Intel has been working on its "Many Integrated Core" architecture for a quite some time, but the chipmaker has finally taken the code-name gloves off and announced that Knights Corner will be the first in a new family of Xeon processors — the Xeon Phi. These co-processors will debut later this year (Intel says "by the end of 2012"), and will come in the form of a 50-core PCIe card that includes at least 8GB of GDDR5 RAM. The card runs an independent Linux operating system that manages each x86 core, and Intel is hoping that giving developers a familiar architecture to program for will make the Xeon Phi a much more attractive platform than Nvidia's Tesla.
The Phi is part of Intel's High Performance Computing (HPC) program, where the company is readying itself to compete with Nvidia and ATI's stranglehold on cloud computing and analyzing "big data." Research firms and businesses alike use HPC to analyze large datasets to predict anything from weather patterns to stock market changes. Financial forecasting in particular needs strong double-precision floating point performance — something that the x86 architecture excels at.
Larrabee was Intel's first attempt at competing with the highly parallelized "general purpose GPU," but the plans for the chip were abandoned back in 2009. That doesn't mean the market hasn't been expanding — services like cloud gaming are opening a new consumer-facing side of the HPC market.
There are still plenty of questions to be answered, however. Intel hasn't announced a price for the Phi, and these new numbers show that Tesla's K20 card is capable of two teraflops of double-precision performance, possibly beating the Phi before its even been released. On the other hand, developers' familiarity with the x86 architecture may be enough in its own right to woo developers away from Nvidia's CUDA.