#oneapi

[govindhtech]

SynxFlow Project: A Smooth Migration From CUDA To SYCL

The SynxFlow Project

SynxFlow, an open-source GPU-based hydrodynamic flood modeling software, in CUDA, C++, and Python Data pre-processing and visualization are done in Python while simulations are executed on CUDA. SynxFlow can simulate floods quicker than real-time with hundreds of millions of computational cells and metre-level precision on many GPUs. An open-source software with a simple Python interface, it may be linked into data science workflows for disaster risk assessments. The model has been widely utilized in research and industry, such as to assist flood early warning systems and generate flood maps for (re)insurance firms.

SynxFlow can simulate flooding, landslide runout, and debris flow. Simulations are crucial to emergency service planning and management. A comprehensive prediction of natural disasters can reduce their social and economic costs. In addition to risk assessment and disaster preparedness, SynxFlow flood simulation can help with urban planning, environmental protection, climate change adaptation, insurance and financial planning, infrastructure design and engineering, public awareness, and education.

- Advertisement -

Issue Statement

Several variables make probabilistic flood forecasting computationally difficult:

Large dataset storage, retrieval, and management

Complex real-time data processing requires high-performance computation.

Model calibration and validation needed as real-world conditions change.

Effective integration and data transfer between hydrological, hydraulic, and meteorological models, and more.

For speedier results, a flood forecasting system must process data in parallel and offload compute-intensive operations to hardware accelerators. Thus, the SynxFlow team must use larger supercomputers to increase flood simulation scale and cut simulation time. DAWN, the UK’s newest supercomputer, employs Intel GPUs, which SynxFlow didn’t support.

These issues offered researchers a new goal to make the SynxFlow model performance-portable and scalable on supercomputers with multi-vendor GPUs. They must transition SynxFlow code from CUDA to a cross-vendor programming language in weeks, not years.

Solution Powered by oneAPI

After considering several possibilities, the SynxFlow project team chose the Intel oneAPI Base Toolkit implementation of the Unified Acceleration Foundation-backed oneAPI protocol. All are built on multiarchitecture, multi-vendor SYCL framework. It supports Intel, NVIDIA, and AMD GPUs and includes the Intel DPC++ Compatibility Tool for automated CUDA-to-SYCL code translation.

- Advertisement -

SynxFlow code migration went smoothly. This produced code that automatically translated most CUDA kernels and API calls into SYCL. After auto-translation, some mistakes were found during compilation, but the migration tool’s error-diagnostic indications and warnings made them easy to rectify. It took longer to switch from NVIDIA Collective Communications Library (NCCL)-based inter-GPU communication to GPU-direct enabled Intel MPI library calls because this could not be automated.

To summarize, there has been a promising attempt to transfer a complicated flood simulation code that was built on CUDA to SYCL, achieving both scalability and performance-portability. The conversion has been easy to handle and seamless thanks to the Intel oneAPI Base Toolkit.

Intel hosted a oneAPI Hackfest at the DiRAC HPC Research Facility

DiRAC

The High Performance Super Computer facility in the United Kingdom serving the theoretical communities of Particle Physics, Astrophysics, Cosmology, Solar System and Planetary Science, and Nuclear Physics.

DiRAC’s three HPC services Extreme Scaling, Memory-Intensive, and Data-Intensive are each designed to support the distinct kinds of computational workflows required to carry out their science program. DiRAC places a strong emphasis on innovation, and all of its services are co-designed with vendor partners, technical and software engineering teams, and research community.

Training Series on oneAPI at DiRAC Hackfest

On May 21–23, 2024, the DiRAC community hosted three half-day remote training sessions on the Intel oneAPI Base Toolkit. The training series was designed for developers and/or researchers with varying degrees of experience, ranging from novices to experts.

The cross-platform compatible SYCL programming framework served as the foundation for a variety of concepts that were taught to the attendees. The students were also introduced to a number of Base Kit component tools and libraries that facilitate SYCL. For instance, the Intel DPC++ Compatibility Tool facilitates automated code migration from CUDA to C++ with SYCL; the Intel oneAPI Math Kernel Library (oneMKL) optimizes math operations; the Intel oneAPI Deep Neural Networks (oneDNN) accelerates hackfest and the Intel oneAPI DPC++ Library (oneDPL) expedites SYCL kernels on a variety of hardware. Additionally, the training sessions covered code profiling and the use of Intel Advisor and Intel VTune Profiler, two tools included in the Base Kit for analyzing performance bottlenecks.

DiRAC Hackfest’s oneAPI Hackath on

In order to complete a range of tasks, including parallelizing Fortran code on Intel GPUs, accelerating math operations like the Fast Fourier Transform (FFT) using oneMKL’s SYCL API, and resolving performance bottlenecks with the aid of Intel Advisor and Intel VTune Profiler, the participants improvised their cutting-edge projects using oneAPI tools and libraries.

