"As a part of aligning our resources with our largest growth opportunities, we are taking a number of targeted steps," an AMD spokesperson told Reuters on Tuesday.

Kettle/batch cooker for potato and vegetable chips is offered by FOODesign Machinery & Systems with proven capacities of up to 650 – 700 pounds per hour.  

Great Plains Holdings Inc. will acquire 51 percent of Bonjoe Gourmet Chips LLC, as well as have to option to acquire another 20 percent interest in the gourmet chip maker.

Two new popchips varieties from popchips inc. will hit store shelves this summer: Crazy Hot Potato and Cinnamon Twist Sweet Potato.

Light, leavened naan bread is the model for Maya Kaimal Fine Indian Foods’ double-baked Maya Kaimal Naan Chips.

Axium Foods’ 130,000-sq.-ft, corn-based snacks operation in South Beloit, Ill., has expanded several times, thanks in part to the success of the many private-label products it packs and products it copacks. It’s also launching its own line of Mystic Harvest purple tortilla chips, which contain powerful antioxidants. Outfitted with seven highly flexible production lines, Axium produces 150 different stock-keeping units of snacks each week.

How an Upstate NY school now heats with locally produced wood chips.

The Hurricanes landed four-star defensive lineman Hayden Lowe, flipping him from USC.

OKI Develops Ultracompact Photonic Integrated Circuit Chip Using Silicon Photonics Technology to Realize Various Optical Sensors  Business Wire

La 9e édition du Dictionnaire de l’Académie française: La 9e édition du Dictionnaire de l’Académie française, dont la publication a commencé dans les années 1980, s’est achevée en novembre 2024, avec la parution du tome 4 aux éditions Fayard. The 9th edition of the Dictionary of the French Academy, whose publication began in the 1980s, […]

"Giving up on Nvidia here after its hit — Hopper — is like giving up on Apple at iPhone 1 or 2," Melius Research said.

The US is considering expanding its export limits on AI chips to China and its allies. Experts say the race for AI dominance is like a new Cold War.

SEOUL, South Korea, Nov. 13, 2024 — SEMIFIVE, a leading design solution provider and pioneer of platform-based custom silicon solutions, has announced its collaboration with Synopsys to develop a cutting-edge […]

The post SEMIFIVE and Synopsys Collaborate on HPC Chiplet Platform for Advanced Multi-Die Designs appeared first on HPCwire.

Snapdragon 8 Elite chipset was launched by Qualcomm at its Snapdragon Summit in Hawaii last month as a successor to the Snapdragon 8 Gen 3. The rumour mill now suggests that its successor, the purported Snapdragon 8 Elite 2 SoC, might offer an even bigger jump in single and multi-core performance, despite recent changes suggested in its development plans.

The new chips are designed for smart home setups.

The post Qualcomm Focuses on IoT Tech with its Newest Chips appeared first on Phandroid.

Washington, United States — More than a dozen US-based employees of Taiwanese chipmaking giant TSMC have sued the company for apparent discrimination against non-Asian workers, according to a recent lawsuit. The suit alleges that the world’s largest contract manufacturer of chips — used in everything from Apple iPhones to Nvidia’s AI hardware — unfairly favors […]...

Keep on reading: Taiwan chip giant TSMC sued for alleged discrimination

Nvidia, Oracle, Google, Dell and 13 other companies reported how long it takes their computers to train the key neural networks in use today. Among those results were the first glimpse of Nvidia’s next generation GPU, the B200, and Google’s upcoming accelerator, called Trillium. The B200 posted a doubling of performance on some tests versus today’s workhorse Nvidia chip, the H100. And Trillium delivered nearly a four-fold boost over the chip Google tested in 2023.

The benchmark tests, called MLPerf v4.1, consist of six tasks: recommendation, the pre-training of the large language models (LLM) GPT-3 and BERT-large, the fine tuning of the Llama 2 70B large language model, object detection, graph node classification, and image generation.

Training GPT-3 is such a mammoth task that it’d be impractical to do the whole thing just to deliver a benchmark. Instead, the test is to train it to a point that experts have determined means it is likely to reach the goal if you kept going. For Llama 2 70B, the goal is not to train the LLM from scratch, but to take an already trained model and fine-tune it so it’s specialized in a particular expertise—in this case, government documents. Graph node classification is a type of machine learning used in fraud detection and drug discovery.

As what’s important in AI has evolved, mostly toward using generative AI, the set of tests has changed. This latest version of MLPerf marks a complete changeover in what’s being tested since the benchmark effort began. “At this point all of the original benchmarks have been phased out,” says David Kanter, who leads the benchmark effort at MLCommons. In the previous round it was taking mere seconds to perform some of the benchmarks.

