Alleged benchmarks of Intel i7-6700K Skylake CPU published

[HEXUS]

Shows an approximate 15 per cent improvement over Haswell-based predecessor.

Read more.

[DanceswithUnix]

The ideal processor for people who's job involves running PCMark every day

I think the FX8350 comes out looking pretty good there tbh, considering how cheap it is compared to the i7. I suspect there is more to it than that though.

[Corky34]

If they turn out to be true I'm going to be a little disappointed, I was hoping for a little more than what now seems the standard 10% improvement on previous gen CPUs.

Maybe I'm just longing for days gone by.

[TRUTHnTRINITY]

I was actually hoping for a little more increase but still better than what AMD have on the market. I'm hoping for good things from the Zen cores they should release close enough to Skylake that Intel coups see a hit if it's actually something good.

[marshalex]

Another processor series, another new socket on top...

Although AMD chips are generally cheaper and inferior to the intel counterparts at least they don't demand a new socket therefore new motherboard each time a new one comes out, especially seeing as the performance gains now are fairly minimal compared to what they used to be

[superp]

Why so happy? Its not a 10% improvement, it could be as little as a 0% or even 5% improvement. Comparing it to the 4790k is not fair. That processor is two generations old, you guys missed the Broadwell generation, the unreleased (desktop) architecture.

These comparisons are useless when you skip a generation, you can not say that it has improved 10% over the last big chip because it was not release. Skylake is compared to Broadwell, Not Haswell. If Haswell was on 22 nm process, and broadwell was on the 14nm, What was the improvement here. 0% maybe or 5% maybe or even the standard 10%. Once we see the Broadwell improvement we can then compare against skylake and only then. If we assume the 10% bump per generation we should be looking at a 20% increase over Haswell and that is worse case possible.

Two years of R&D funding, god knows how much in production and we get a maybe a hand picked 10% improvement. IMO I feel like this is a massive flop on Intels part.

[GuidoLS]

I'll believe the benchmarks once the chip(s) are either actually released, or they're done by a credible site with a little more recognition - say, Hexus... or even Anandtech or similar.

I'm also not seeing anything that indicates a new MB will be required. Still the same old socket 1151. I guess if you have DDR3 ram, and want to swap to DDR4, you'll need a new board, but that's pretty standard regardless of what or who's chip you use.

It should also be noted that the graphic at the top of this article is NOT from the site providing the benchmarks. The only graphics there ARE the benchmarks, and while the benchmarks are in English, the rest of the site is most assuredly not (Turkish).

A little more due diligence before we rage might be in order...

[Corky34]

I didn't think there were other CPUs that used LGA 1151, wasn't the last one 1150?

[Badbonji]

Yeah LGA1151 is a new socket that Skylake will use, you cannot use it in en existing LGA1150 motherboard. This socket apparently supports both DDR3 and DDR4 according to http://en.wikipedia.org/wiki/LGA_1151 .

[DanceswithUnix]

If they turn out to be true I'm going to be a little disappointed, I was hoping for a little more than what now seems the standard 10% improvement on previous gen CPUs.

Maybe I'm just longing for days gone by.

10% would be nice. That is showing 3% in GTA V, 6% in BF4 and a big fat zero in Crysis.

Sorry, the days gone by are gone. Back around the 486 days I would only upgrade when I could get double the processing power. I think that for me would now be around the 11700K

[CeeGee]

The 3DMark scores are cribbed from here http://www.futuremark.com/hardware/cpu
The Cinebench "R15" supposed scores are in the format used by Cinebench "R11.5", R15 uses a different scoring system.

I don't think the people the made these graphs had access to any of that hardware let alone an engineering sample of Skylake, I call bull****.

[GuidoLS]

I didn't think there were other CPUs that used LGA 1151, wasn't the last one 1150?

My bad - for some reason, not once, but twice, mis-read it as 1155.

But the lack of a list of benchmark hardware really set off some red flags for me, such as MB, memory, etc - so I went back to the linked page, and read it again. Several times. For anyone interested, here's the page, in all of its Google translate glory -

Intel Skylake 1151 with the name of the new processor in the socket ( 6700 Core i7 and Core i5 6600) we announced before the next in June. So i7 Processor 6700 the predecessor of the 4790 i7 performance when compared with how much to reveal farku? Skylake movement curve associated with the following processors possible , we have prepared performance graphics. Intel i7 6700 benchmark test that takes place in this comparison were prepared on the basis of performances given by Intel's processors in the previous year.

Once you get around the horrible grammar of the translation, it becomes obvious that they never benchmarked a single thing. They made a guess, based off of other benchmarks, and then rounded up based on reports of what the chip is supposed to do.

This isn't reporting... this is National Enquirer/Fox News/Rupert Murdock level 'news'. Bah.

[directhex]

No reason to feel sad about buying a 5820k then

[petrucius]

Nice!
I can't wait for the prices of DDR4 to drop.

[EvZ_2]

So, am I reading this correct the 5820K is still better!? Why would you bother with the 6700k?

[shaithis]

So, am I reading this correct the 5820K is still better!? Why would you bother with the 6700k?

6700k is a desktop chip with 4 cores. It costs desktop prices and with desktop priced boards.

5820K is a workstation/server chip with 6 cores. It and it's motherboards (x99) cost substantially more. I would hope it performs better.