The participants reported that it was easy to adjust to using oneAPI components and that the code migration process went smoothly. The teams saw a noticeable increase in workload performance with libraries like Intel MPI. Approximately 70% of the teams who took part indicated that they would be open to using oneAPI technologies to further optimize the code for their research projects. Thirty percent of the teams benchmarked their outcomes using SYCL and oneAPI, and they achieved a 100% success rate in code conversion to SYCL.

Start Programming Multiarchitecture Using SYCL and oneAPI

Investigate the SYCL framework and oneAPI toolkits now for multiarchitecture development that is accelerated! Use oneAPI to enable cross-platform parallelism in your apps and move your workloads to SYCL for high-performance heterogeneous computing.

Intel invite you to review the real-world code migration application samples found in the CUDA to SYCL catalog. Investigate the AI, HPC, and rendering solutions available in Intel’s software portfolio driven by oneAPI.

Read more on govindhtech.com

[incredefy]

[likejazz]

지난번 라마3 모델의 순수 NumPy 구현에 이어 이번에는 라마3 모델을 순수 C/CUDA로 구현해봤습니다.

GitHub: https://github.com/likejazz/llama3.cuda

Reddit 소개글: https://www.reddit.com/r/LocalLLaMA/comments/1d6q7qh/llama3cuda_pure_ccuda_implementation_for_llama_3/

NumPy 구현과 달리 CUDA 구현은 고성능을 목표로 했기 ���문에 M2 맥북 에어에서 33 tokens/s로 처리되던 것에 비해 CUDA 버전은 4080 SUPER에서 2,823 tokens/s로 처리되어 약 85배 더 빠른 속도를 보입니다. 왜 GPU를 사용해야 하는지를 몸소 체감할 수 있도록 했습니다.

무엇보다 이 모든 구현에 오픈소스 커뮤니티의 많은 도움을 받았습니다. 앞으로도 몇 차례 추가 개선을 통해 다양한 실험을 이어나갈 생각입니다. 피드백과 기여 또한 환영합니다.

상세한 기술적인 내용은 다음과 같습니다.

No dependency 코드는 간결함과 가독성을 유지하도록 구성했으며, 의존성이 없는 구조로 개발하여 어디서든 간단하게 컴파일이 가능하게 했습니다. 아울러 Makefile과 CMake를 모두 지원합니다.

No C++ C++를 사용하지 않은 순수 C 구현이며, 모든 값은 포인터와 더블 포인터로 CPU에서 GPU의 값을 가리키는 형태로 구현되어 있습니다.

One single file UTF-8 바이트 시퀀스 처리 같은 상용구 코드를 포함하고 있지만 전체 코드를 단일 파일에서 900줄 미만으로 유지했습니다.

Same result NumPy 구현과 동일한 결과를 얻기 위해 logit 값을 일일이 디버깅하여 부동 소수점 연산 오차를 0.5% 미만으로 줄였습니다.

이외에도 추후 개선 계획은 다음과 같습니다.

NumPy 버전과 동일한 결과를 보장하기 위해 토크나이저를 비롯한 일부 패치를 진행하였는데 재밌게도 토크나이저 코드에 안드레이 카파시가 “I don't have the energy to read more of the sentencepiece code to figure out what it's doing”라고 주석을 달아둔 걸 볼 수 있었습니다. 이 부분을 깔끔하게 마무리 하기 위해 많은 시간을 투자해 봤으나 아쉽게도 좋은 결과를 얻진 못했고 지저분하게 몽키 패치로 처리할 수밖에 없었습니다. 추후에 추가 개선을 진행하면서 다시 한번 도전해 볼 예정입니다.

이후에는 AMD의 ROCm 구현, Intel의 oneAPI 구현을 통해 CUDA 이외에 다른 플랫폼에서도 각각 검증을 진행해 볼 예정입니다. 또한 현재도 단일 커널로 Multi-Head Attention을 처리하여 Flash Attention과 비슷한 효과를 내고 있으나 GEMM 연산이 아닌 GEMV 연산을 수행하여 다소 비효율적으로 동작하는 문제가 있습니다. 추후에는 이 부분을 개선하여 Flash Attention을 정확하게 구현해 볼 예정입니다.

모델 구조와 UTF-8 토크나이저 구현은 안드레이 카파시가 지난번 llama2.c에 구현한 것을 대부분 사용했으며 CUDA 구현은 포틀랜드의 CUDA 전문 프로그래머인 rogerallen이 구현한 커널을 활용했습니다. 이외에 초창기에 인도의 ankan-ban이 구현한 CUDA 커널도 많이 참고하였습니다.

전 세계 오픈소스 커뮤니티의 도움으로 저 또한 간결하게 CUDA 구현을 마무리할 수 있었습니다. 많은 도움을 받았던 만큼 저 또한 코드를 공개하여 오픈소스 혁신이 전 세계에 이어질 수 있도록 보탬을 더해봅니다. 많은 분들에게 도움이 되길 바랍니다.