Performance of the best machine learning systems on various benchmarks has outpaced what would be expected if gains were solely from Moore’s Law [blue line]. Solid line represent current benchmarks. Dashed lines represent benchmarks that have now been retired, because they are no longer industrially relevant.MLCommons

According to MLPerf’s calculations, AI training on the new suite of benchmarks is improving at about twice the rate one would expect from Moore’s Law. As the years have gone on, results have plateaued more quickly than they did at the start of MLPerf’s reign. Kanter attributes this mostly to the fact that companies have figured out how to do the benchmark tests on very large systems. Over time, Nvidia, Google, and others have developed software and network technology that allows for near linear scaling—doubling the processors cuts training time roughly in half.

First Nvidia Blackwell training results

This round marked the first training tests for Nvidia’s next GPU architecture, called Blackwell. For the GPT-3 training and LLM fine-tuning, the Blackwell (B200) roughly doubled the performance of the H100 on a per-GPU basis. The gains were a little less robust but still substantial for recommender systems and image generation—64 percent and 62 percent, respectively.

The Blackwell architecture, embodied in the Nvidia B200 GPU, continues an ongoing trend toward using less and less precise numbers to speed up AI. For certain parts of transformer neural networks such as ChatGPT, Llama2, and Stable Diffusion, the Nvidia H100 and H200 use 8-bit floating point numbers. The B200 brings that down to just 4 bits.

Google debuts 6th gen hardware

Google showed the first results for its 6^th generation of TPU, called Trillium—which it unveiled only last month—and a second round of results for its 5^th generation variant, the Cloud TPU v5p. In the 2023 edition, the search giant entered a different variant of the 5^th generation TPU, v5e, designed more for efficiency than performance. Versus the latter, Trillium delivers as much as a 3.8-fold performance boost on the GPT-3 training task.

But versus everyone’s arch-rival Nvidia, things weren’t as rosy. A system made up of 6,144 TPU v5ps reached the GPT-3 training checkpoint in 11.77 minutes, placing a distant second to an 11,616-Nvidia H100 system, which accomplished the task in about 3.44 minutes. That top TPU system was only about 25 seconds faster than an H100 computer half its size.

A Dell Technologies computer fine-tuned the Llama 2 70B large language model using about 75 cents worth of electricity.

In the closest head-to-head comparison between v5p and Trillium, with each system made up of 2048 TPUs, the upcoming Trillium shaved a solid 2 minutes off of the GPT-3 training time, nearly an 8 percent improvement on v5p’s 29.6 minutes. Another difference between the Trillium and v5p entries is that Trillium is paired with AMD Epyc CPUs instead of the v5p’s Intel Xeons.

Google also trained the image generator, Stable Diffusion, with the Cloud TPU v5p. At 2.6 billion parameters, Stable Diffusion is a light enough lift that MLPerf contestants are asked to train it to convergence instead of just to a checkpoint, as with GPT-3. A 1024 TPU system ranked second, finishing the job in 2 minutes 26 seconds, about a minute behind the same size system made up of Nvidia H100s.

Training power is still opaque

The steep energy cost of training neural networks has long been a source of concern. MLPerf is only beginning to measure this. Dell Technologies was the sole entrant in the energy category, with an eight-server system containing 64 Nvidia H100 GPUs and 16 Intel Xeon Platinum CPUs. The only measurement made was in the LLM fine-tuning task (Llama2 70B). The system consumed 16.4 megajoules during its 5-minute run, for an average power of 5.4 kilowatts. That means about 75 cents of electricity at the average cost in the United States.

While it doesn’t say much on its own, the result does potentially provide a ballpark for the power consumption of similar systems. Oracle, for example, reported a close performance result—4 minutes 45 seconds—using the same number and types of CPUs and GPUs.

Red Magic 10 Pro+ and Red Magic 10 Pro were launched in China as the latest gaming-focused smartphones from the ZTE sub-brand Nubia. Both models have similar internals but their battery packs and...

Hats off to Louis Gerbarg on his excellent write-up of how a Flash chip manages reads, writes, and most interestingly – deletes.  I have a relatively comprehensive understanding of how file systems work from classes back at CMU.  But back when I was taking those classes, Flash storage was basically non-existent in a file system [...]

(EMAILWIRE.COM, November 01, 2024 ) Artificial Intelligence Chipset Market size was valued US$ 20.76 Bn in 2023 and the total revenue is expected to grow at 40% from 2024 to 2030, reaching US$ 218.85 Bn. by 2030. The AI chipset market is experiencing rapid growth, driven by increasing adoption...

