Intel Details ‘Biggest Changes of a Generation’ at Architecture Day


Intel Architecture Day 2021

Picture: Intelligence

This week Intel hosted its annual Architecture Day and even for an event that often features major changes coming for the company’s silicon, this year Intel is hailing its new advancements as some of the “the biggest changes in a generation ” across its CPUs, GPUs, and more.

When it comes to consumer devices, the big new product lines to watch out for are Alder Lake, Intel’s next-gen CPU architecture, and Alchemist, which is the codename for the next big graphics card. Intel’s audience and the first product of the new Intel brand. High performance arc line gaming products (HPG).

However, before we talk about new processors, we need to talk about the design of Intel’s new processors themselves, as the big changes to Intel’s core architecture (no pun intended) are what really open up the provides a large part of Intel’s performance and energy efficiency gains. .

Image of article titled Intel Details 'Biggest Changes in a Generation' at Architecture Day 2021

Picture: Intelligence

Instead of just packing more cores on a smaller die, for the first time, Intel is shifting to a new hybrid architecture of faster performance cores augmented by power-hungry efficiency cores, very similar to what many ARM chipmakers have been doing for years on mobile devices.

Previously named Golden Cove, Intel says its new Performance-core microarchitecture includes more decoders (4-6), more execution ports (10-12), larger registry files, and improved branch prediction. . All in all, compared to current 11th Gen Cypress Cove processors running at the same frequency, Intel says its new Golden Cove performance core offering about 19% better overall performance. In addition, thanks to the addition of new advanced matrix expansions from Intel, these cores are also receiving a big boost in the treatment of AI.

And as you would expect from a hot new processor architecture, Alder Lake is also designed to better support new component standards including DDR5 RAM, PCIe Gen 5, Thunderbolt 4, Wi-Fi 6th. , etc.

Intel Architecture Slides of the Day

Picture: Intelligence

Meanwhile, completing its performance cores, Intel’s new Efficient-core microarchitecture (previously named Gracemont) would support enhanced multitasking through enhanced multithreaded performance with what Intel claims to be the most. efficient x86 microarchitecture at the moment. In short, Intel says its new efficiency cores deliver 40% more single-threaded performance at the same power levels. or book similar performance while using 40% less energy. And compared to Intel’s Skylake chips for overall throughput, four of Intel’s new efficiency cores would provide 80% more performance compared to two Skylake cores running four threads.

Additionally, to better manage the workloads between its performance and efficiency cores, Intel introduces its new Thread Director Scheduler, which provides Support for dynamic assignment of processes to the appropriate kernel for the job.

As for Alder Lake itself, new chips will be built on the Intel 7 process, with three different main designs intended to cover desktops, high-performance laptops, and a range of high-performance laptops. devices. The new Alder Lake chips will be available with varying number of cores and threads with the most powerful processors reaching 16 cores, 24 threads and 30MB of LL cache.

Intel Arc Roadmap

Here’s a roadmap showing Intel’s future consumer GPUs, starting with Alchemist which is expected to arrive early next year.
Picture: Intelligence

Regarding the alchemist, Intel hasn’t delved so deeply into the architecture behind its next GPU but he revealed a number of important features and designs points. One of the biggest revelations is that instead of making the Alchemist chip itself, Intel is relying on TSMC’s N6 node to produce the silicon for its next graphics card, signaling a big change for a company that previously relied heavily on its own in-house manufacturing.

Additionally, instead of focusing on the number of threads, Alchemist’s architecture is based on what Intel calls XE cores, which will include 16 vector engines and 16 matrix engines, with units of additional ray tracing that supports both DirectX Raytracing and Vulkan. Laser trace.

Intel XE cores

Picture: Intelligence

TWith updated graphics driver components and memory managers / compilers, Intel says Alchemist can deliver 15% throughput (or up to 80% better in some cases) for CPU-bound games , while improving loading times by up to 25%.

Alchemist already supports Unreal Engine 5 and DirectX 12 Ultimate and Intel also showed disabled its own super-sampling technology called XE SS, designed to improve graphics quality or graphics performance up to 2 times under ideal conditions. Intel even says the SDK for XE SS will be available to developers starting this month.

XE HPG overview

Picture: Intelligence

Beyond traditional consumer computing products, Intel is also making big waves with the introduction of its new IPU (Infrastructure Processing Unit) and Mount Evans being Intel’s first UIP ASIC. And for enterprise graphics, Intel also detailed its new Ponte Vecchio GPU, which Intel says delivers record AI performance and up to 45 TFLOPS of FP32 throughput, with the chip based on the N5 process the most advanced from TSMC.

Wwith Intel processors now on the back foot compared to AMD’s current chips in terms of performance and value (assuming you can find them in stock) the company needs Alder Lake and Alchemist will be a hit if it happens to recover the superiority of silicon.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *