Initially announced by Intel executives at a fair held last September, the chip includes 80 cores in 275 square millimeters (practically the size of a fingernail), while it requires a consumption of 62 watts, which means a lower consumption than required by the processor used in a desktop PC.
The manufacturer does not plan to launch this “teraflop chip” on the market, but will use it to test new technologies such as broadband interconnections, power management techniques, and a design method for building multi-core chips. according to Jerry Bautista, director of Intel’s tera-scale research program.
Intel engineers are also using the chip to explore new forms of teraflop-scale computing, in which future users could process terabytes of data on their desktops to perform real-time speech recognition operations, multimedia data mining, photorealistic games and artificial intelligence.
Until now, this level of computer performance was only available to scientists and schools using machines such as the ASCI Red, the teraflop computer built by Intel and its partners in 1996 for the United States. This system is capable of performing a quantity of calculation comparable to that of the new chip, although it requires a consumption of 500 kilowatts and 500 kilowatts of cooling for its nearly 10,000 Pentium Pro.
Running at a speed of 3.16 GHz, the new processor achieves a computing performance of 1.01 teraflops, which implies an efficiency of 16 gigaflops per watt. It can even run at a higher speed, although in that case it would lose efficiency by offering performance of 1.63 teraflops at 5.1 GHz and 1.81 teraflops at 5.7 GHz.
