Intel Skylake / Kaby Lake

[Arachnotronic]

Nice! I see you found a 6700K..

I was very lucky in that I happened to check NowInStock and, lo and behold, the 6700K was available on Newegg w/ a combo deal.

I honestly had planned to get a cheaper board and memory (I don't see much point in expensive boards for Intel's mainstream platform), but the $60 combo savings on the DDR4-3000 and ASUS Maximus VIII Ranger seemed pretty solid.

 

[steve wilson]

Every time I mentioned the 40% number I said it 'suggests' higher clocks, never proclaimed it as a fact. Meanwhile this does sound like proclaiming a fact to me:





Oh, I have no problem with the predictions, I just find it funny that you accuse people of doing the same things you do on a regular basis when it comes to AMD products with much less available info than we have about Skylake. At least Intel mentioned the products and the benchmark used used in the Core-M comparison so that we will be able verify if their claim is true or not. A vague 40% better IPC claim can be justified in 1000x different ways if Zen doesn't live up to the hype next year.

Why do you even bother discussing anything with Fjodor?... it's like banging your head against a brick wall.

 

[Burpo]

I don't see much point in expensive boards for Intel's mainstream platform), but the $60 combo savings on the DDR4-3000 and ASUS Maximus VIII Ranger seemed pretty solid.

Dude, you got the right board.. The Asus Pro Clock feature shows what is possible for overclocking. Allows up too 400 bclk overclocking.

 

[Arachnotronic]

Dude, you got the right board.. The Asus Pro Clock feature shows what is possible for overclocking. Allows up too 400 bclk overclocking.

Well then...awesome

 

[Fjodor2001]

Every time I mentioned the 40% number I said it 'suggests' higher clocks, never proclaimed it as a fact.

Here is what you actually said:

So Core m7 6Y75 is 40% faster than Core M-5Y71 @ 3DMark Sky Diver Overall.

I.e. you are stating it as a fact.

Oh, I have no problem with the predictions, I just find it funny that you accuse people of doing the same things you do on a regular basis when it comes to AMD products with much less available info than we have about Skylake.

The difference is that I do not state it as a fact. However I have said what I expect, estimate, or guess Zen could be like. Don't you still get the difference between stating something as a fact vs saying it's an estimate/guess?

 

[Fjodor2001]

All his Skylake predictions went straight into the trash.

We've already been over this like 100 times, and you've been proven wrong on it every time, yet you bring it up over and over again. Can you please stop lying about this just to derail the thread?

Also, maybe we should bring up your failed PS4 predictions, or one of the hundreds of other failed predictions you've made?

BTW: You should put yourself on your own ignore list if you want to come closer to reality.

 

[steve wilson]

Dude, you got the right board.. The Asus Pro Clock feature shows what is possible for overclocking. Allows up too 400 bclk overclocking.

Is it a lot better than the Asus 170 Pro gamer?

 

[Burpo]

No, has the same feature..

Advantages of ASUS Pro Clock
Reduced boot up time
Lower jitter under extreme conditions for higher overclocks
Increased stability under heavy overclocking

*BCLK overclocking range will vary according to CPU capabilities, cooling, motherboard support and tuning options. Ranges up to 650MHz are possible under select conditions.

Means you can safely crank up clock and memory speeds for extreme overclocks..

"A dedicated base-clock (BCLK) generator designed for 6th-generation Intel® processors that allows overclocked base clock frequencies up to or beyond 400MHz*. This custom solution works in tandem with the ASUS TurboV Processing Unit (TPU), to enhance voltage and base-clock overclocking control — providing an exciting new way to boost performance to extreme heights."

 

[Sweepr]

First post updated with the full Skylake lineup so that it doesn't get lost in hundreds of pages.

Here is what you actually said:

I.e. you are stating it as a fact.

Look at this troll attempt guys, he purposedly deleted the rest of my post.

So Core m7 6Y75 is 40% faster than Core M-5Y71 @ 3DMark Sky Diver Overall.
This test is 5+ minutes long and Broadwell-Y can only run it at ~600-700MHz. If true this suggests Skylake-Y can sustain higher clocks.

The difference is that I do not state it as a fact. However I have said what I expect, estimate, or guess Zen could be like. Don't you still get the difference between stating something as a fact vs saying it's an estimate/guess?

I do, but apparently you don't. Saying Skylake-Y Core-M operates at higher clocks based on an Intel claim where they mentioned which products were compared (and the particular benchmark) might not qualify as a fact because there is no design win tested just yet, but it's a safe guess, unless you can provide us another reasonable explanation. Now saying Zen will smoke whatever Intel has planned for mainstream in 2016 at this point is not just speculation but pure fud, and asking others for evidence/results when you post this kind of stuff is bad faith IMHO.

