Exhibition: Intel, a new powerhouse of high-speed computing

A train of new processors with characteristics each more amazing than the next. In Dallas, on the occasion of its Intel Vision 2022 show, finally in person, Intel returned to the great hours of the IDF, this annual event hosted by Pat Gelsinger in the early 2000s and which drew attention each time. computing towards new technological horizons. Ten years, fifteen years later, Pat Gelsinger is backhe is now the CEO and he has every intention of returning Intel to its driving role.

“Intel’s challenge now is to launch the great stuff we have in our drawers as quickly as possible. Before, we were limited in our production capacity, in our ability to provide what our customers expected. We had management problems, we weren’t supplying the best products in certain markets. To solve all this, we reorganized Intel into six divisions. Each in an IT sector, each with its own accounting and each with the objective of gaining as much market share as possible, against very specific competitors,” launched the CEO during the event.

“Honestly, how can we explain that we had nothing to offer for 10 years in the data center accelerator card market? This era is over, we are coming today with a very aggressive offer! “, he outbid during a press briefing.

The announcements are a processor for portable stations, the Core HX, which has 16 cores and operates at up to 5 GHz. A 52-core Xeon Sapphire Rapids, about 10% faster in tests than the 64-core AMD Epyc Milan. An Artic Sound M GPU designed to render server-side graphics and supposed to power all the immersive environments that the cloud will soon offer, including Facebook’s metaverses. A Gaudi 2 accelerator which accelerates artificial intelligence algorithms, with a score of 5425 in the ResNet-50 benchmark, while Nvidia’s A100, the ultimate in the field, only reaches 2930.

And also IPUs – the equivalent of DPUs at Intel – which should make it possible to generalize intelligent networks at 200 Gbit/s per Ethernet connector.

A new industrial strike force

As for the sixth activity, it is quite simply the factories, those that Intel has already modernized in Ireland and Arizona, like those which must come out of the ground by 2027 in Germany and Ohio.

“We learned more from the production lines in one year than we did in ten years. »

Pat GelsingerCEO, Intel

” We have raised the technical level of our factories, changing equipment suppliers, or investing in these suppliers to improve their products. But also by changing the way of directing this activity. We learned more from the production lines in one year than we did in ten years”, swears Pat Gelsinger.

“There you go, the more automated equipment we needed is there, we are rolling it out to our sites. Implicitly, he recalls that he had been let go to EMC during these famous ten years and that Intel begged him to come back when he had become the all-powerful boss of VMware.

“What’s left for us to do?” Change our industrial approach towards our customers. Our new strategy is tobring an ecosystem to where our customers are, in France, Italy, Poland, Spain. We will build the chips they need on site,” adds the CEO.

In this case, several European countries should see the emergence of an industrial structure on their territory. In France, it will be the R&D center where the next data center chips will be developed. Officially, this center will work on the design of the Xeons and GPUs that Intel will sell around the world and, at the same time, will welcome car manufacturers or any other industrialists to co-design the special components that they imagine putting at the heart of their future on-board equipment. Unofficially, hyperscalers – AWS, Azure, Google GCP – will come to this center to order custom builds of Xeons.

“Hyperscalers want to develop the technologies they use themselves. But their infrastructures are still processor-based. They are therefore counting on us to provide them with the eco-responsible and elastic processors that they will deploy in their datacenter and which will make the difference with the serial servers that companies can install in their own computer rooms”, argues Sandra Rivera, the boss of the Datacenter division at Intel.

“By starting from a base of standard Intel processors, they can benefit from lower prices thanks to an economy of scale. We will manufacture the same circuits, whatever the product or the customer, but we will assemble them differently. »

“Hyperscalers […] rely on us to provide them with the eco-responsible and elastic processors they will deploy in their data center. »

Sandra RiveraDatacenter Division Manager, Intel.

“We are going to innovate together on a circuit and this circuit will end up integrated, or not, in some of our processor models which are now assembled by chiplets. In fact, we have many more options to innovate than before,” she explains.

16 cores for laptops

The first day of the event gave pride of place to new processors for portable workstations, a real technological showcase in the race between Intel and its competitor AMD. These new Core HX processors will be available in seven models, including a high-end, the Core i9, which has 16 cores, i.e. eight efficient and eight economical.

The budget cores can only run one process each, compared to two on the high-performance cores, giving a total of 24 processes simultaneously. The performance cores run at 2.3 GHz with peaks at 5 GHz. The economical run at 1.7 GHz with peaks at 3.6 GHz. Power consumption ranges from 55 to 157 watts depending on core usage. The cache memory is 30 MB and the integrated GPU has 32 graphics units at 1.55 GHz.

The i7 and i5 models have a more or less reduced number of cores, frequencies and cache memory. At the entry level, the Core i5 has 8 cores (4 + 4), 12 MB of cache, 16 graphics units, slightly better base frequencies (2.4 and 1.8 GHz) and less substantial turbos (4 .4 and 3.1 GHz). Fuel consumption is the same on all models. Three models are available in a “vPro” version, that is to say which mainly integrates the additional security functions of professional PCs.

The use of this or that type of heart is not very clear. Intel seems to say that high performance cores are used by default for applications in the foreground and economic ones for those running in the background. A different approach to that of Apple – which invented the principle of high performance and economic cores on its M1 processors – where the economic cores are used by default and only hand over to the high performance cores when they are saturated with calculations, typically for graphics applications.

It remains to be seen what is the point of having so many cores on a laptop PC. “More and more applications are developed in multithread. But beyond the threads of execution in parallel which, admittedly, remain little used in the execution of a main algorithm, the enormous advantage of a large number of cores on a portable station is the multiplication of the inputs-outputs . It becomes realistic to use simulation applications that read or write in parallel to several NVMe SSDs,” explains Daniel Rogers, product manager of the processors division for client devices.

Information taken, a machine equipped with a Core HX can have four NVMe SSDs mounted in parallel on its PCIe buses.

These Core HX processors are the laptop version of the Core Alder-Lake for desktop stations introduced last fall. Like them, they support PCIe 5.0 buses twice as fast (especially on NVMe SSD access) and 128 GB of RAM per socket. Their threads also execute 17% more instructions in a given time than previous generation processors.

We want to thank the author of this article for this awesome material

Exhibition: Intel, a new powerhouse of high-speed computing

Take a look at our social media profiles and also other pages related to themhttps://yaroos.com/related-pages/