (AIE) for high-performance compute, Queue Direct Memory Access (QDMA) for efficient data streaming, and Multi-rate MAC (MRMAC) for high-speed networking.
www.amd-haccs.io
Torben Kalkhof, TU Darmstadt
The Task Parallel System Composer (TaPaSCo) is an open-source framework designed to simplify the construction of heterogeneous computing systems by efficiently offloading tasks to hardware accelerators. In this presentation, we explore how TaPaSCo leverages key features of the AMD Versal architecture, including AI Engines (AIE) for high-performance compute, Queue Direct Memory Access (QDMA) for efficient data streaming, and Multi-rate MAC (MRMAC) for high-speed networking.
มุมมอง: 66
วีดีโอ
AMD HACC Tech Talk: RoCE Balboa - RDMA Deep Packet Inspection at Line Rate with ML-models on FPGAs
มุมมอง 7514 ชั่วโมงที่ผ่านมา
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io RoCE Balboa - RDMA Deep Packet Inspection at Line Rate with ML-models on FPGAs Maximilian Heer, Benjamin Ramhorst, ETH Zurich FPGAs are becoming increasingly important in the cloud and data centers, especially as network-attached accelerato...
AMD HACC Tech Talk: Energy-efficient 2.5D System for LLM Inference
มุมมอง 14614 วันที่ผ่านมา
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Dr. Sumit Kumar Mandal, IISc Large Language Models (LLMs) are used to perform various tasks, especially in the domain of natural language processing (NLP). State-of-the-art LLMs consist of a large number of parameters that necessitate a hig...
AMD HACC Tech Talk: Pruning in the Datafree regime, A Distributional approach to Structured Pruning
มุมมอง 5814 วันที่ผ่านมา
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Presenter: Dr. Chiranjib Bhattacharyya, IISc Minimizing Inference time in Deep Networks is one of the biggest challenges in modern day AI systems. Model pruning is considered an important approach in addressing this issue. In this talk we h...
AMD HACC Tech Talk: Profiling Tools for AMD Hardware
มุมมอง 972 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Presenter: Giacomo Capodaglio, AMD HPC Solutions & Performance Analysis This talk will introduce three powerful open-source profiling tools: rocprof profiles all hardware counters on your GPU during execution. Omnitrace profiles and traces ...
AMD HACC Tech Talk: ROCm Ecosystem and HIP Programming
มุมมอง 2112 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Presenter: Tom Papatheodore AMD University Program ROCm is an open-source software stack that consists of a collection of drivers, development tools, and APIs that enable GPU programming from low-level kernel to end-user applications. The R...
AMD HACC Tech Talk: HiHiSpMV: with Hierarchial. Row Reduction on HBM FPGAs, 17 July 2024 (Part 2)
มุมมอง 962 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Presenter: Dr. Tobias Kenter, University of Paderborn Using the full potential of Alveo U280 for Sparse Matrix Vector Multiplications FPGAs with High Bandwidth Memory are a promising architecture for sparse linear algebra, but so far it has...
AMD HACC Tech Talk: Serpens and Callipepla, 17 July 2024 (Part 1)
มุมมอง 1322 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io Presenter: Dr. Linghao Song, UCLA Two FPGA-based HBM accelerators for enhanced computational efficiency will be introduced. Callipepla accelerates preconditioned conjugate gradient solvers, achieving significant speedup and energy efficienc...
AMD HACC Tech Talk: XDNA™ Architecture and Programming Model, 26 June 2024 (Part 1)
มุมมอง 5584 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io XDNA™ Architecture and Programming Model Presenter: Mario Ruiz (AMD) AMD XDNA™ is a multi-generation spatial dataflow architecture consisting of two-dimension array of VLIW and SIMD AI Engine cores that provide high compute density in addit...
