Nvidia's Grace AI chip leaves Intel processors behind

Nvidia's Grace AI chip leaves Intel processors behind

Nvidia has a new chip in the works for boosting artificial intelligence and other high-performance computing work: Grace, a design slated to arrive in mammoth supercomputers in 2023. Instead of accelerating conventional Intel-powered servers, though, the design includes its own built-in Arm processors.

Nvidia's current brainiest chip, the A100, is typically yoked to Intel Xeon processors. Nvidia chips do the grunt work, but Intel chips oversee it. With Grace, named after pioneering programmer Grace Hopper, the company opted to embed several Arm Neoverse processor cores within the chip to speed up processing, said Paresh Kharya, an Nvidia senior director. The chip news arrived at Nvidia's GTC 2021 conference this week.

The new chip should let AI customers run computing tasks that are vastly more complex than is possible with today's chip designs, a step toward the general artificial intelligence that is the holy grail of today's machine learning research, said Cambrian AI Researach analyst Karl Freund in a blog post.

The design illustrates Nvidia's dramatic ascent -- and Intel's struggles. Even decades of dominance in technology don't guarantee success when the rules of computing are constantly being rewritten. Your laptop likely comes with an Intel chip, but an Nvidia chip was more likely responsible for important AI work like filtering spam, improving image quality or recognizing your voice when you call your bank.

Not so many years ago, Nvidia was just a component supplier, a designer of graphics chips called GPUs to boost PC performance. Intel's family of processors, or perhaps compatible rival AMD chips, shouldered most of the computing work. Intel, though, has struggled in recent years to keep pace with chip miniaturization and to capitalize on the exploding use of AI.

The result: Nvidia's market capitalization vaulted over Intel's, reaching $357 billion compared with Intel's $278 billion. Much of the growth has been propelled by the fact that GPUs also turned out to be pretty good at AI work, specifically the computationally intense training process that builds the models that later run in data centers, PCs and phones.

Also in the ascendant is Arm, which licenses the chip designs and technology that power every smartphone, new M1-based Apple Macs and the world's fastest supercomputer. Nvidia is seeking to acquire Arm for $40 billion, a move some rivals like Qualcomm object to. Grace's integrated Arm chips let Nvidia read data from memory many times faster than with current designs, the company said.

Nvidia's Selene machine, currently the world's fifth-fastest supercomputer, pairs A100 chips with AMD Epyc CPUs. A 2023 Grace-based machine called Alps at Switzerland's National Supercomputing Center should be seven times faster, Kharya said. The Los Alamos National Laboratory in the US also will buy a Grace-powered supercomputer.

Under new Chief Executive Pat Gelsinger, Intel is working to reclaim its manufacturing lead, planning to tap into others' manufacturing abilities while it works on miniaturizing its circuitry inscribing technology.

Intel is building AI abilities into its main processors while working on dedicated hardware, too. It folded its Nervana chips operation, but its Habana AI acceleration processors are still under active development.

One hot area for AI chips is autonomous vehicles, whose self-driving algorithms rely on processing in camera imagery and other sensor data. It's a core focus for Nvidia AI chip work, for example with its Orin chip scheduled to debut in 2022 vehicles.

Nvidia CEO Jensen Huang announced new processors for AI, graphics and supercomputing at the company's GTC event.

Screenshot by Stephen Shankland/CNET

At GTC, Nvidia announced a new chip called Atlan with quadruple the performance. It should arrive in 2025 vehicles, said Danny Shapiro, Nvidia's senior director of automotive work. Like Orin and Grace, Atlan relies on Arm cores, too.

Nvidia also announced a grander autonomous vehicle technology package called Hyperion 8. It combines two Orin processors with a host of sensors: eight exterior cameras, four exterior wider-angle fisheye cameras, three interior cameras, nine radar scanners and one lidar 3D scanner. The technology should arrive later in 2021.

Nvidia extended a partnership with Volvo, the companies said. Volvo plans to use Orin chips in its next-generation vehicles.

Intel has its own autonomous vehicle division, Mobileye. Tesla develops its own AI chips for its cars. 

