Eighty Cores Is Enough?

By Geoff Duncan February 12, 2007

Eighty Cores Is Enough?

If Intel‘s Core Duo technology quickened your pulse and the company’s quad-core offerings positively got your blood rolling…then the company’s Tera-scale Computing Research Program has developed a chip which is going to send you to the doctor for some industrial-strength blood pressure medication.

Intel has announced its first silicon tera-scale research prototype, which the company claims is the first programable chip to deliver more than one trillion floating point operations per second (that’s one teraflop) while consuming less power than many of today’s home appliances. The research chip implements 80 simple processor cores, each of which contain two programmable floating point engines, on a design no larger than a typical fingernail. Each core also contains a 5-port message router connected in a2D network the implements inter-core message passing. Scaled up, this sort of core intermesh network could be staled up to even larger multicore setups, enabling even better processor performance.

Although the chip is a prototype and doesn’t represent a product the company plans to bring to market, it could be a glimpse of technologies lying five to ten years down the road of processor development. Running at 3.16 GHz, the chip managed 1.01 teraflops and consumed just 62 Watts of power, which is comparable with processors used in today’s desktop systems; cranked to 5.7 GHz, the chip managed 1.81 teraflops and consumed 265 Watts. The processor also implements power management features which enable individual cores to be put to sleep based on the performance requirements of a particular application.

Editors' Recommendations

Former Digital Trends Contributor

Geoff Duncan writes, programs, edits, plays music, and delights in making software misbehave. He's probably the only member…

Computing

Intel’s new Core Ultra chips needed to be more than this

The Acer Swift Go 14 on a table in front of a window.

A lot has been made of Intel's new Meteor Lake chips. They have a new brand, naming scheme, and even new technology in the form of the NPU (neural processing unit).

And they come at an important time. The stakes have never been higher for Intel, as it faces competition from all sides. They're also an important litmus test for the idea of the NPU, or neural processing unit.

Computing

Confused about Core Ultra? We were too, so we asked Intel about it

Intel's new Intel Core Ultra badge.

It’s the start of something new for Intel: the Core Ultra era. With the launch of new Meteor Lake processors, Intel is ditching its old naming scheme for something new. Intel CPU names are going to start looking a little different, and you might be confused by what you see on a spec sheet. We’re here to get you up to speed.

At first glance, Intel is just dropping the “i” that has defined its lineup of CPUs for 15 years. There are a few new details in the naming scheme, though, and knowing them can help you navigate this new era for Intel.
The basics of Core Ultra

Computing

With the arrival of Core Ultra, a new era for Intel has begun

A tray of Intel Core Ultra CPUs.

Intel's Meteor Lake processors represent more than just a new generation. It's a fresh start for Intel -- a cycle it seems to find itself in every few years -- and the birth of the AI PC. Most importantly, it's a sign that Intel can deliver on its road map with a fresh generation of processors built on a new node.

Meteor Lake ushers in the Intel 4 node, which we first heard about nearly three years ago with the launch of 11th-gen CPUs. It also marka the start of the Core Ultra era of CPUs, along with a consolidation of Intel's massive mobile CPU lineup. Now, processors fit into two camps: U-series for thin and light laptops, and H-series for more powerful machines.