AMD HACC Tech Talk: CHARM: Composing Heterogeneous Matrix Multiply Accelerators, 26 June 2024
มุมมอง 1884 หลายเดือนก่อน
The HACC Tech Talks are a series of virtual talks covering a broad range of topics related to Heterogeneous Accelerated Computing research. www.amd-haccs.io CHARM: Composing Heterogeneous Accelerators for Matrix Multiply on Versal ACAP devices Presenter: Jinming Zhuang AMD Versal offer a powerful platform for dense Matrix Multiplication for deep learning. Large MM operations can be efficiently ...
Introduction to Riallto and Ryzen AI
มุมมอง 1K5 หลายเดือนก่อน
Ryzen AI is a new Neural Processing Unit (NPU) from AMD, found in selected Ryzen processors. Riallto is a new open-source exploration framework for first-time Ryzen AI developers that lets you explore the NPU architecture and programming tools available for Ryzen AI. This video introduces Ryzen AI and Riallto, and motivates the need for NPUs. To learn more and install Riallto on your Ryzen AI l...
SDR with the Zynq RFSoC; Section 6: RF ADCs, DACs, DDCs & DUCs
มุมมอง 6K2 ปีที่แล้ว
Software Defined Radio Teaching & Research with the Xilinx Zynq Ultrascale RFSoC
MicroRec: Efficient Recommendation Inference on FPGAs
มุมมอง 3452 ปีที่แล้ว
The XACC Tech Talks are a series of virtual talks covering a broad range of topics related to Adaptive Compute. xilinx.github.io/xacc/ MicroRec: Efficient Recommendation Inference on FPGAs Wenqi Jiang, ETH Zurich This talk will introduce MicroRec, a high-performance FPGA deep neural network inference engine for recommendation systems. Two high-performance recommendation inference clusters will ...
ACCL: an open-source FPGA accelerated communication library for scale-out applications
มุมมอง 5512 ปีที่แล้ว
The XACC Tech Talks are a series of virtual talks covering a broad range of topics related to Xilinx Adaptive Compute. xilinx.github.io/xacc/ ACCL: an open-source FPGA accelerated communication library for scale-out applications Dr. Lucian Petrica, AMD AECG Research Labs This talk will present ACCL, an open-source FPGA-accelerated communication collectives library designed to enable scale-out o...
Portable, high-performance Python on CPUs, GPUs, FPGAs (XACC Winter school 2022)
มุมมอง 1.1K2 ปีที่แล้ว
Portable, high-performance Python on CPUs, GPUs, FPGAs (XACC Winter school 2022)
Innovative FPGA Approaches to AI (XACC Winter school 2022)
มุมมอง 6622 ปีที่แล้ว
Innovative FPGA Approaches to AI (XACC Winter school 2022)
System Building with FPGAs at ETH Zurich (XACC Winter school 2022)
มุมมอง 3002 ปีที่แล้ว
System Building with FPGAs at ETH Zurich (XACC Winter school 2022)
XACC Tech Talk7: Part 2 Lucent; 16 December 2021
มุมมอง 1852 ปีที่แล้ว
XACC Tech Talk7: Part 2 Lucent; 16 December 2021
XACC Tech Talk7: Part 1 InAccel; 16 December 2021
มุมมอง 2702 ปีที่แล้ว
XACC Tech Talk7: Part 1 InAccel; 16 December 2021
Tutorial (TVMCon 2021) - Neural Network Quantization with Brevitas
มุมมอง 2.9K2 ปีที่แล้ว
Tutorial (TVMCon 2021) - Neural Network Quantization with Brevitas
ITU AI for Good/ML in 5G - Awards 2021 (RadioML)
มุมมอง 6302 ปีที่แล้ว
ITU AI for Good/ML in 5G - Awards 2021 (RadioML)
Xilinx Adaptive Compute Clusters (XACC) Academia-Industry Research; Xilinx University Program
มุมมอง 5403 ปีที่แล้ว
Xilinx Adaptive Compute Clusters (XACC) Academia-Industry Research; Xilinx University Program
PYNQ: Extending the Python Ecosystem to Xilinx platforms (Patrick Lysaght)
มุมมอง 4.9K3 ปีที่แล้ว
PYNQ: Extending the Python Ecosystem to Xilinx platforms (Patrick Lysaght)
Research Talk (FPL'2021 Keynote) Innovative FPGA Approaches to AI
มุมมอง 5433 ปีที่แล้ว
Research Talk (FPL'2021 Keynote) Innovative FPGA Approaches to AI
PYNQ Composable Overlay overview and demo
มุมมอง 2.3K3 ปีที่แล้ว
PYNQ Composable Overlay overview and demo
Announcement of the 2021 Awards for the Xilinx Women in Technology University Grants
มุมมอง 1763 ปีที่แล้ว
Announcement of the 2021 Awards for the Xilinx Women in Technology University Grants
SDR with the Zynq RFSoC; Section 11: Wrap-up and call to participate
มุมมอง 9253 ปีที่แล้ว
SDR with the Zynq RFSoC; Section 11: Wrap-up and call to participate
Miller Dorothy Garcia Gary Lee Brian
DBF example by using RF-SoC ??
