I heard about that, I want to see it compared to this chip and see a new gen comparison when they come out. I like the idea of power saving honestly, but I was hoping for more of the under load variant, not idle. I leave my machine off when I'm not gaming, and when I am gaming it's ring pushed, that's where I think the focus should lie. But yea, on the mobile platform, I have a feeling this will shine there.Unfortunately man. AMD is now rumored to be releasing a 5GHz CPU. They have to focus on architecture, not clock speeds. That is their issue. They need to snatch more Intel guys. That wont be hard with the money they just got from Xbox and Sony console deals.
A couple of points:
1. AMD is always rumoured to releasing something.
2. A 5GHz Piledriver based chip made in limited quantities for the LN2 crowd wouldn't impact the market at all. AMD already has 8+ GHz chips in the record books, and it hasn't exactly improved AMD's outlook.
3. The mythological FX-9000 (there's a hint in that name) is supposedly a 220 Watt part. Just for the sake of putting the oddball rumour to rest, there are plenty on analysis on the net to why this isn't feasible on the 32nm process using in-place hardware.
Well, that seems at odds with launching 5GHz / 220W monster golden sample doesn't it ?
Hahahahahahahahahaha.................................sorry.
Intel will spend over sixty percent more on R&D in this quarter than AMD's total financial cap is valued at
($4.7 billion in R&D versus AMD's market cap of $2.86 billion)
Well, no they aren't, and no it isn't.
Intel clarified that HEDT remains a socketed solution for the foreseeable future - hardly surprising since the CPUs share commonality with workstation and server chipsets. As for mainstream, Skylake (2015-16) is still LGA (Flip Chip-Land Grid Array) which means a socketed CPU
You're probably thinking of the hoo-hah regarding Intel's supposed shift to embedded (BGA) processors. Intel pretty much cleared up the issue saying that certain SKU's would be embedded only. A prime example would be the 4570R and 4670R because of the embedded DRAM on package.
FWIW, you'll find that there is a distinct lack of socketed processors made that feature eDRAM- mostly because the complexity of pin-outs and criticality of contact required between pins and mainboard.
Steve Can you guys test the performance of 1600MHz RAM on Haswell vs IB. Intel said they "redesigned" the 1600MHz memory controller.
But they redesigned the 1600MHz controller, not the 1866MHz lol.All configurations were tested with the memory clocked at 1866MHz if they supported it which Ivy Bridge and Haswell did.
There is a 1600MHz memory controller? That's news to me, can you link me to this info.But they redesigned the 1600MHz controller, not the 1866MHz lol.
Note - that's only under peak load. On average it is way better than Ivy Bridge
There is a 1600MHz memory controller? That's news to me, can you link me to this info.
Do you have any proof of that claim because from everything I have seen that is a load of rubbish.
What is funny is that it is in a TechSpot article: https://www.techspot.com/news/52398-intel-haswell-cpus-to-arrive-june-3.html. Lol.
Okay its not a 1600MHz memory controller, its just a memory controller that supports DDR3-1600 memory officially. That doesn't mean 1600MHz memory is faster than 1866MHz memory and it doesn't mean it cannot use 1866MHz memory.
It certainly doesn't mean you need to set both platforms at 1600MHz to look for a difference.
The whole point as I understood it was the main advantage to the new memory controller is that it could be overclocked to operate at high frequencies such as DDR3-2400. There are no performance advantages at 1600MHz and evidently at 1866MHz either.
Personally speaking we don't have the right to look down at anyone with capabilities we ourselves don't understand.AMD confirmed as a big letdown.
Come on AMD, step up. Intel needs to stop getting away with this ****.
Well, technically the architecture of both AMD and Intel have been on 2+ year cycles for a while. AMD's arch introduction is on a four year timetable of late ( K7 in 1999, K8 in 2003, K10 in 2007, Bulldozer in 2011), and Intel's if anything have been accelerating of late (Netburst in 2000, Core in 2006, Nehalem in 2008, Sandy Bridge in January 2011). Of course, CPUs like any technology become a case of diminishing returns- more so if you count process shrinks as architectures in their own right. Back in the early days of CPU evolution-as with any tech- the initial gains are impressive. Using the tried and true automobile analogy, cars were capable of 200+ mph in the mid-late 1980's. What has the next 25 years added in terms of absolute speed for production cars?Ugh, I find myself missing the days when CPU arcitechtures would launch at a frequency double that frequency over the course of the next 12-18 months, then a new aritechture could come out and do it again... It used to be that there would be no way in hell a chip "3 generations" back could ever hope to come close to whatever was current. It seems that sadly the days of leap frogging hardware are over... :/
Hasn't stopped a large percentage of forum posters is the past.Personally speaking we don't have the right to look down at anyone with capabilities we ourselves don't understand.
From a benchmarking point of view, Asus aren't the best indicators of a representative performance since their UEFI implements an "all core turbo" by default. The Intel specification is to enable max turbo on one core, and drop one multiplier for every successive core coming into turbo state. This is the primary reason that Asus motherboards generally top the performance charts (stock clocks) against their competitors - SATA third party controllers excepted.
The opportunity has always been there. But the ability... That's another story.AMD has the opportunity to kick Intels ***, let us see if they screw it up.
The opportunity has always been there. But the ability... That's another story.