Published at LXer: The Radxa ROCK 2F is a small computing device designed for a wide range of uses, from development projects to multimedia setups. It�s packed with features, including multiple...

Nutritious and healthy snack alternative with almond flour and whey containiing 10 grams of protein.

The grudge-match between AMD and Intel continues - although AMD is the clear aggressor in all of it. This week, in a follow up to AMD's antitrust lawsuit over Intel's "abuse of monopoly power" the guantlet was thrown down for...

February 14, 2006 – ACCESS has been concerned for months that the State Department’s plan to include RFID chips in passports has not been well thought out. The original plan would have placed an RFID chip in every passport which would have contained unencrypted data about the passport holder. After a storm of protest, the State Department revised the standard to include some encryption. Now, a Dutch television news program has along and broken that encryption in less than two hours. The ramifications of this to passport holders are anything but positive.

Dr Gordon Moore, one of the three co-founders of Intel Corporation has passed away in Hawaii at the age of 94.  I had the privilege of interviewing him in London in March 1997, and he came across as a genuinely humble man, typical of the genuine ‘boffin’ type of the time. The thing I remember mostly from the interview was his professed love for antiques. Apparently he and his wife, Betty, spent a lot of their free time browsing for...

The post Gordon Moore RIP – Goodbye Mr Chips appeared first on The Red Ferret Journal.

In which Nicole shares what she learned at the recent Whole Good News Conference and the talk she gave there. Also a lesson in time about praying while sick. Mentioned in this episode: The Whole Good News Conference hosted by MissioAlliance (https://www.missioalliance.org/twgn) The Life of Moses by Gregory of Nyssa The Emotionally Healthy Woman by Lisa Scazzero Irresistible: the Rise of Addictive Technology and the Business of Keeping Us Hooked by Adam Alter The Ruthless Elimination of Hurry by John Mark Comer The intro and outro of this podcast are the songs "Idea" and "Remedy for Melancholy" by Kai Engel, available under a CC BY-NC 4.0 license.

Archippus was the son of Saints Philemon (Nov. 22) and Apphia (Feb. 15), and, like them, was a disciple of the Apostle Paul, who calls him "our fellow soldier" (Philm. 2). He and his father preached the Gospel at Colossae, and Archippus probably served as a priest for the church that gathered there at his family's house (Col. 4:17). Archippus' fervor in preaching the Gospel of Christ so angered the pagans that they seized him and brought him before the governor Androcles. When the Saint refused to sacrifice to Artemis, he was stripped, beaten, tormented in various ways, and finally stoned to death.

Philemon is addressed in the Epistle of the holy Apostle Paul that bears his name. He was a nobleman from Colossae; Apphia was his wife. Archippus was Bishop of Colossae. All three were disciples of the Apostle Paul. Onesimus was a pagan slave of Philemon, who stole from his master and fled to Rome. There St Paul led him to faith in Christ, and wrote the Epistle to Philemon, urging Onesimus' master to forgive him and take him back as a brother in Christ. This Philemon did, and Onesimus later became a bishop. In Greece he is venerated as the patron Saint of the imprisoned. All of these holy followers of Christ died as martyrs, stoned to death by pagans.

Archippus was the son of Saints Philemon (Nov. 22) and Apphia (Feb. 15), and, like them, was a disciple of the Apostle Paul, who calls him "our fellow soldier" (Philm. 2). He and his father preached the Gospel at Colossae, and Archippus probably served as a priest for the church that gathered there at his family's house (Col. 4:17). Archippus' fervor in preaching the Gospel of Christ so angered the pagans that they seized him and brought him before the governor Androcles. When the Saint refused to sacrifice to Artemis, he was stripped, beaten, tormented in various ways, and finally stoned to death.

Archippus was the son of Saints Philemon (Nov. 22) and Apphia (Feb. 15), and, like them, was a disciple of the Apostle Paul, who calls him "our fellow soldier" (Philm. 2). He and his father preached the Gospel at Colossae, and Archippus probably served as a priest for the church that gathered there at his family's house (Col. 4:17). Archippus' fervor in preaching the Gospel of Christ so angered the pagans that they seized him and brought him before the governor Androcles. When the Saint refused to sacrifice to Artemis, he was stripped, beaten, tormented in various ways, and finally stoned to death.

We will be judged on what we know and how we live our lives according to that knowledge.

Intel was once unrivalled as the world's chip maker. Now it's fallen behind its rivals.

One of the least successful Pro-Am teams in history reflects on where it all went wrong.

A Solheim Cup preview and reviews of Eddie’s time in Switzerland and Texas at the O2.