Comments from the press Q & A
Pat Gelsinger, CEO on Increased Manufacturing Capacity:
“We are building all expansion projects on schedule or ahead of schedule, but it is very important for the US House of Representatives and Senate to complete the CHIPS Act. This hassle because it is necessary to accelerate the industry. Do something. “”
Hyperscaler Sandra Rivera, CVP and GM, Data Center and AI to build this unique custom silicon:
“Hyperscalers are part of our high-capacity drivers and we expect to provide our data centers with sustainable and durable features. For large-scale deployments, we still rely on Xeon. doing.”
“But they also want lasting innovation. Xeon can offer some of it, and the foundry business also has the opportunity to allow hyperscalers to innovate with their own IP.”
Gelsinger, about the importance of software:
“Currently, there are more software engineers working than VMware. The role of software has become dramatically more important and is unlikely to change.”
“We are making more SaaS and SaaS acquisitions to differentiate Silicon through SaaS services. Silicon and SaaS are comparable to solutions and we see them running more. prize.”
Patrick Gelsinger is back on stage to conclude the keynote.
“There are still many things that are impossible, but we are setting a blazing pace to create world-changing technologies that improve the lives of everyone on the planet,” he said. say.
“Today’s Renaissance sets the stage for tomorrow.”
Gaudi 2 Accelerator Demo
The COO of Habana Labs is on stage to provide a demonstration of the benefits of deep learning training performance for the new Gaudi 2 accelerator mentioned earlier in the blog.
Compared to Nvidia’s A100 GPU, Gaudi 2 offers twice the training throughput on both the most popular vision and NLP models, the ResNet-50 and BERT.
“The higher the throughput, the faster the training and the faster the customer can reach their goals,” said Eitan Medina, COO of Habana.
Koduri invites Rick Stevens, Associate Director of the Argonne National Laboratory, home of the Aurora supercomputer currently under development, to the stage.
Aurora consists of a number of computing blades, each with a 2x Sapphire Rapids CPU and a 6x Intel Ponte Vecchio GPU, each connected to thousands of other blades. Once in operation, supercomputers will be one of the most powerful in the world.
publication: Argonne invites scholars to submit a time request with Aurora while Aurora is still under development.
Intel Arctic Sound-M GPU
Koduri owns Arctic Sound-M, Intel’s new data center GPU, announced earlier this year in Aloft.
The Artic Sound-M GPU features up to 4xXe media engine and up to 32xXe core and ray tracing unit, and seems to be able to support up to 40+ HD game streams and 30+ HD video streams simultaneously. “This is a media supercomputer on a single chip,” says Koduri.
Arctic Sound-M will be available to customers next quarter.
Raja Koduri, Head of Accelerated Computing and Graphics (AXG), will be on stage to talk about the next steps in Intel’s high performance computing space. This is an Intel-dominated area, but AMD has taken hold in recent years.
Announcement: Intel’s Habana Labs New AI Accelerator
Intel has opened the lid of a second-generation Gaudi accelerator that can significantly reduce the time it takes to train large AI models.
Built on a 7nm process, the Gaudi 2 processor features 24 integrated 100GbE RoCE ports and boasts the largest amount of memory on the market for accelerators (96GB HBM2e).
The new processor is a product of Israel-based Habana Labs, which was absorbed by Intel in 2019 and is designed for servers dedicated to deep learning workloads.
Read our full article here.
Rivera holds Xeon Sapphire Rapids wafers in the air. It’s very shiny.
The company has already offered 4th generation Xeon SKUs to its customers, but there is still no news about when wide availability can be expected.
Announcement: IPU Roadmap
Sandra Rivera, Head of Data Centers and AI, will be on stage to keep you up to date on Intel’s data center business line.
Rivera has announced that it will extend its Infrastructure Processing Unit (IPU) roadmap to 2026.
The company is said to offer three new generations of Infrastructure Processing Units (IPUs) within the next four years. The first new products, Oak Springs Canyon FPGA and Mount Evans ASIC, will be available by the end of the year.
We’ll talk more about this later.
Announcement: 12th Generation Intel Core Mobile Workstation Processor
Some news! Intel has announced the launch of its final product in the family, the 12th Generation Intel Core-HX Mobile Workstation Processor.
The new chip is said to offer up to 16 cores, up to 5Ghz clock speeds, and “unmatched performance”.
It’s all about the PC
Pat Gelsinger goes off stage to the right and talks about a humble PC on behalf of Michelle Johnston-Holthouse (GM of EVP and client computing).
“The PC is the human touchpoint we rely on to focus, create, connect and drive our business. I really believe it’s here,” she says. increase.
“If you had to work from your phone, it’s a good idea to imagine how it’s been in the last two years. PCs enable the workforce of the future.”
Johnston-Holthaus adds that there are 140 million commercial PCs over four years old, which are blocking business users.
IDM 2.0 and supply chain
Gelsinger nods to the company’s IDM 2.0 strategy. Its purpose is to put Intel at the forefront of chip manufacturing during times of unprecedented demand.
“We will build state-of-the-art capabilities in the United States to support the demands of these four superpowers,” he pledged.
“We are just beginning to offer geographically balanced supply chain capacity.”
The company has already invested billions of dollars in new manufacturing facilities in the US and Europe this year, but the implicit suggestion here is that more will come.
“The world is at a strategic turning point, a moment when things go incredibly well or incredibly unsuccessfully,” Gelsinger adds.
“Transformation is inevitable and applies to everything. Every business is becoming a technology business.”
According to Gelsinger, there are four superpowers that drive innovation. It’s pervasive connectivity, ubiquitous computing, artificial intelligence, and cloud-to-edge infrastructure.
“Each of these superpowers is impressive in itself, but when they come together, it’s magic.”
“If we don’t apply AI to every business process, we’re lagging behind. This is seen in all industries.”
Intel’s CEO will be on stage to welcome the audience to the company’s first major face-to-face event since the pandemic began (and returned to Intel).
“Today, the pace of technology is the fastest in your life, but the slowest in the rest of your life,” he says. “We are sitting on the cliffs of the Digital Renaissance.”
We are waiting for Intel CEO Pat Gelsinger to take the stage for the keynote. He should appear after about 20 minutes.
Stay tuned for updates on new products and services, and perhaps the progress of the company’s IDM 2.0 manufacturing strategy announced last year.
Good morning from Dallas
Intel Vision starts today and TechRadar Pro is here with all the latest announcements.
The first is a keynote by CEO Pat Gelsinger, followed by a more rigorously focused session between client computing and the data center and AI business groups.
The day ends with a round table meeting between Raja Koduri and Jeff McVeigh. In this session, we need to hear more about the company’s efforts in the graphics space, both from a device and data center perspective.
What to expect
Intel keeps the card close to its chest, but it’s clear that it will be contacted by spokespersons for its client computing, accelerated computing, and data center business groups.
Basically, there should be something exciting for everyone with product and service updates, from the processors that power the device to the chips that power the cloud.
In the meantime, here are some articles about the latest news from Intel.