[jcmarchi]

Intel’s Aurora achieves exascale to become the fastest AI system

New Post has been published on https://thedigitalinsider.com/intels-aurora-achieves-exascale-to-become-the-fastest-ai-system/

Intel’s Aurora achieves exascale to become the fastest AI system

.pp-multiple-authors-boxes-wrapper display:none; img width:100%;

Intel, in collaboration with Argonne National Laboratory and Hewlett Packard Enterprise (HPE), has revealed that its Aurora supercomputer has exceeded the exascale computing threshold reaching speeds of 1.012 exaflops to become the fastest AI-focused system.

“The Aurora supercomputer surpassing exascale will allow it to pave the road to tomorrow’s discoveries,” said Ogi Brkic, Intel’s VP and GM of Data Centre AI Solutions. “From understanding climate patterns to unravelling the mysteries of the universe, supercomputers serve as a compass guiding us toward solving truly difficult scientific challenges that may improve humanity.”

Aurora has not only excelled in speed but also in innovation and utility. Designed from the outset as an AI-centric supercomputer, Aurora enables researchers to leverage generative AI models, significantly accelerating scientific discovery.

Groundbreaking work has already been achieved using Aurora, including mapping the 80 billion neurons of the human brain, enhancing high-energy particle physics with deep learning, and accelerating drug design and discovery through machine learning.

At the core of Aurora lies the Intel Data Center GPU Max Series, built on the innovative Intel Xe GPU architecture, optimised for both AI and HPC tasks. This technological foundation enables parallel processing capabilities, crucial for handling complex neural network AI computations.

Details shared about the supercomputer highlight its grand scale, consisting of 166 racks, 10,624 compute blades, 21,248 Intel Xeon CPU Max Series processors, and 63,744 Intel Data Center GPU Max Series units, making it the largest GPU cluster worldwide.

Alongside the hardware, Intel’s suite of software tools – including the Intel® oneAPI DPC++/C++ Compiler and an array of performance libraries – boost developer flexibility and system scalability.

Intel is also expanding its Tiber Developer Cloud, incorporating new state-of-the-art hardware and enhanced service capabilities to support the evaluation, innovation, and optimisation of AI models and workloads on a large scale.

Looking forward, the deployment of new supercomputers integrated with Intel technologies is expected to transform various scientific fields. Systems like CMCC’s Cassandra will advance climate change modelling, while ENEA’s CRESCO 8 supports breakthroughs in fusion energy, among others, underscoring Intel’s dedication to advancing HPC and AI into new realms of discovery and innovation.

(Image Credit: Intel)

See also: Chuck Ros, SoftServe: Delivering transformative AI solutions responsibly

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Tags: ai, artificial intelligence, aurora, exaflop, exascale, intel

[tomreview]

Write Compelling Product Descriptions & Boost Sales

Intelli AI Kit Unleashed: Your Gateway to Online Success for Only $17.97!Embark on a journey of innovation and growth with the Intelli AI Kit, now available at the unbelievable price of $17.97! This isn't just a purchase; it's a gateway to a world of cutting-edge tools that will revolutionize your online business. Seize this fleeting opportunity to elevate your strategies and soar ahead in the digital landscape.The buzz around the Intelli AI Kit is reaching a crescendo, highlighting its exceptional value. Act swiftly before the winds of change alter its price. Take charge now and harness the transformative power of Intelli AI Kit at this jaw-dropping cost of $17.97.Don't miss this chance to redefine your marketing approach and unlock endless possibilities. Time is of the essence - act decisively, secure your competitive edge, and embrace the future of online success before this incredible offer transforms

Unlock Limitless Potential with Intelli AI Kit Upgrades!

♥ OTO 1 Intelli AI Kit Pro ($47)

Access 30 Extra AI Tools for Unprecedented Growth

Target Global Audiences with 30 Additional Languages

Enhance Results with 3 Variations of Each Output

Customize Output Tones for Tailored Content

Elevate Quality Across Content, Videos, and Ad Copy

Save Big, Increase Profits, and Eliminate Third-Party Reliance

Backed by a Solid 30-Day Money-Back Guarantee

♥ OTO 2 Intelli AI Kit Unlimited ($67)

Break Free from Limits and Unleash Unlimited Growth Potential

Send Endless Emails to Unlimited Subscribers

Create Boundless Content Across Diverse Niches

Craft Unlimited YouTube Descriptions and Facebook Ads

Drive Unlimited Targeted Traffic and Generate New Leads

Ironclad 30-Day Money-Back Guarantee

♥ OTO 3 Intelli AI Kit MAX ($97)

Explore Limitless Creativity with 125 Premium Chat Bots

Create Custom Chat GPT 4 Prompts for Any Niche

Enjoy Unlimited Prompts and Output Customization

Deploy 125 Virtual Assistants for Seamless Website Integration

Enhance Efficiency, Reduce Errors, and Save Time

Backed by a Reliable 30-Day Money-Back Guarantee

♥ OTO 4 Intelli AI Kit Click Design ($97)

Access ClickDesigns Software with Commercial Edition

Utilize 3,000+ Customizable Templates Across 28 Niche Categories

Design Logos, Book Covers, Reports, and More with Ease

Gain Commercial Rights for Profitable Graphic Sales

Comprehensive Support and Training Included

♥ OTO 5 Intelli AI Kit AI Smart News ($67)