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

 

[Fjodor2001]

Look at this troll attempt guys, he purposedly deleted the rest of my post.

I clearly linked to it so everyone can check the rest if they like.

But the rest was just a benchmark table without any 6Y75 benchmark, just Broadwell Y benchmarks and a comments on those benchmarks. That's why I asked for 6Y75 benchmarks, which you then said you didn't have. I.e. the rest of your post did not add any evidence to back up your claim that 6Y75 was 40% faster anyway.

I do, but apparently you don't. Saying Skylake-Y Core-M operates at higher clocks based on an Intel claim where they provided which products were compared (and the particular benchmark) might not qualify as a fact because there is no design win tested just yet, but it's a safe guess, unless you can provide us another reasonable explanation.

Guess yes, fact no. See the difference?

Now saying Zen will smoke whatever Intel has planned for mainstream in 2016 at this point is not not just speculation but pure fud, and asking others for evidence when you post this kind of stuff is bad faith IMHO.

Speculation yes, which is all I have ever claimed it to be. Also, note that I only said MT performance is likely to be better (not ST performance), and also when comparing it to 4 core mainstream Intel CPUs. Do you consider that to be incorrect, if the guesstimates about Zen are true? If so, why?

Fud no, why would it be?

 

[ShintaiDK]

First post updated with the full Skylake lineup so that it doesn't get lost in hundreds of pages.

Super. Easy to keep track :thumbsup:

Look at this troll attempt guys, he purposedly deleted the rest of my post.

Dont think it surprises anyone

 

[Sweepr]

I clearly linked to it so everyone can check the rest if they like.

But the rest was just a benchmark table without any 6Y75 benchmark, just Broadwell Y benchmarks. That's why I asked for 6Y75 benchmarks, which you said you didn't have. I.e. the rest of your post did not add any evidence to back up your claim that 6Y75 was 40% faster anyway.

Which is why I said 'suggests'. The images just illustrate the fact that Broadwell-Y sustains lower clocks than Broadwell-U. It's pretty obvious you need higher clocks to provide a 40% better score given that Broadwell-Y and Skylake-Y pack a very similar iGPU (24 EUs, Gen 8 vs Gen 9). If you want me to provide test results for any kind of reasonable assumption perhaps you should start backing up your generous AMD predictions with more than personal hopes.

Guess yes, fact no. See the difference?

Reasonable/educated guess yes, fud no. See the difference to this post?

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

 

[Fjodor2001]

Which is why I said 'suggests'. The images just illustrate the fact that Broadwell-Y sustains lower clocks than Broadwell-U. It's pretty obvious you need higher clocks to provide 40% better score given that Broadwell-Y and Skylake-Y pack the a very similar iGPU (24 EUs, Gen 8 vs Gen 9).

You said "If true this suggests Skylake-Y can sustain higher clocks.", so I agree you are clear about that this part of your statement was a guess. However the claim you made about 40% higher performance was absolute.

If you want me to provide test results for any kind of reasonable assumption perhaps you should start backing up your generous AMD predictions with more than personal hopes.

The difference is that I've never claimed the Zen guesstimates to be a fact. Don't you still get that?

As for reasonable assumptions on possible Zen performance, there are lots of people who have provided estimates on that in the Zen threads. I suggest you check those threads out if you are interested.

Guess yes, fud no. See the difference to this post?

Huh? I've never claimed your statements to be fud. It was you who attacked me for that, without any explanation.

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

Which is something I stand by, assuming the predictions that most forum members have made about Zen are true, and under the conditions I mentioned earlier. But I have already explained that in several posts to you now. Why don't you read them instead of just repeating the same quote over and over again?

Also, do you have any objections to the actual technical aspects of that? I.e. if the assumptions are true (Zen being 8 cores, with SB to Haswell performance level), don't you think Zen will be better in MT performance, compared to Intel mainstream 4 core desktop CPUs? And don't you consider SB to Haswell performance level to be close to what Intel will have to offer in 2016 (effectively Skylake)?

 

[Arachnotronic]

Also, do you have any objections to the actual technical aspects of that? I.e. if the assumptions are true (Zen being 8 cores, with SB to Haswell performance level), don't you think Zen will be better in MT performance, compared to Intel mainstream 4 core desktop CPUs?

It might be (depending on many factors), but even if we assume this is the case, what makes you think that it will be comparable in terms of price/product positioning to a "mainstream" Skylake?

