INTEL HAS FINALLY done what we had all expected it to do, and pulled the plug on the consumer version of Larrabee. In a statement today, Intel said that the chip will be a development platform and an HPC part, but there will be no retail version, at least not any time soon.
This change of direction was pretty well assured because the first Larrabee chip was so late. If you are a year late in the GPU business, that is an unrecoverable deficit. While some may say that this is a failure, they probably don’t understand the magnitude of the task that Intel undertook. It is not simply a GPU, it is the next generation of vector compute parts.
Looking into the crystal ball, we see that Larrabee 1 and 2 were very similar, and Larrabee 3 was a very different part. Given that, we would expect Larrabee 2 to be dropped as well, and effort to focus on Larrabee 3 as a GPU.
The more interesting thing is the release of an HPC SKU, so there will be Larrabees available for that task. Given that it was always going to be a better HPC part than a GPU, this is a good step to get the parts out into the field.
At SemiAccurate, we have been waiting to see Larrabee for a long time, the author was the first to break the code name more than three years ago, and the first to describe the architecture almost three years ago. It was a good idea then, and it is a good idea now.
Given the likely downplaying of two generations of parts, I guess we will have to wait a while longer. The end result is still the right thing for Intel, and the future of computing, it will just take a bit longer to get there.S|A
Latest posts by Charlie Demerjian (see all)
- More on Intel’s 10nm process problems - Sep 17, 2018
- Intel puts out another 14nm 2020 server platform - Sep 11, 2018
- Why Can’t Intel Supply Enough 14nm Xeons? - Sep 10, 2018
- Intel can’t supply 14nm Xeons, HPE directly recommends AMD Epyc - Sep 7, 2018
- AMD reintroduces the Athlon name with two CPUs - Sep 6, 2018