---

Source

Tags:

• Nvidia S Grace Ai Chip Leaves Intellicast
• Nvidia S Grace Ai Chip Leaves Inteletravel
• Nvidia S Grace Ai Chip Leaves Intel Drivers
• Nvidia S Grace Ai Chip Leaves Intellectual Property
• Nvidia S Grace Ai Chip Leaves Of Grass
• Nvidia S Grace Ai Chip Leaves Clip
• Nvidia S Grace Aimi
• Nvidia S Grace Air Dryer
• Nvidia S Grace Aitchison
• Nvidia S Grace Airport
• Nvidia S Grace Ai Ling
• Nvidia Stock

Nvidia to buy SoftBank's Arm chip division for $40 billion

Nvidia to buy SoftBank's Arm chip division for $40 billion

Nvidia has acquired SoftBank's Arm chip division for $40 billion in cash and stock in the chip industry's largest deal ever. As part of the deal, made this Sunday, SoftBank will take an ownership stake in Nvidia that's expected to be less than 10%, the companies said in a joint statement.

Bloomberg reported last week that Nvidia and SoftBank were in advanced talks, with Nvidia the lone potential buyer. That followed an earlier report by The Wall Street Journal that SoftBank was considering a sale of Arm. The Journal also reported Saturday's news of an imminent Nvidia-SoftBank agreement.

Arm isn't as well-known as mega chip companies such as Qualcomm and Intel, but its work lies behind the processors inside many of the world's mobile phones.

Arm licenses designs to companies like Qualcomm but also licenses its chip instruction set -- the collection of commands software can use to control it -- to companies like Apple that design their own. Arm's designs are also used as the basis for chips made by Samsung and Nvidia.

In June, Apple said it would overhaul its Mac computers with its own Arm chips, which are similar to the ones it designs for iPhones and iPads, moving away from the Intel processors it has used for the past 14 years. Arm licenses its chip instruction set -- the collection of commands software can use to control a chip -- to companies like Apple that design their own processors.

SoftBank purchased the UK-based Arm in 2016 for $32 billion with the intent of bolstering its internet of things division. Nvidia said it expects the tie-up to boost its artificial intelligence ambitions.

"AI is the most powerful technology force of our time and has launched a new wave of computing," Nvidia founder and CEO Jensen Huang said in a statement. "In the years ahead, trillions of computers running AI will create a new internet-of-things that is thousands of times larger than today's internet-of-people."

The companies said they expect the deal to close in 18 months, noting that it will require approval of the US, UK, EU and China.

---

Source

Tags:

• Arm Bought By Nvidia
• Arm Sale To Nvidia
• Nvidia To Buy Arm
• Nvidia And Arm Deal
• Geforce Now For Softbank
• Geforce Now Softbank Au
• Is Softbank A Buy
• Geforce Now Softbank Au
• Where To Buy Nvidia Stock
• Best Buy Nvidia Graphics Card

Apple's M1 Pro and M1 Max chips mean new trouble for Intel

Apple's M1 Pro and M1 Max chips mean new trouble for Intel

A year ago, Apple announced it was taking on Intel's most efficient chips by introducing lightweight MacBook laptops powered by the M1, a homegrown processor. On Monday, the consumer electronics giant expanded its challenge, launching MacBook Pro laptops built around the new M1 Pro and M1 Max that take on Intel's beefier chips.

The new MacBook Pros bode well for Apple's attempt to take firmer control over its products. And they're bad news for Intel, whose chips Apple is ejecting from its Macs after a 15-year partnership. It's a loss of revenue, prestige and orders to keep its factories running at full capacity.

"Intel has completely lost the Mac and is unlikely to regain it any time soon," New Street Research analyst Pierre Ferragu said in a research note Tuesday.

Intel didn't lose this big customer overnight. The company that was once synonymous with consumer computers -- remember Intel Inside? -- fell on hard times because of difficulties upgrading its manufacturing. New CEO Pat Gelsinger has started an Intel recovery plan, including an effort to revitalize manufacturing progress. But turning around a behemoth requires patience. 

Meet the Mac's new chips

Intel's troubles encouraged Apple to develop its own chip expertise and technology for computers. (It already designed its own A-series chips for the iPhone and iPad, and indeed the M-series chips capitalize on that investment.) The company's M1 processors, which came in last year's MacBook Air and low-end 13-inch MacBook Pro, were evidence it wanted to take control of its own future.

