Based on a leaked Intel roadmap, you would be really surprised to know that Intel Broadwell won’t be launching in 2014. Instead it’s the Haswell (Haswell-E server part on X99 chipset, Ivy Bridge E, the six cores LGA2011 CPU on the chipset X79) which will be launching later this year with undefined performance characteristics.
The drawback of the story is that it completely assumes that Intel’s business is to focus on the desktop and that company will obviously set out its new process nodes and technologies before releasing them out elsewhere. Well that has been true for the past so many years but if we look at it from Intel’s point of view, investing money initially in desktops gives in lesser and smaller returns. Earlier in May when the company did the major Haswell meetings, the company talked frankly about 6-8year replacement cycle for desktops.
Coming down to the Intel’s Haswell latest architecture to date, it is directed for minimum power consumption, increased thermal density, improved IPC. It offers 8-10% IPC improvement over the Ivy Bridge which is actually quite praiseworthy. Sadly these new desktop chips have not been able to gather the attention of readers and people still think Haswell is more exciting for mobile which drives us to that fact that desktop performance alone isn’t exciting enough to attract buying.
Intel’s decision of prioritizing mobile and low power consumption over desktop performance isn’t what users really require. Having a probability of hardware failure within 3 to 4 years of owning laptop, Intel can rely on a replacement cycle that is quite reliable. Tablets and smart phones are most likely to follow the same path and can win more SKUs in the market which gives us a reason to push forward to Broadwell.