Knowing AMD (*cough* R9 Fury X pricing *cough*), if they do have something that handily bests a Skylake 4C/8T on their hands, they won't sell it for $350. They'll try to price it up there with an 8C/16T Intel CPU.

Remember, this is the company that is now telling investors that they're tired of being the cheaper solution (paraphrasing). This should be a very big clue as to what AMD would charge for a part like the one you have described.

 

[Sweepr]

You said "If true this suggests Skylake-Y can sustain higher clocks.", so I agree you are clear about that this part of your statement was a guess. However the claim you made about 40% higher performance was absolute.

First of all this is Intel's claim and not mine. Second 40% higher performance refers to a specific benchmark (3DMark Sky Diver Overall) so don't start trolling my thread when overall results don't match the results of this particular test.

Huh? I've never claimed your statements to be fud. It was you who attacked me for that, without any explanation.

Which is something I stand by, assuming the predictions that most forum members have made about Zen are true, and under the conditions I mentioned earlier. But I have already explained that in several posts to you now. Why don't you read them instead of just repeating the same quote over and over again?

Asking everyone (not just me) to provide facts/results while you feel free to post predictions of your own without 'concrete evidence' to back them up is a contradiction.

Also, do you have any objections to the actual technical aspects of that? I.e. if the assumptions are true (Zen being 8 cores, with SB to Haswell performance level), don't you think Zen will be better in MT performance, compared to Intel mainstream 4 core desktop CPUs?

This may very well be true but saying Intel's 2016 mainstream products will be slow and expensive in comparison does sound like flame/fud to me, considering how little we know about Summit Ridge performance/price so far.

 

[Fjodor2001]

It might be (depending on many factors), but even if we assume this is the case, what makes you think that it will be comparable in terms of price/product positioning to a "mainstream" Skylake?

Knowing AMD (*cough* R9 Fury X pricing *cough*), if they do have something that handily bests a Skylake 4C/8T on their hands, they won't sell it for $350. They'll try to price it up there with an 8C/16T Intel CPU.

Remember, this is the company that is now telling investors that they're tired of being the cheaper solution (paraphrasing). This should be a very big clue as to what AMD would charge for a part like the one you have described.

Yes, that's a good question. Unfortunately we have no idea how AMD intends to position Zen, or at what price.

As for Fury, I think they had no option but to price it quite high, since it's such a huge die and thus costs a lot to produce. Also, AMD and nVidia have in general been seen as equals (even if nVidia has pulled away a bit lately). Therefore it's also easier to charge a similar price to their competitor's products at the same performance levels.

But in the x86 space, AMD is the underdog, so they have to price their products a bit lower. In addition, an 8 core Zen die should not be much larger than a 4 core Skylake die on 14 nm, since Zen will not have any iGPU. So the cost of producing an 8 core Zen does not have to be that high. In other words AMD at least have the option of selling Zen at a lower price than Intel's 8 core 5960X which costs $1000.

Now whether AMD actually will sell Zen at a lower price or not is another story. But if they don't and try to charge $1000 for it, then I don't think it's be a success sales-wise.

As for the technical performance claims vs Intel 4 core CPUs I made earlier, those are still valid though regardless (if the Zen guesstimates are true). But I agree the comparison to a mainstream 4 core Intel CPU will not be fair, if AMD decides to charge $1000 for an 8 core Zen.

 

[Fjodor2001]

First of all this is Intel's claim and not mine. Second 40% higher performance refers to a specific benchmark (3DMark Sky Diver Overall) so don't start trolling my thread when overall results don't match the results of this particular test.

Asking everyone (not just me) to provide facts/results while you feel free to post predictions of your own without 'concrete evidence' to back them up is a contradiction.

If someone states something as a fact, I require higher standards of evidence than if someone is clear about that they are just making a guess or estimate.

This may very well be true but saying Intel's 2016 mainstream products will be slow and expensive in comparison does sound like flame/fud to me, considering how little we know about Summit Ridge performance/price so far.

I said that if the guesstimates about Zen are true, it will have much higher MT performance and close to same ST performance vs Intel 4 core mainstream CPUs. What part of that is fud?

 

[Sweepr]

If someone states something as a fact, I require higher standards of evidence than if someone is clear about that they are just making a guess or estimate.

The performance claim was made by Intel, not me (I'm quoting Intel), so ask them (after you ask AMD for evidence about Zen's IPC, unless this rule only applies to Intel products). And I didn't state higher sustained clocks as a fact, hence the use of the word 'suggests'.

