Connect with us

Tech

Graphcore’s Colossus GC200 7nm Chip Competes Against The NVIDIA A100 GPU With Colossal Design & 250 TFLOPs AI Performance – 59.4 Billion Transistors In An 823mm2 Die

Published

3 years ago

on

The AI segment is seeing rapid progress with major tech companies pooling in lots of resources to keep up with the demand for higher performance each year. We’ve seen NVIDIA and AMD actively building next-generation GPUs specifically with AI and HPC in mind but it looks like competition has arrived from British AI chip designer, Graphcore, who has unveiled its second-generation chip for AI which directly competes against NVIDIA’s A100 Tensor Core GPU accelerator.

Graphcore’s GC200 Is A Massive 7nm Chip For AI Tasks Which Is Designed To Compete Against NVIDIA’s A100 GPU – IPU Delivers Up To 250 Teraflops of AI Compute

For this purpose, Graphcore has announced its new Colossus MK2 GC200 IPU or an Intelligence Processing Unit which is designed exclusively to power machine intelligence. Just like its name, the chip itself features a colossus design and delivers an 8x performance bump over its predecessor, the MK1.

“We’re 100% focused on silicon processors for AI, and on building systems that can plug into existing centers. Why would we want to build CPUs or GPUs if those already work well? This is just a different toolbox.” via Graphcore’s CEO, Nigel Toon

The Colossus MK2 GC200 is fabricated on TSMC’s 7nm process node and features a die size of 823 mm2. For comparison, that’s almost as big as the NVIDIA A100 GPU accelerator which measures at 826mm2. The chip is not only a behemoth in terms of size but also in terms of density with a total of 59.4 Billion transistors onboard compared to 54.2 Billion transistors on the NVIDIA A100 GPU. It shows a higher density on the Graphcore chip than NVIDIA’s flagship chip accelerator.

To make the GC200 work, it is configured with 1472 IPU titles, each with an IPU core & In-processor memory. Each IPU core has 6 threads executing in parallel which put the total number of threads in the chip at 8832 (1472 cores / serial processor). For memory, the chip makes use of an on-die solution which offers 900 MB capacity per IPU and a memory bandwidth of 47.5 TB/s. Graphcore has gone with a smaller capacity but the higher-bandwidth solution and stated that you can theoretically get more capacity when using several racks at once and the memory pool would end up higher when compared to a rack composed of A100 GPUs.

Read More

————–

By: Hassan Mujtaba
Title: Graphcore’s Colossus GC200 7nm Chip Competes Against The NVIDIA A100 GPU With Colossal Design & 250 TFLOPs AI Performance – 59.4 Billion Transistors In An 823mm2 Die
Sourced From: wccftech.com/graphcores-colossus-mk2-gc200-7nm-ai-chip-rivals-nvidia-a100-gpu/
Published Date: Wed, 15 Jul 2020 14:56:20 +0000

Did you miss our previous article…
https://getinvestmentadvise.com/tech/whats-in-a-word-overhead/

Related Topics:
Continue Reading

Tech

NVIDIA GeForce RTX 30 Gaming Laptops Powered Massive Cryptocurrency Mining Farm Spotted in China, Over Hundreds of Ampere Laptops

Published

2 years ago

on

February 11, 2021

By

Last week, it was reported how cryptocurrency miners are rushing after NVIDIA GeForce RTX 30 Gaming Laptops to earn that sweet digital currency profits. Well, now a video tour of one of the mining farms has surfaced showcasing several hundred gaming laptops being used to mine the cryptocurrencies.

Several Hundred NVIDIA GeForce RTX 30 Gaming Laptops Spotted In Chinese Cryptocurrency Mining Farm Video Tour

Earlier, we only got to see a glimpse of the cryptocurrency mining farm but a brand new video has surfaced which shows the complete scale of the setup. The setup is based in China but the user has not mentioned the location for obvious reasons. As you can tell, the mining farm is set up with and utilizes NVIDIA GeForce RTX 30 Gaming laptops from the Chinese manufacturer, Hasse.

The Hasse laptop is based on the GeForce RTX 3060 mobility GPU. The GeForce RTX 3060 laptop solution is definitely not similar to the desktop variant which offers much better specifications and doesn’t come at a lower price either. According to NVIDIA themselves, the RTX 3060 laptops start at $999 US & the miners can save a penny by purchasing in bulk but they end up paying at least $600 US more for the laptop variant. Plus, they also end up with unnecessary hardware that the laptops feature. At the same time, the laptop variants do feature reduced TGPs which might save electricity costs & offer higher cryptocurrency mining rates.

pic.twitter.com/XJrfH9Fx7V

— 포시포시 (@harukaze5719) February 9, 2021

What’s interesting here is that the user has bought these laptops in a bulk order which means that either the retailer, distributor, or the manufacturer itself is openly supporting cryptocurrency miners over actual customers such as gamers which these laptops are primarily designed for. The selling party could be taking advantage of crypto miners by selling these at scalped prices to earn a profit.