Launch Self-Updating Viral News Websites in Any Niche

Choose from 300+ Premium Templates for Engaging Sites

Monetize News Sites on Flippa, eBay, and Facebook

Real-Time Trending News Updates and Multilingual Support

Drag & Drop Editor, Lifetime Hosting, and SEO Optimization

30-Day Money-Back Guarantee for Peace of Mind

♥ OTO 6 Intelli AI Kit 1-Hour Profits ($67)

Dive into a DFY $50-100K Monthly Profit System

Access Proven Sales Pages and Drive Massive Traffic

Start Profiting Within 60 Minutes with Guaranteed Returns

Enjoy a 30-Day Money-Back Guarantee for Risk-Free Investment

♥ OTO 7 Intelli AI Kit 1-Click Traffic Generator ($97)

Drive Targeted Traffic to Your Intelli AI Kit and eShops

Generate Additional $50-100K Monthly Profits in Just One Hour

♥ OTO 8 Intelli AI Kit Reseller ($97)

Become a Reseller and Keep 100% Profits

Launch Your Software Business with Minimal Hassles

Offer a High-Demand Product for Lucrative Returns

Budget-Friendly Option with Quick ROI and Zero Maintenance Costs

Who Should Use Intelli AI Kit:

Data Scientists:

Accelerate end-to-end data science and analytics pipelines on Intel® architecture.

Train on Intel® CPUs and GPUs, integrate fast inference, and optimize models efficiently.

AI Developers:

Access familiar Python* tools and frameworks for AI development.

Achieve drop-in acceleration for data preprocessing and machine learning workflows.

Researchers:

Utilize Intel's AI Tools to maximize performance from preprocessing through machine learning.

Develop and optimize oneAPI multiarchitecture applications without hardware installations or complex configurations

Advantages of Artificial Intelligence: Reduction in Human Error: AI can significantly reduce errors and increase accuracy by relying on gathered information and algorithms

Zero Risks: AI can undertake risky tasks like defusing bombs or exploring hazardous environments, providing accurate work with greater responsibility 24x7 Availability: AI can work endlessly without breaks, handling multiple tasks simultaneously with accurate results, such as online customer support chatbots providing instant assistance Disadvantages of Artificial Intelligence: Job Displacement: AI can lead to job losses as automation replaces human roles Ethical Concerns: Issues like bias, privacy, and security risks from hacking are raised with the use of AI Lack of Human-like Creativity: AI lacks human creativity and empathy, limiting its capabilities in certain areas

[hackernewsrobot]

FreeBSD has a(nother) new C compiler: Intel oneAPI DPC++/C++

https://briancallahan.net/blog/20240306.html

[lanshengic]

Intel continues to innovate edge AI technology and joins hands with ecological partners to promote urban digital transformation

【Lansheng Technology News】On October 24, 2023, the 2023 Intel Digital Intelligence Park and Community Ecology Conference with the theme of "Initiating Digital Intelligence to Create a Smart City" was held in Shenzhen.

At this conference, Intel comprehensively demonstrated its vision and strategic layout in the field of smart cities, and talked about the future of smart cities with many technical experts and industry elites from smart cities, smart parks, smart communities and other professional fields. Trends and development directions, and jointly explore the latest technologies and products in the field of smart cities.

Intel said that in order to solve the security issues faced by smart cities during the development process, as well as the challenges caused by road congestion and insufficient infrastructure, the company has built a strong edge artificial intelligence software and hardware product portfolio and vertical industry solutions, and cooperated with a wide range of Cooperate with ecological partners to accelerate the digital and intelligent transformation of cities.

In terms of product layout, Intel has built a diversified hardware product portfolio for the edge, including Intel® Xeon® processors, Intel® Core® processors, and ARC series and Flex series independent graphics cards; on the other hand, Intel can provide tools, software , reference solutions and tools, including OpenVINO, FlexRAN, and oneAPI.

At the same time, Intel also provides critical value to a variety of business types and end customers by creating innovative solutions. For example, the VPPSDK solution provided by Intel can enable customers to quickly develop application software for NVR/VPP products on Intel CPU platforms, allowing developers to directly use the interfaces provided by VPPSDK to complete corresponding functional development without further knowledge of Intel. Chip-related underlying video processing software. The Intel® Edge AI Computing Box integrates Intel's advanced software and hardware technology and commercial AI algorithms that have been deployed in thousands of industries, helping industry partners accelerate the launch and deployment of solutions.

Intel cooperates with a wide range of ecological partners to promote the large-scale implementation of innovative technologies in cities and create advanced solutions for future smart cities for many scenarios. For example, based on Intel processors, Gosuncn Technology Group has built a holographic smart video application solution to help realize holographic processing of smart videos and multi-dimensional data; Intel and Entropy Technology have jointly created the Entropy-based intelligent edge server BioCVBox, which is targeted at small and medium-sized enterprises. Micro-projects integrate storage, computing, management and other applications. They have the characteristics of low construction cost, simple deployment and operation and maintenance, powerful performance, safety, stability and high reliability.