The M1 Pro and M1 Max demonstrate the company's increasing power as a chip designer. Both are designed for more capable models, the 14-inch and 16-inch Pros, geared for video editors, programmers and others with intense computing needs. The heft of the chips -- each of which sports eight performance and two efficiency cores, compared with the M1's four-by-four design -- is intended to sustain heavy work. They also come with much more powerful graphics processing power and memory, up to 16GB for the M1 Pro and 64GB for the M1 Max.

Miniaturization is what lets chip manufacturers economically squeeze in more transistors, a chip's electronic circuitry elements. The new M1 models are doozies of miniaturization, with 34 billion transistors in the M1 Pro and 57 billion in the M1 Max. That's how it could add special chip modules for graphics, video, AI, communications and security into its high-end MacBook Pros.

Intel's troubles

Intel, which for decades has led the world in chip technology, suffered for the last half decade as an upgrade to its manufacturing technology dragged on longer than the usual two years. The company's problem came as it tried to move from a 14-nanometer manufacturing process to 10nm, the next "node" of progress. (A nanometer is a billionth of a meter.)

Intel didn't respond to a request for comment. Apple didn't comment for this story.

Apple's chip foundry, Taiwan Semiconductor Manufacturing Co., took advantage of Intel's lag to the benefit of Apple, Nvidia, AMD and other Intel rivals. It now leads in electronics miniaturization and the all-important measurement of performance per watt of power consumed. 

The result is the M1 Pro and M1 Max, which according to Apple's measurements are 1.7 times faster than Intel's current eight-core Tiger Lake chips, formally called 11th generation Core. Compared differently, the M1 Pro and Max consume 70% less power than the Tiger Lake chips at the same performance level.

Apple doesn't reveal which speed tests it uses, so the results are hard to validate at this stage. The consensus, however, is that the performance claims are valid in broad terms.

"I am overall impressed at what Apple has been able to do on the latest process from TSMC," said Patrick Moorhead, analyst at Moor Insights and Strategy. He estimates that Apple saves a few hundred dollars per laptop because it doesn't have to buy Intel processors, although it spends a lot of that money designing its chips.

Don't count Intel out yet

To be sure, Intel won't be hurt badly by the loss of Apple's business. The company has plenty of other business. The vast majority of Windows PCs still use x86 processors from Intel and AMD. And customers only rarely change from Windows to MacOS or vice versa.

It also doesn't have a lot of competition. Apple doesn't license its chips to others, and Qualcomm's efforts to sell processors to PC makers has been a limited success at best. 

Intel mostly has to worry about AMD, which makes increasingly capable chips but still trails in market share.

Intel also has its Alder Lake processor, scheduled for later this year, and Meteor Lake processor, coming in 2023, to generate excitement. The chips will bring speed boosts in part by adopting a combination of performance and efficiency cores, just like the M1 does, and by adopting the new Intel 7 and Intel 4 manufacturing processes.

Still, Apple has taken wind out of Intel's sails. Intel may narrow the gap as its new chips hit the market. But in the meantime, Apple's M series could help it steal market share from Windows computers, Intel's stronghold.

---

Source

Tags:

• Apple M1 Max Chips
• Apple Macbook Pro With M1 Max Chip
• Apple M1 Pro And M1 Max Chips
• Apple M1 Pro And Max
• Apple M1 Processor
• Apple Stock
• Apple Stock Price
• Apple Support

NVIDIA's new GPU is better than the Titan X and $500 cheaper

NVIDIA's new GPU is better than the Titan X and $500 cheaper

NVIDIA's GTX 1080 is no longer the king of the GPU mountain. That honour is now the company's newest graphics card, the GTX 1080 Ti.

Announced at a GDC press conference on Wednesday in San Francisco, the $700 1080 Ti is 35 percent more powerful than the GTX 1080, says NVIDIA CEO Jen-Hsun Huang.

It comes with 11GB of GDDR5X memory, and with 11GB of frame buffering it is designed for 4K gaming. In fact, Huang reassured his audience that they could play Watch Dogs 2 in 5K -- that's 5,120×2,880 -- using the GPU.

Most impressively, the company claims the 1080 Ti is more powerful than the Titan X, a $1,200 GPU developed for AI and deep learning. Like the Titan X, the 1080 Ti use 3,584 of NVIDIA's CUDA cores. The GTX 1080, by comparison, has 8GB of memory and 2,560 CUDA cores.

Huang also announced at the conference that the 1080 was getting a price drop down to $500.

The GTX 1080 Ti will be available from March 10 for $700. No international pricing has been announced, but that converts to AU$915 and £565.

