At first glance, Intel doesn’t look like a besieged company. Last year’s sales were $ 77.8 billion and profits were $ 20 billion. The market capitalization in mid-September 2021 is $ 220 billion.
Still so. When you are a leader, all your competition is aimed at you. Intel is working to lose leadership on manufacturing and process nodes, losing market share to the highly resurrected AMD, relentless Nvidia challenges Intel for AI advantage, Atom processor Arm in the mobile market CEO three years later, who failed brilliantly against and became the third.
But Intel is winning the competition. Lisa Spellman, Corporate Vice President and General Manager of Xeon and Memory Group, said: “So it’s no wonder there are competitors who want some of it.”
“I’m not saying Intel is’siege’, but it’s certainly being attacked from various angles, including internal turmoil. Glenn O’Donnell, Vice President and Research Director of Forrester Research, said:
Other factors are difficult to fix.
With AMD gaining momentum, Nvidia GPUs collapsing, Arm architecture now a great alternative, and countless other chips such as communication interfaces, microcontrollers, and DSPs all on the battlefield, competition is unprecedented. It is no exaggeration to say that it is getting fierce.
Intel got the shots that were terribly needed with Gelsinger’s return to CEO in February. Gelsinger has long been admired as one of the most important engineers in the company’s history and was considered a leading candidate for the top spot when exiled in 2009.
After returning to Japan, Gelsinger began to run hard and made a series of moves to repair Intel’s lost luster.
“I think he’s doing the right thing. And you know, he’s a new guy. He has a license to make bold moves and they were bold. IDC Computing And Shane Lau, Research Vice President of Semiconductors, said:
Some of Gelsinger’s advocacy projects actually started under his predecessor, Bob Swan. Bob Swan couldn’t get the credit he deserved, given that he was the former CFO of the company and was rich rather than an engineer. “Bob has survived many difficult times,” Spellman said.
Gelsinger has extensive experience since leaving Intel in 2009 and is now helping him by modifying the company’s policies. He was the COO of EMC, taking over the CEO of VMware, trebling revenue and expanding far beyond core hypervisor technology.That experience Ready for him According to O’Donnell, for a newer job than if he stayed at Intel. “It gave him not only a way to lead a major tech company, but also a non-Intel perspective,” O’Donnell said. This will help the gelsinger figure out what he needs to do. “So I think it was very valuable.”
IDM 2.0 shakes manufacturing
In 2008, AMD split into two companies, one developing chips and the other called GlobalFoundries, which manufactured them. Struggling to compete with Intel, the company couldn’t afford to maintain manufacturing facilities in New York and Dresden, Germany, and sold the fabs to two investment funds owned by the Abu Dhabi government. The concept of a chip company called “fabless”, which outsources manufacturing to a third party just by designing the chip, became popular, and Nvidia proved that the concept worked.
Wall Street and other analysts have begun whispering that Intel should probably do the same, as Intel was behind production. Instead, Intel has launched a strategy called Integrated Device Management 2.0, or IDM 2.0. First, IDM 2.0 includes a $ 20 billion investment in two new chip fabs at its Chandler, Arizona facility. Besides that, IDM consists of three components:
- Intel’s global in-house factory network. Intel has fabless in the US, Europe, Central America, and Asia, and IDM 2.0 promises to create its own chips to avoid fabless.
- Expanded use of third-party foundry capacity. Intel plans to manufacture some chips in collaboration with other foundries, including Taiwan’s giant TSMC.
- Build a world-class foundry business called Intel Foundry Services. Intel will manufacture chips for other companies, including competing companies. This is new to Intel and already has two customers side by side. Both are competitors. Amazon Web Services and Qualcomm. AWS manufactures its own Arm-based server processor called Graviton, but Qualcomm is competing for 5G chipspace.
Spellman said Intel had previously manufactured for other parties. IDM 2.0 makes it more strategic to discuss products, solutions, and opportunities to build solutions. “
IDM 2.0 also includes additional technology elements such as leadership software and foundry business, Rau said. He feels that these will probably be the most difficult components of the IDM 2.0 strategy.
“There is a craving for leadership processes and manufacturing throughout the industry,” he said. “The connection must be made between Intel’s foundry company and its customers, not Intel itself, which means it brings different skills, different IP portfolios, different designs, and therefore foundries. Can enable customer success. “
Spelman states that IDM is more than just a business alliance between chip makers, it also benefits end users. “Customers who provide infrastructure as a service may be able to drive their own IP, their own customizations, or simply different data flows that are best suited for their environment,” she says. This potentially means lower latency, increase the density of compute that can be delivered, and increase the sustainability of compute. All of these are beneficial to the end user.