In addition, Intel is working with the National Intelligent Building and Residential Area Digital Standardization Technical Committee to develop reference standards for the digital campus industry. This specification can ensure that the construction of the park follows unified standards, enables interconnection between various systems, and avoids the formation of information islands. It can also help the park create unified workflow, collaboration, scheduling and sharing mechanisms, thereby promoting the park to improve operational efficiency and service quality. , improve the overall operational level.

Lansheng Technology Limited, which is a spot stock distributor of many well-known brands, we have price advantage of the first-hand spot channel, and have technical supports. 

Our main brands: STMicroelectronics, Toshiba, Microchip, Vishay, Marvell, ON Semiconductor, AOS, DIODES, Murata, Samsung, Hyundai/Hynix, Xilinx, Micron, Infinone, Texas Instruments, ADI, Maxim Integrated, NXP, etc

To learn more about our products, services, and capabilities, please visit our website at http://www.lanshengic.com

[tamarovjo4]

The Linux Foundation launches the Unified Acceleration Foundation to create an open standard for accelerator programming, an evolution of the oneAPI initiative (Frederic Lardinois/TechCrunch)

http://dlvr.it/SwMjZg

[jacquelineknowle]

Professional Marketplace API Integration - DigitalAPICraft

If you seek a reliable and robust API service provider, nothing could be better than DigitalAPICraft. We provide a professional marketplace API integration where developers can easily find the APIs, and providers can easily publish them for consumption to earn greater revenue. For the best OneAPI marketplace experience, visit DigitalAPICraft. 

Visit Here - https://digitalapicraft.com/one-api-market-place/

[cheerchain]

祺荃企業 2023 年 5 月 科研軟體產品最新購買指南(1) - 來看看有什麼新應用軟體! #Software #軟體 #Bandicam #質性 #科研 #SmartPLS #MAXQDA #NVivo #oneAPI #SigmaPlot #Grapher #Camtasia #MindManager 線上閱覽 >>> https://cheerchain.benchmarkurl.com/c/v?e=1633534&c=80A69&t=1&l=93DC2539&email=u%2FFmAYrLEJPlJ4AvQgqL%2B3wDN63%2FgZq6&relid=AFA2136

[yoji-ono]

via Publickey

[govindhtech]

Utilizing llama.cpp, LLMs can be executed on Intel GPUs

The open-source project known as llama.cpp is a lightweight LLM framework that is gaining greater and greater popularity. Given its performance and customisability, developers, scholars, and fans have formed a strong community around the project. Since its launch, GitHub has over 600 contributors, 52,000 stars, 1,500 releases, and 7,400 forks. More hardware, including Intel GPUs seen in server and consumer products, is now supported by llama.cpp as a result of recent code merges. Hardware support for GPUs from other vendors and CPUs (x86 and ARM) is now combined with Intel’s GPUs.

Georgi Gerganov designed the first implementation. The project is mostly instructional in nature and acts as the primary testing ground for new features being developed for the machine learning tensor library known as ggml library. Intel is making  AI more accessible to a wider range of customers by enabling inference on a greater number of devices with its latest releases. Because Llama.cpp is built in C and has a number of other appealing qualities, it is quick.

16-bit float compatibility

Support for integer quantisation (four-, five-, eight-, etc.)

Absence of reliance on outside parties

There are no runtime memory allocations.

Intel GPU SYCL Backend

GGM offers a number of backends to accommodate and adjust for various hardware. Since oneAPI supports GPUs from multiple vendors, Intel decided to construct the SYCL backend using their direct programming language, SYCL, and high-performance BLAS library, oneMKL. A programming model called SYCL is designed to increase hardware accelerator productivity. It is an embedded, single-source language with a domain focus that is built entirely on C++17.

All Intel GPUs can be used with the SYCL backend. Intel has confirmed with:

Flex Series and Data Centre GPU Max from Intel

Discrete GPU Intel Arc

Intel Arc GPU integrated with the Intel Core Ultra CPU

In Intel Core CPUs from Generations 11 through 13: iGPU

Millions of consumer devices can now conduct inference on Llama since llama.cpp now supports Intel GPUs. The SYCL backend performs noticeably better on Intel GPUs than the OpenCL (CLBlast) backend. Additionally, it supports an increasing number of devices, including CPUs and future processors with  AI accelerators. For information on using the SYCL backend, please refer to the llama.cpp tutorial.

Utilise the SYCL Backend to Run LLM on an Intel GPU

For SYCL, llama.cpp contains a comprehensive manual. Any Intel GPU that supports SYCL and oneAPI can run it. GPUs from the Flex Series and Intel Data Centre GPU Max can be used by server and cloud users. On their Intel Arc GPU or iGPU on Intel Core CPUs, client users can test it out. The 11th generation Core and later iGPUs have been tested by Intel. While it functions, the older iGPU performs poorly.