Read More

————–

By: Hassan Mujtaba
Title: NVIDIA GeForce RTX 30 Gaming Laptops Powered Massive Cryptocurrency Mining Farm Spotted in China, Over Hundreds of Ampere Laptops
Sourced From: wccftech.com/nvidia-geforce-rtx-30-gaming-laptops-powered-massive-cryptocurrency-mining-farm-spotted-in-china-over-hundreds-of-ampere-laptops/
Published Date: Thu, 11 Feb 2021 04:17:52 +0000

Did you miss our previous article…
https://getinvestmentadvise.com/tech/intel-3rd-gen-xeon-ice-lake-sp-cpu-rumors-10nm-with-up-to-40-cores-8-channel-ddr4-3200-64-pcie-gen-4-0-lanes-max-270w-tdp/

Continue Reading

Tech

Intel 3rd Gen Xeon Ice Lake-SP CPU Rumors – 10nm With Up To 40 Cores, 8-Channel DDR4-3200, 64 PCIe Gen 4.0 Lanes, Max 270W TDP

Published

2 years ago

on

February 11, 2021

By

A couple of rumors regarding Intel’s Ice Lake-SP Xeon server CPUs have surfaced over the web, mostly from known leaker, Momomo_US. The information consists of specifications, configs & power ratings of several SKUs that will be part of the 3rd Gen Xeon Scalable processor family.

Intel 3rd Gen Ice Lake-SP Xeon CPU Rumors Point Out Up To 40 Cores, 270W TDPs & Brand New Core Architecture

Starting off with the details, Intel’s 3rd Gen Ice Lake-SP Xeon CPUs will be based on the 10nm+ process node and utilize the Sunny Cove core architecture. The Intel Sunny Cove x86 architecture has been around since 2019 and was first featured on Intel’s 10th Gen Ice Lake processors for the notebook segment. Intel has since moved over to Tiger Lake which is based on the Willow Cove x86 architecture and makes use of the 10nm SuperFin process node.

Some of the major upgrades that Intel’s 10nm+ for Ice Lake-SP Xeon CPU will deliver include:

2.7x density scaling vs 14nmSelf-aligned Quad-PatterningContact Over Active GateCobalt

Read More

————–

By: Hassan Mujtaba
Title: Intel 3rd Gen Xeon Ice Lake-SP CPU Rumors – 10nm With Up To 40 Cores, 8-Channel DDR4-3200, 64 PCIe Gen 4.0 Lanes, Max 270W TDP
Sourced From: wccftech.com/intel-3rd-gen-xeon-ice-lake-sp-cpu-rumors-10nm-up-to-40-cores-270w-tdp/
Published Date: Thu, 11 Feb 2021 10:02:51 +0000

Did you miss our previous article…
https://getinvestmentadvise.com/tech/intels-raja-koduri-teases-xe-hpg-enthusiast-gaming-graphics-cards-in-action-within-3dmark-mesh-shader-test-launch-imminent/

Continue Reading

Tech

Intel’s Raja Koduri Teases Xe HPG Enthusiast Gaming Graphics Cards In Action Within 3DMark Mesh Shader Test, Launch Imminent?

Published

2 years ago

on

February 11, 2021

By

Intel’s Senior Vice President, Raja Koduri, has been dropping some major bombs on his Twitter feed for a while now. The most recent one being the Xe HPC chip shot which is an engineering marvel but it looks like Raja has switched gears and now dropped a teaser of Intel’s first gaming discrete graphics cards, the Xe HPG.

Intel Teases Xe HPG Enthusiast Gaming Graphics Cards, Running 3DMark’s Upcoming Mash Shader Feature

It’s been some time since we last heard about Intel’s Xe HPG (High-Performance Gaming) discrete GPUs or graphics cards. Intel’s main focus has been Xe-LP and they also dive in deeper within the tech that Xe HPC has to offer considering it’s their biggest chip design to date and the flagship Xe product. But the Xe HPG is where the gaming community is more focused at and we have got the first teaser of Intel’s discrete graphics cards based on the HPG GPUs running in action.

Xe HPG mesh shading in action, with the UL 3DMark Mesh Shader Feature test that is coming out soon pic.twitter.com/fnYeWoM08c

— Raja Koduri (@Rajaontheedge) February 10, 2021

The teaser from Raja shows an undisclosed Intel Xe HPG GPU based discrete graphics card running the UL 3DMark Mesh Shader Feature test. This test is coming out soon and will be integrated within the 3DMark test suite. Mesh Shader is one of the major features introduced within DirectX 12 Ultimate API back in 2020. The feature is supported by NVIDIA’s Turing, Ampere, and AMD’s RDNA 2 graphics architectures. Based on this tweet, Intel will also be adding Mesh Shaders support on its Xe-HPG class discrete GPUs.

The technology helps to dramatically improve performance and image quality when rendering a scene with a substantial number of very complex objects. Take for instance a very complex & triangle-heavy mesh and what Mesh Shaders would essentially do is segment it into smaller meshlets. Each meshlet ideally optimizes the vertex re-use within it. Using the new hardware stages and this segmentation scheme, devs can render more geometry

Read More

————–

By: Hassan Mujtaba
Title: Intel’s Raja Koduri Teases Xe HPG Enthusiast Gaming Graphics Cards In Action Within 3DMark Mesh Shader Test, Launch Imminent?
Sourced From: wccftech.com/intels-raja-koduri-teases-xe-hpg-enthusiast-gaming-graphics-cards-in-action-3dmark-mesh-shader-test/
Published Date: Wed, 10 Feb 2021 07:21:19 +0000

Continue Reading

Trending