Raphaelle Drive
Batz Shoal
The sound quality of this video is completely messed up on multiple levels. Makes it very hard to follow.
You are using the term "demodulation" when you should be using "downconverting"
So pynq can't be used to create the fpga circuit logic?
Will this work on my AMD 6800 or my 7800X3d since it does include a GPU? The gpu should be good enough for small AI tasks or at least for developmental purposes and I'm not expecting great performance. Cheers.
crazy no comments wtf
hello. is the system generator project open source?
This is an awesome hardware platform, the problem is the costs are extremely prohibitive for many startups and smaller companies. $9K for a RFSoC eval board... They should make it more accessible and there will be a faster move towards it by RF designers.
where can i get these ipcores from?
Am facing problem in rfsoc-ofdm pacakage. Could you please provide mail address to send my issue in details?
The links seem not working.
hello I am getting no module named finn.core error using finn 0.0.1 how can I fix this.
英语不好,要听晕了
Based.
Why is the PLL frequency 409.6 MHz for the sampling rate of 1.024 GSPS?
Hello, I have applied to get the access, but has almost one month without any response. May you assist me
I cannot see the console elsewhere in Jupyter as you can in 30:40... how do I access the /usr/local/share/pynq-venv/lib/python3.8/site-packages/board.py if I only can see the folders inside the network folder "xilinx"? I used the command ! pip3 install adafruit-circuitpython-pca9685 To install the files, but as I'm not in console (And I can't find the option to access it from Jupyter) I'm not sure where to look for the Path... with ! ls I get the same folders and files as in Xilinx one
Any luck? Where to start?
Nice training
Curious how did you determine the NCO @750M. Mixer produce sum and difference of two carrier
Thank you
Hi, Vamshi Are you doing any projects related to Zynq SDR?
@@sridharbheemani4292 no not now
@@vamsikrishnaajjuri planning in the future?
@@sridharbheemani4292 I can't say now. I had plans but not now
@@sridharbheemani4292 can I know the reason why
Would this work with any RFSOC device?I was trying to find a way to work with the ZCU208 board.
You need to build PYNQ for your board by yourself.
thanks
Can other research organizations use this program?
Go to the XUP webpage and contact us for more information: www.xilinx.com/support/university.html
Great job!
th-cam.com/play/PLUn6cqainH8jZxS3ppSGPi3rNScz9cFZf.html
I can't find sellers with this board. Where can i buy it??
Please contact xup@xilinx.com
Where can we download the slides from this video?
Where can I get one? The links on TUL's site don't link to anyone that is selling them.
Please contact xup@xilinx.com
It would have been better if someone really demonstrates an example rather than simply putting the Audio instructions.
Are you having problems with setup? You can post any questions to the PYNQ support forum discuss.pynq.io/
Keep up the good work!!! You need more views. Did you ever check Promosm!?? You could use it to promote your videos!
interesting FPGA dev work!
can this presentation being gotten?
Impressive talk. Thanks for the introduction