The memory is the only restriction. Shared memory on the host is used by the iGPU. Its own memory is used by the dGPU. For llama2-7b-Q4 models, Intel advise utilising an iGPU with 80+ EUs (11th Gen Core and above) and shared memory that is greater than 4.5 GB (total host memory is 16 GB and higher, and half memory could be assigned to iGPU).

Put in place the Intel GPU driver

There is support for Windows (WLS2) and Linux. Intel suggests Ubuntu 22.04 for Linux, and this version was utilised for testing and development.

Linux:sudo usermod -aG render username sudo usermod -aG video username sudo apt install clinfo sudo clinfo -l

Output (example):Platform #0: Intel(R) OpenCL Graphics -- Device #0: Intel(R) Arc(TM) A770 Graphics

orPlatform #0: Intel(R) OpenCL HD Graphics -- Device #0: Intel(R) Iris(R) Xe Graphics \[0x9a49\]

Set the oneAPI Runtime to ON

Install the Intel oneAPI Base Toolkit first in order to obtain oneMKL and the SYCL compiler. Turn on the oneAPI runtime next:

First, install the Intel oneAPI Base Toolkit to get the SYCL compiler and oneMKL. Next, enable the oneAPI runtime:

Linux: source /opt/intel/oneapi/setvars.sh

Windows: “C:\Program Files (x86)\Intel\oneAPI\setvars.bat\” intel64

Run sycl-ls to confirm that there are one or more Level Zero devices. Please confirm that at least one GPU is present, like [ext_oneapi_level_zero:gpu:0].

Build by one-click:

Linux: ./examples/sycl/build.sh

Windows: examples\sycl\win-build-sycl.bat

Note, the scripts above include the command to enable the oneAPI runtime.

Run an Example by One-Click

Download llama-2–7b.Q4_0.gguf and save to the models folder:

Linux: ./examples/sycl/run-llama2.sh

Windows: examples\sycl\win-run-llama2.bat

Note that the scripts above include the command to enable the oneAPI runtime. If the ID of your Level Zero GPU is not 0, please change the device ID in the script. To list the device ID:

Linux: ./build/bin/ls-sycl-device or ./build/bin/main

Windows: build\bin\ls-sycl-device.exe or build\bin\main.exe

Synopsis

All Intel GPUs are available to LLM developers and users via the SYCL backend included in llama.cpp. Kindly verify whether the Intel laptop, your gaming PC, or your cloud virtual machine have an iGPU, an Intel Arc GPU, or an Intel Data Centre GPU Max and Flex Series GPU. If so, llama.cpp’s wonderful LLM features on Intel GPUs are yours to enjoy. To add new features and optimise SYCL for Intel GPUs, Intel want developers to experiment and contribute to the backend. The oneAPI programming approach is a useful project to learn for cross-platform development.

Read more on Govindhtech.com

[craigbrownphd-blog-blog]

Advancing HPC and AI through oneAPI Heterogeneous Programming in Academia and Research https://www.datasciencecentral.com/advancing-hpc-and-ai-through-oneapi-heterogeneous-programming-in-academia-and-research/?utm_source=dlvr.it&utm_medium=tumblr

[usacounselingcredit]

Houston Texas Appliance Parts: AI Software Developers Extol the Power of Standards

Houston Texas Appliance Parts

AI Software Developers Extol the Power of Standards

by Houston Texas Appliance Parts on Wednesday 08 February 2023 10:21 AM UTC-05

AI software and its industry have grown so painfully big and complex that standards are needed to simplify things and ease the discomfort.

That was the key message from participants in a panel discussion on identifying and solving developers' pain points, which was held during EE Times's recent AI Everywhere Forum.

Andrew Richards (Source: Autosens Conference 2016/Bernal Revert)

"All AI systems, from tinyML to supercomputers, need software, but the state of AI software across the industry is at best variable," said panel moderator Nitin Dahad, editor-in-chief of embedded.com and an EE Times correspondent. "What are the biggest pain points developers face today with AI software stacks? What problems are common to all applications, snd where should we start trying to fix them?"

Standardization may be the start of a fix, according to panelists.

"Industry standards let you plug things together, right?" said Codeplay CEO Andrew Richards, who also became a VP at Intel after it bought the specialist software company in 2022. "If you think about something like USB or Wi-Fi, they just make things work together, even though they're designed by different companies, and different technologies, different skill sets."

Richards and his fellow panelists had other issues on their mind, too:

Fragmentation of frameworks;

Performance;

The need to debug data and the lack of data debuggers;

Managing the disparate professionals and skill sets needed for AI software, along with the complexity of the applications themselves, and

Code portability.

"The answer to all of these questions, from my point of view, is industry standards," Richards said. "And that's what we do with SYCL, and that's what we're doing across the oneAPI project, what we've been doing with other standards: enabling different sector people to work together."

He encouraged developers to become active in the groups writing the standards, including The Khronos Group and the oneAPI Community Forum.

"If you look at the standards and you think, 'Oh, that's not a good fit for us,' come and join," he said. "You can change them. You can actually change the standards and take them down a direction that you want, and it's going to work much better for you."