Until the last decade, Intel has been the only x86 show when it comes to computing. I didn’t use x86 for communications, network chips, etc., but I did use it elsewhere ( Itanium processor of fate Intolerable). The Atom was an attempt at a mobile processor, essentially stripping off the x86 core. The first two attempts at GPU, Larrabee and Xeon Phi, were x86 based. They all failed.
Then the flow changed. In 2014, Intel acquired FPGA vendor Altera, AI chip makers Nervana and Habana, and networking expert Barefoot Networks, and invited top AMD GPU experts to finally build the right GPU.
The multi-architecture approach is not unique to Intel. AMD is acquiring FPGA maker Xilinx. Nvidia has acquired networking specialist Mellanox and is about to acquire Arm Holdings. Marvell has also moved from manufacturing hard drive controllers to Arm server processors and smart networking adapters. No one is offering a single architecture anymore.
“It’s a religious debate for a company like Intel, but as I say, Gelsinger needs to gladly eat his youth, so to speak, to take Intel in another direction. It doesn’t mean moving away from x86, but you should consider x86plus, “says O’Donnell.
Recently, Intel has added a new RISC architecture to its family. In June, we plan to sign a contract with RISC-V maker SiFive to create a new development platform called Horse Creek with SiFive’s new Performance P550 core. Intel has also registered SiFive with Foundry Service. That same month, Bloomberg reported that Intel tried to buy SiFive for $ 2 billion. However, Intel remained silent on the matter and refused to comment except to confirm its partnership with SiFive.
According to O’Donnell, the mobile computing ship has set sail, but Intel still has great opportunities in edge computing with RISC-based chips. “One of the biggest criticisms of x86 is power consumption,” he said. “But IoT devices and edge computing devices will be huge. It’s a huge growth market. And power consumption is important there. And now it’s everyone’s market to win.”
Introduction of IPU
Intel has also introduced a new line of smart networking controllers called Infrastructure Processing Units (IPUs) in addition to the SmartNICs it already offers. The idea behind SmartNIC is to offload the task of routing network traffic from the CPU and free the CPU to run jobs that process data. Unlike traditional network controllers, SmartNICs have some form of programming potential to perform tasks that traditional dam NICs cannot handle, such as packet processing...
IPU is a Smart NIC that takes you to the next step. In the past, SmartNICs were on-premises, but IPUs are specifically designed for data movement in the cloud and communications services. The IPU is designed to separate the cloud infrastructure from tenant or guest software, allowing guests to have full control of the CPU with the software and service providers to maintain control of the infrastructure and trust routes.
The two IPUs have Intel Xeon-D and Agilex FPGA cores for processing. The third codename, MountEvans, is the first type of Intel. Designed in collaboration with top cloud service partners, Mount Evans is based on Intel’s packet processing engine and features up to 16 Arm Neoverse N1 cores with a dedicated compute cache and up to 3 memory channels. The ASIC can support up to four hosts, Xeon, with a full dual bandwidth of 200 Gb / s between them.
XPU and oneAPI connect them all
Each processor architecture has its strengths and weaknesses, all suitable for a particular use case. Announced last year, Intel’s XPU project aims to provide a programming model that is integrated into all types of processor architectures and to fit all applications to the optimal architecture. XPU allows you to mix x86, FPGAs, AI, machine language processors, and GPUs all into your network, which means your app will be compiled to the best processor for your job.
This is done through a oneAPI project that works with XPU. XPU is the silicon part, and oneAPI is the software that connects it all together. oneAPI is a heterogeneous programming model that uses common languages such as C, C ++, Fortran, and Python, and code written in standards such as MPI and OpenMP.
The oneAPI Base Toolkit includes compilers, performance libraries, analysis and debugging tools for general purpose computing, HPC, and AI. It also provides compatibility tools to help migrate code written in Nvidia’s CUDA to Intel’s GPU language, Data Parallel C ++ (DPC ++).
“This is a branded version of the trend that is well established across semiconductor manufacturers,” says Rau. “That is, you can no longer be CPU-centric or GPU-centric. You can’t have one major chip in your portfolio. And to be successful, you can connect multiple types of processors. Requires processing and interconnection. “
Spelman has confirmed that the jury is still under development, so oneAPI is still under consideration.
And she says never say about the possibility of adding architecture. “We never say it ended there. The industry is moving too fast and too dynamic, so we don’t know where to invest next,” she said.
Copyright © 2021 IDG Communications, Inc.
Intel: Attacked and counterattacked in many ways
Source link Intel: Attacked and counterattacked in many ways