X99 is an odd platform:
1. It doesn't do high end computing well, because it doesn't support ECC and isn't certified for 24/7 operation;
2. It doesn't provide any noticeable gaming performance, nor any rendering performance that is worth the price premium (check TechofTomorrows benchmarks) in Windows;
3. It doesn't benefit from the OpenCL acceleration and Share Virtual Memory support of the next gen Broadwell CPU's;
4. It uses a lot of power for an Intel chip with those specs.
In my opinion, it's a serious sidestep, not a step forward, not a step backward, but a hyped status quo with a premium price tag.
It's clearly targeted at Windows gamers that use multiple nVidia-GPU SLI setups. So it's a luxury product, no real benefit, but something to show off with.
There is a market for that, but not such a big market I think.
When Broadwell comes out with the new iGPU's that have 48 streaming processors on board, that will be next desktop-CPU by Intel worth buying in my opinion, and that platform, once it comes to 2011-3, will also provide a system that will last a few years.
Haswell-E is just what Haswell should have been from the start, it's nothing special in 2014 any more, and certainly not worth the money. Just like AMD, Intel has made adjustments to their range, but nothing new has been brought out. Difference is that AMD has dropped the prices, and Intel has upped them.
For linux users, X99 isn't beneficial yet either. Difference with software console users is that X99, once Intel solves the problems with X99 support, will become pretty good in linux, whereas that will probably never happen for software console users. Usually, that takes a year or more nowadays, because Intel definitely has issues keeping up with their development.
The next AMD range will probably have hybrid CPU's, that also contain high performance computing ARM cores, that use much less power and have much higher specific computing performance. The hybrid computing model that AMD uses, has also been tried and tested, and has proven itself over the last 4 years or so, so it's a pretty safe investment for next-gen computing needs. Intel hasn't proven anything in terms of next-gen computing yet, they've only spent a lot of money on marketing. With Broadwell, Intel is going after AMD's Kaveri line, which is quite old at this point, and the Intel top model Broadwell will probably just about equal the compute performance in hybrid computing of the present top model Kaveri at stock clock (but they overclock so nice, nobody uses them at stock speed anyway...). That means that Intel is jumping through hoops and doing spins on it's head to compete with a 150 USD APU... they're not doing the same kind of thing to develop a faster high end desktop CPU... enough said.
Chances are very real that with linux and HSA, a 150 USD chip will outperform any X99 system in the next 12 to 18 months in most applications, thereby using only half the power. Two years ago, when HSA was still being tested, speed increases averaged between 60 and 80% for a quad core setup. That's a lot, especially knowing that speed increases will be higher with more CPU cores and GPUs with FPU units. Compare that to the Intel advertised performance increase of X99 of maximum 20% (of which almost none is really noticeable in Windows gaming and accelerated video rendering), and it's clear that X99 is not really the deal of the century...
I would buy nothing if you can avoid it right now, and wait for the next generation HSA systems.