David Kanter (Source: MLCommons)

Necessary beyond writing or re-writing standards, though, is trying them, Richards said.

"What we find with SYCL is people go, 'It's never going to work, it's never going to work,' and then you try," he said. "And it does actually work, and you actually get really good performance. Now we can run massive, large-scale software."

Don't forget standards for data, said David Kanter, executive director of MLCommons, a consortium formed to grow ML from a research field into a mature industry via benchmarks, public datasets and best practices.

 "One thing I would say is, standards for data, not a thing, right?" he said.

Kanter cited the example of building an ML model for speech. There are no standardized inputs developers always use for speech, he said, so everyone has a different pipeline munging things in different ways.

Anders Hardebring (Source: Imagimob)

"It also depends on your target," he said. "So, to talk about tinyML, in some cases you may have an on-device component that can do some portion of the pre-processing. Sometimes you don't. So, there's a lot of things there where standards could totally help. That's another example of an area where software has a rich history, but we don't see it on the data side, and sort of need to develop that intuition and those muscles."

The two parts of AI applications—software and data—are mismatched in their development, panelists said, leading to post-spinach Popeye "muscles" of code, while the ML training information remains a 98-pound weakling.

"We believe that the biggest pain point is to be able to collect enough amounts of data—well annotated, high quality—and to have the software to do that," said Anders Hardebring, the CEO of Imagimob, a development platform for ML on edge devices.

Popeye muscles or no, software development always lags behind new hardware, Hardebring noted.

Alex Grbic, VP of software engineering at chip company Untether AI, agreed with Hardebring. "In order to take advantage of the novel architectures, the whole reason these new architectures are coming out are to meet a certain performance requirement that more traditional architectures can't," he said. "But they are complex, right?"

Alex Grbic (Source: Untether AI)

Untether's customers who use frameworks like TensorFlow don't want to do the underlying parallel programming needed for spatial architectures, he added.

"And, in that case, they rely heavily on the software to take advantage of that," Grbic said. "And we're the ones that provide it."

While standards are being worked out, Untether's software simplifies the use of its hardware and facilitates the better performance it makes possible.

 The post AI Software Developers Extol the Power of Standards appeared first on EE Times.

Pennsylvania Philadelphia PA Philadelphia February 08, 2023 at 09:00AM

Hammond Louisiana Ukiah California Dike Iowa Maryville Missouri Secretary Maryland Winchester Illinois Kinsey Alabama Edmundson Missouri Stevens Village Alaska Haymarket Virginia Newington Virginia Edwards Missouri https://unitedstatesvirtualmail.blogspot.com/2023/02/houston-texas-appliance-parts-ai.html February 08, 2023 at 10:34AM Gruver Texas Glens Fork Kentucky Fork South Carolina Astoria Oregon Lac La Belle Wisconsin Pomfret Center Connecticut Nason Illinois Roan Mountain Tennessee https://coloradovirtualmail.blogspot.com/2023/02/houston-texas-appliance-parts-ai.html February 08, 2023 at 04:41PM from https://youtu.be/GuUaaPaTlyY February 08, 2023 at 04:47PM

[jcmarchi]

Dell, Intel and University of Cambridge deploy the UK’s fastest AI supercomputer

New Post has been published on https://thedigitalinsider.com/dell-intel-and-university-of-cambridge-deploy-the-uks-fastest-ai-supercomputer/

Dell, Intel and University of Cambridge deploy the UK’s fastest AI supercomputer

.pp-multiple-authors-boxes-wrapper display:none; img width:100%;

Dell, Intel, and the University of Cambridge have jointly announced the deployment of the Dawn Phase 1 supercomputer.

This cutting-edge AI supercomputer stands as the fastest of its kind in the UK today. It marks a groundbreaking fusion of AI and high-performance computing (HPC) technologies, showcasing the potential to tackle some of the world’s most pressing challenges.

Dawn Phase 1 is the cornerstone of the recently launched UK AI Research Resource (AIRR), demonstrating the nation’s commitment to exploring innovative systems and architectures.

This supercomputer brings the UK closer to achieving the exascale; a computing threshold of a quintillion (10^18) floating point operations per second. To put this into perspective, the processing power of an exascale system equals what every person on Earth would calculate in over four years if they were working non-stop, 24 hours a day.

Operational at the Cambridge Open Zettascale Lab, Dawn utilises Dell PowerEdge XE9640 servers, providing an unparalleled platform for the Intel Data Center GPU Max Series accelerator. This collaboration ensures a diverse ecosystem through oneAPI, fostering an environment of choice.

The system’s capabilities extend across various domains, including healthcare, engineering, green fusion energy, climate modelling, cosmology, and high-energy physics.

Adam Roe, EMEA HPC technical director at Intel, said:

“Dawn considerably strengthens the scientific and AI compute capability available in the UK and it’s on the ground and operational today at the Cambridge Open Zettascale Lab.

Dell PowerEdge XE9640 servers offer a no-compromises platform to host the Intel Data Center GPU Max Series accelerator, which opens up the ecosystem to choice through oneAPI.

