This cant be serious. Good guy Intel saving the global warming is your argument for not shipping a stock cooler?
Skylake K buyers can use the stock solution of several past generations from Intel, even 775 with some teaking. The subjects revolves minimizing BoM and maximizing profit, not the ECO BS you are trying to say.
You're probably right, unfortunately. That was one of the great things about the G3258 budget overclocker. It came with a copper-cored stock heatsink, that would generally take the chip as far as a cheap H81 mobo could push it.
Now that's just an added expense for the i3-7350K. So much for being an ideal budget OC CPU.
$180 for a dual core in 2017 ??? no thanks, better go for the i5.
This is getting ridiculous now, instead of progressing with higher core count at the same or lower prices, we have a Dual Core at $180 in 2017 that will be decimated in every MT benchmark by an OCed 6 year old Core i5 2500K.
Really i mean if you bought a Core i5 2500K in January 2011 at $220 and got it OCed to 4.5GHz and 6 years later you want to upgrade to a new CPU spending $220 again what do you get ???
$180 for a dual core in 2017 ??? no thanks, better go for the i5.
This is getting ridiculous now, instead of progressing with higher core count at the same or lower prices, we have a Dual Core at $180 in 2017 that will be decimated in every MT benchmark by an OCed 6 year old Core i5 2500K.
Really i mean if you bought a Core i5 2500K in January 2011 at $220 and got it OCed to 4.5GHz and 6 years later you want to upgrade to a new CPU spending $220 again what do you get ???
Hell, according to the leaked price table at $230 your choice is a locked Quad Core (again) with a turbo of 4.1GHz that will not even be 10-20% faster than your 6 year old OCed CPU, really pathetic.
No wonder Desktop sales are going to hell.
Despite your efforts to downplay this product, a lowly i3-6100 still holds its own compared to competing APUs, 130-220W FX and older Core i5s in some of the latest titles - so an OC capable 4.2 GHz 2C/4T Kaby Lake should be fine. Also stop saying it costs $180, those are pre-order prices that usually go down after launch.
Of course this product is not for everybody, but the added OC support for $10-20 over current Core i3-6320 isn't bad at all. And again, Pentium G4620 will deliver most of this performance at sub $100. Crossing fingers for a $80-85 Pentium G4600 with HT support.
Here's some alternatives if you want to avoid Core i5s, even though the latest Skylake (soon Kaby Lake) variants provide a nice boost over Sandy Bridge in many games:
Used or New IB/Haswell Core i7s
Used or New 4C/8T Xeon CPUs for $240 - lower when on sale
Used 6C/12T+ Xeon CPUs on the cheap, insane MT performance/$
Coffee Lake CPUs as soon as late 2017 / early 2018 (DigiTimes) - possibly increasing core/thread count at all segments
As Intel traditionally does, they will sample to big consumers earliers, like the cloud guys.Mid 2017? That sucks; I thought it was going to be earlier.
As Intel traditionally does, they will sample to big consumers earliers, like the cloud guys.
And it's about one year after BDW, right?
Assuming the i3-7350K has 2c/4t, turbo, and HT - that would make for a very interesting budget gaming rig. All of the games in my library except Ashes want high single-core perf, and won't see perf improvements beyond four hardware threads. This puts the i5 market in a squeeze.
Uh, yeah, sure, we already know that by now Intel and delay are synonyms, right? But Intel has never lied so outrageously about their process timeline as TSMC has done.Might want to update your sig that bashes other companies for not hitting timelines, Intel is just as guilty
Uh, yeah, sure, we already know that by now Intel and delay are synonyms, right? But Intel has never lied so outrageously about their process timeline as TSMC has done.
But the first roadmap placed SKL Xeon in Q4'17, so that hasn't been all that bad I would say. Xeon Phi slipped by a lot longer. Just like silicon photonics.
$180 for a dual core in 2017 ??? no thanks, better go for the i5.
This is getting ridiculous now, instead of progressing with higher core count at the same or lower prices, we have a Dual Core at $180 in 2017 that will be decimated in every MT benchmark by an OCed 6 year old Core i5 2500K.
Really i mean if you bought a Core i5 2500K in January 2011 at $220 and got it OCed to 4.5GHz and 6 years later you want to upgrade to a new CPU spending $220 again what do you get ???
Hell, according to the leaked price table at $230 your choice is a locked Quad Core (again) with a turbo of 4.1GHz that will not even be 10-20% faster than your 6 year old OCed CPU, really pathetic.
No wonder Desktop sales are going to hell.
And by Xeons, they mean Skylake-X/E right?
Skylake-EP. Skylake-X will be derived from Skylake-EP LCC die.And by Xeons, they mean Skylake-X/E right?
Absolutely false. http://core0.staticworld.net/images/article/2015/09/intel-kdm-roadmap-1-100615948-large.idge.jpg1H 2017
Now let me remind you how the situation looked from Intel's POV end of 2013. They were projecting they were 1 quarter behind in yield learning, so BDW was delayed by 1 quarter. But as we all know by now 14nm was more like 6-9 months later and in terms of yield parity maybe a few years. So a 1Q delay is no big deal and would certainly not affect 10nm cadence. That's just what common sense + Intel track record suggested. That's not lying. Intel has always been relatively open about the 14nm and 10nm progress, and they publicly announced that they ditched Tick-Tock.
Absolutely false. http://core0.staticworld.net/images/article/2015/09/intel-kdm-roadmap-1-100615948-large.idge.jpg
And mid-2017 is still H1'17.
Wait, did you mean to put in any new information in that post?14nm was a big mess, and people literally lost their jobs because of how much of a disaster it was.
And at least maybe we might imagine that 14nm just sucked so much that Intel has done a lot to ensure they don't repeat this anymore. It's like ASML who thought EUV would be breeze to develop and thus procrastinated the heck out of it. Although I hope lengthening the cycle isn't the only change they made . We'll see how the 10nm transition goes. Should be fun. Ticks always the most interesting.14nm was a big mess, and people literally lost their jobs because of how much of a disaster it was.
And I hope many more people literally got jobs because of how a disaster 14nm was. To fix it.
Those taking the buyout in Oregon, though, include Sanjay Natarajan, an Intel vice president and top engineer. He has been with the company since 1993 and led development of the company's newest, 14-nanometer manufacturing technology.
Otellini already left ship, so someone's got to take the blame I suppose..http://www.oregonlive.com/silicon-forest/index.ssf/2015/05/intel_trims_manufacturing_grou.html
They paid this so-called "rock star engineer" to take a hike.
Otellini already left ship, so someone's got to take the blame I suppose..
But how are these related to 14nm as they're called "voluntary"?
Skylake-EP. Skylake-X will be derived from Skylake-EP LCC die.