Login

Feed Post

Payal Manghnani

Stealth • 4m

Computing has come a long way with the introduction of NPUs, GPUs, and CPUs. Let's break it down in simple terms: 🧠 CPUs: They handle general instructions and system management. Think of them as the brain of your computer, evolving with more cores and speed. 🎮 GPUs: These are like supercharged CPUs that excel at parallel processing tasks, like matrix operations and vector calculations. They are ideal for graphics and AI because they have thousands of cores. 🤖 NPUs: NPUs are specifically designed for AI tasks, such as matrix multiplication and low latency. Google calls them TPUs, while Apple refers to them as Neural Engines. So, why are NPUs important? They are crucial for edge computing, which allows local AI processing. This reduces costs, enhances privacy, and increases efficiency. NPUs can handle many tasks locally and offload complex ones to the cloud, making AI more energy-efficient and capable. In a nutshell, CPUs, GPUs, and NPUs each have their unique roles, with NPUs be

4 replies6 likes
Replies (4)

More like this

Recommendations from Medial

Image Description

Haran Naresh K

Stealth • 17d

Harnessing GPU Power with CUDA CUDA (Compute Unified Device Architecture) is a parallel computing platform by Nvidia that unleashes the power of GPUs for more than just graphics rendering. Initially developed in 2007, CUDA enables massive parallel p

See More
1 replies4 likes
Image Description
Image Description

Inactive

Stealth • 4m

Could AI workload processing be the new crypto mining? Some companies think so, offering ways to earn by lending your GPU power for AI tasks, just like the old crypto mining days. Key points: - Services like Salad pay users to contribute their GP

See More
5 replies13 likes

Vaibhav Babruwan Shingde

Stealth • 8m

Let's Start With Today's Topic : How AI Chips Work , Uses , Companies Working On It? ° How AI Chips Work? •Large arrays of specialized processing cores for AI operations. •High-speed on-chip memory for storing AI models and data. •Fast interconn

See More
0 replies7 likes

Banhishikha Ghosh

Stealth • 1m

Edge AI and Edge Computing: Transforming Technology Edge AI and Edge Computing are revolutionizing how data is processed by enabling local computation on devices, minimizing reliance on centralized cloud servers. Edge computing processes data closer

See More
0 replies6 likes
Image Description
Image Description

DIVYANSHU MHATRE

Stealth • 4m

Kids nowadays think that the school is matrix but the real matrix is the social media they are using

12 replies26 likes
Image Description
Image Description

Vaibhav Babruwan Shingde

Stealth • 8m

Hello everyone, Do you think NVIDIA is the king of the artificial intelligence industry? With the need for GPUs in AI deployment and processing, will NVIDIA continue to dominate the market? Let me know your thoughts.🤔🙂

4 replies5 likes
Image Description
Image Description

Afifa

Stealth • 2m

NVIDIA- THE DOMINANT FORCE 1. NVIDIA is a leading force in AI and GPU technology. 2. Their GPUs, like the H100 Tensor Core, are critical for AI development, including training models like ChatGPT. 3. NVIDIA's stock has surpassed a $1 trillion ma

See More
2 replies6 likes
1
Image Description

Mr Z

 • 

Medial • 2m

Krutrim, the AI startup founded by Ola Cabs' CEO Bhavish Aggarwal, plans to design, build, and launch its first AI chip by 2026:🇮🇳 Bodhi 1 (For AI): 2026 Bodhi 2 (For AI): 2028 Sarv 1 (General computing): 2026 Ojas (Edge computing): Date not re

See More
2 replies12 likes
2

Sunil Sharma

Stealth • 6m

AI is the new future and doing the tasks which are may be not possible for humans. so in which race you are: Building AI to be the part of innovation and future. Using AI to make life easier and wealthier.

0 replies6 likes
Image Description

Rahul Gupta

Stealth • 4m

The Scaling Laws are doubling the power of AI systems every 6 months, leading to an intelligence revolution and a golden age of computing. ~ Satya Nadella

1 replies3 likes

Download the medial app to read full posts, comements and news.