I’m very excited to see the sorts of early science this machine can deliver and continue to strengthen the Open Zettascale Lab partnership between Dell Technologies, Intel, and the University of Cambridge, and further broaden that to the UK scientific and AI community.”

Glimpse into the future

Dawn Phase 1 is not just a standalone achievement; it’s part of a broader strategy.

The collaborative endeavour aims to deliver a Phase 2 supercomputer in 2024, promising tenfold performance levels. This progression would propel the UK’s AI capability, strengthening the successful industry partnership.

The supercomputer’s technical foundation lies in Dell PowerEdge XE9640 servers, renowned for their versatile configurations and efficient liquid cooling technology. This innovation ensures optimal handling of AI and HPC workloads, offering a more effective solution than traditional air-cooled systems.

Tariq Hussain, Head of UK Public Sector at Dell, commented:

“Collaborations like the one between the University of Cambridge, Dell Technologies and Intel, alongside strong inward investment, are vital if we want the compute to unlock the high-growth AI potential of the UK. It is paramount that the government invests in the right technologies and infrastructure to ensure the UK leads in AI and exascale-class simulation capability.

It’s also important to embrace the full spectrum of the technology ecosystem, including GPU diversity, to ensure customers can tackle the growing demands of generative AI, industrial simulation modelling and ground-breaking scientific research.”

As the world awaits the full technical details and performance numbers of Dawn Phase 1 – slated for release in mid-November during the Supercomputing 23 (SC23) conference in Denver, Colorado – the UK stands at the precipice of a transformative era in scientific and AI research.

This collaboration between industry giants and academia not only accelerates research discovery but also propels the UK’s knowledge economy to new heights.

(Image Credit: Joe Bishop for Cambridge Open Zettascale Lab)

See also: UK paper highlights AI risks ahead of global Safety Summit

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Tags: ai, ai research resource, ai supercomputer, airr, artificial intelligence, dawn phase 1, dawn supercomputer, dell, intel, oneapi, open zettascale lab, poweredge, sc23, supercomputer, supercomputing 23, uk, uk supercomputer, university of cambridge, xe9640

[allhomeworkassignments1]

Machine Programming - What If Computers Could Self-Program?

What exactly is Machine Programming? The idea of automating the entire software development cycle, including code writing, testing, debugging, and maintenance, is audacious. MP is driven by MIT, Berkley, Intel, Google, and other industry heavyweights and is gaining traction. Need help with your R Programming Assignment Help? Allhomeworkassignments.com has you covered. We offer affordable and reliable support for all your programming needs.

  The main driving force behind this initiative is a vision of the future in which everyone will be able to program computers. It is currently only available to 1% of the world's population. That's right, 99% of the world's population cannot program machines. This could be possible if we can teach machines to understand human intent without having to write code. The machine will do all of the tedious work in the MP vision, including creating the code and ensuring that it achieves the goal.

  Second, as I discussed in my previous post, the world is becoming increasingly heterogeneous. The truth is that no one is capable of programming that many devices. Initiatives such as OneAPI may be of assistance in this regard by providing a simple standardized method of programming various devices. However, creating an efficient implementation of that API will be extremely difficult. I'm thinking of Intel's many performance libraries, which provide highly tuned routines for math primitives, linear algebra, memory management, and so on. This is a coded tone written by experts that is extremely complex. To make the production and maintenance of lower-layer code easier, some form of automation must be implemented. Need help with your Java Programming Assignment Experts? Our experts are here to help! Allhomeworkassignments.com has the best team of programmers to help you get your work done quickly and efficiently. For additional details, visit our site.

  Machine Programming heavily relies on Machine Learning (ML) techniques. However, ML usually allows for some inaccuracy in results. If your iPhone's face recognition feature fails once a month, we can live with it; no one will die. With MP, however, we cannot allow the machine to misinterpret human intent. As a result, Machine Programming employs formal methods to ensure its correctness.

  A High-Level Vision

  Machine Programming is built on three pillars, as defined in the original visionary paper1:

1. It serves as a link between a human and a machine. It enables humans to express their intent in any way they see fit. A UML diagram, pseudocode, or even natural language could be used. The machine should be able to adapt regardless of the format. It's as if you're communicating with the machine. And once it understands what you want, it says, "Okay, give me a few hours, and I'll build it for you." Taking Elon Musk's NeuraLink technology into account, this isn't out of the question.

2. Once the intention is understood, the machine creates all of the components required to achieve the desired goal, such as algorithms and data structures, the need for network communication, and so on. It is still relatively high-level, and the "design" generated by the machine is independent of SW and HW.

  3. The third step in this process is to tailor the "design" (as a result of the invention step) to the specific HW and SW ecosystem, e.g. create an implementation, such as compiling it down to machine code, optimizing it, and verifying that it works. Need help with your C Programming Assignment Help? Allhomeworkassignments.com is the right place for you. We have a team of experienced programmers who can help you with basic questions to more complex problems. Find out more today. Visit our site.