---

Source

Tags:

• Nvidia Titan X Gpu
• Nvidia Titan V Gpu
• Titan X Gpu Cost
• Nvidia Titan V Gpu
• Nvidia Titan X Gpu
• Intel Gpu Vs Nvidia
• Which Is Better Intel Or Nvidia
• Which Is Better Intel Graphics Or Nvidia
• Installing New Nvidia Gpu
• Nvidia S New Gpu
• Nvidia News
• Nvidia Stock News

Intel s core i9 11980hk leads the way for gaming and depression intel s core i9 11980hk leads the way intel s core i9 11980hk leads the efforts intel s core i9 11980hk leads the pack intel s core i9 11980hk leadsonline intel s core i9 11980hk leads definition intel s core i9 11980hk processor intel s core values intel share price intel stock forecast intel supplier

Intel's Core i9-11980HK leads the way for gaming and creative laptops

Intel's Core i9-11980HK leads the way for gaming and creative laptops

If it weren't for the ancillary technologies that come with Intel's latest round of Tiger Lake Core H-series CPUs, today's 11th-gen launch could seem like kind of a snoozefest. Yes, these are the first of the high-power mobile gaming-and-creative-targeted CPUs built on the company's 10-nanometer SuperFin process -- tech that essentially improves current handling to deliver improved performance -- led by an always notable flagship i9, the i9-11980HK. Yes, the i9 is faster than its 10th-gen predecessor. And yes, Intel promises that it's faster than AMD's offerings. Frankly, it would be newsworthy if Intel didn't make those claims. But these chips are basically just an expansion of the line Intel rolled out at CES 2021.

As has become habit with mobile processor launches, Nvidia and Intel have been making them in tandem. At the same time as the new Tiger Lake-H series launch, Nvidia revealed its low-end RTX 3050 and 3050 Ti mobile GPUs.   

More of today's news:

That's where Intel debuted the Tiger Lake-H architecture and related process, which (in conjunction with the 500-series chipset) adds support for Thunderbolt 4, Killer Wi-Fi 6E/Gig+, DDR4-3200 memory, dual built-in displays, Optane H20 and 20 lanes of PCIe Gen 4.  

Because the Gen 4 allows direct connection to the CPU rather than using a separate bus, it brings with it a couple of notable capabilities for power users. One is Resizeable BAR, which allows the system to allocate an optimal amount of video memory for the CPU to use for graphics operations not otherwise run on the GPU. That means it takes less time to move the graphics data for rendering out to the display, and can eke out some extra graphics performance. (It's similar to AMD Smart Access Memory, which debuted with the Radeon RX 6000 series desktop cards in October 2019.)  It also lets manufacturers incorporate bootable SSD RAID arrays using Intel's Rapid Storage Technology. So speedier storage in larger capacities.   

Specifications

CPU	Cores / threads	Cache	TDP	Base frequency (GHz)	Max single core frequency (GHz)	Max all core frequency (GHz)
Core i9-11980HK	8/16	24MB	65W	2.6	5	4.5
Core i9-11900H	8/16	24MB	35W	2.5	4.9	4.4
Core i7-11800H	8/16	24MB	35W	2.3	4.6	4.2
Core i5-11400H	6/12	12MB	35W	2.7	4.5	4.1
Core i5-11260H	6/12	12MB	35W	2.6	4.4	4

Just because the processor and chipset support these capabilities doesn't mean you'll see them in all laptops; some of them, such as implementing PCIe Gen 4, are subject to individual manufacturers' preferences and product-line strategies. The i7 and i9 carry on Intel's incorporation of Turbo Boost 3, notable for its automatic selection of the fastest and most reliable core to boost to the max for single-threaded operations.

There are commonalities across all the CPUs, including integrated Intel UHD Graphics. Intel has stressed that the integrated GPU uses its latest Xe graphics architecture, but as with its desktop 11th-gen (Rocket Lake-S) CPUs chose to brand it with the old, old UHD Graphics nomenclature. That's because one of Intel's requirements for it to carry the Iris Xe brand is at least 80 execution units and these H series chips only have 32 EUs. These CPUs are intended for use in laptops with discrete graphics, so that paucity of EUs can be a minor, if irritating, drawback.

Intel also announced its Tiger Lake-H commercial processors, both Core and Xeon, which use the secure, managed vPro chipset.

---

Source