I said that if the guesstimates about Zen are true, it will have much higher MT performance and close to same ST performance vs Intel 4 core mainstream CPUs. What part of that is fud?

This one:

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

Unless you know something we don't about final clockspeeds and price (assuming 40% better IPC than Excavator holds true), it' pure fud.

 

[frozentundra123456]

Guys, just put him on ignore and dont let the thread get derailed(further)/locked.

Edit: BTW, just got an e-mail in my inbox from Tiger Direct that 6600K was available (I live in the US), but no word on 6700k.

 

[Khato]

Edit: BTW, just got an e-mail in my inbox from Tiger Direct that 6600K was available (I live in the US), but no word on 6700k.

Yeah, Newegg appears to have been keep stock on the i5-6600k consistently as well. In addition some of the other options have them in stock, and for example all of the local Fry's claim to have stock in store. So seems like the i5-6600k inventory might be stabilizing.

As for the i7-6700k, I'm optimistic that the few places which report an ETA are correct (shopBLT showing first shipment incoming on the 10th with additional inventory every week afterwards, B&H saying the 11th.) Seems likely that the majority of initial capacity has been going to OEMs to meet their timelines (aka, be able to ship most of their assembled product via boat rather than air and still make holiday shopping season) with a mere trickle of parts to the boxed processor channel.

 

[dark zero]

Why do you even bother discussing anything with Fjodor?... it's like banging your head against a brick wall.

Unless you talk about AMD and will show slides and slides about the vaporware they make.

 

[Absolute0]

6600K has not been a problem, I got mine almost 2 weeks ago after seeing them regularly in stock.

I expect 6700K will ramp up, I saw many vendors listed ship dates in September.

On another note once more people here get their Skylakes in, I'd like to see a new Skylake thread dedicated to Skylake Overclocking, performance, mobo use, RAM overclocking etc. Everyone agrees this thread is too long but has been a good source for discussing all Skylake news esp. regarding iGPU, new SKUs etc.

 

[Fjodor2001]

The performance claim was made by Intel, not me (I'm quoting Intel), so ask them (after you ask AMD for evidence about Zen's IPC, unless this rule only applies to Intel products).

The performance claim was made by Intel, not me (I'm quoting Intel), so ask them (after you ask AMD for evidence about Zen's IPC, unless this rule only applies to Intel products).[/QUOTE]
But you wrote it as a fact, then posted Broadwell-only benchmarks as if they should prove it, but they didn't. A simple clarification from your side when I requested the corresponding Skylake 6Y75 benchmarks would have done it, saying that you didn't have any and that it was speculation and/or only based on Intel's marketing slides. But instead you turned it all into a huge argument.

And I didn't state higher sustained clocks as a fact, hence the use of the word 'suggests'.

That part I have never complained about, which I already have clarified in earlier posts.

This one:

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

Unless you know something we don't about final clockspeeds and price (assuming 40% better IPC than Excavator holds true), it' pure fud.

For God's sake, how many times do I have to explain it to you? It was written under assumption that the performance guesstimates made by others will be roughly accurate and in the context I have mentioned repeatedly in several posts already. You've already agreed to that it is correct if those assumptions are true. So what exactly are you complaining about?

If you are complaining about the Zen performance guesstimates themselves and call them unrealistic and thus 'fud', then you should attack the people in the Zen threads that have made those estimates instead. And if you plan on doing that, you should let us know the technical reasons why you think they are incorrect, because a lot of people will disagree with you.

 

[Sweepr]

But you wrote it as a fact

Not true, I simply quoted info from this Intel slide:

That is the information Intel provided about their product, if we can't trust them then you better throw your Zen performance predictions based on AMD's claim of 40% better IPC down the toilet too.

...then posted Broadwell-only benchmarks as if they should prove it, but they didn't.

Correction: Sustained clockspeed values for Broadwell-Y and Broadwell-U, not benchmarks results. You were so desperate to vomit your question that you didn't even look at the images.

A simple clarification from your side when I requested the corresponding Skylake 6Y75 benchmarks would have done it, saying that you didn't have them and that it was speculation and/or only based on Intel's marketing slides. But instead you turned it all into a huge argument.

It's not speculation from my part if this info comes from Intel itself and repeating this crap won't make it true. That was Intel's statement comparing the two Core M generations in a specific benchmark (5-minute long). My actual prediction (never stated as a fact) was higher sustained clocks could explain the 40% difference. Can you provide us a simple clarification as to why you think you have the right to ask everyone for facts/results when yours own posts are full of fud and baseless speculation like this one?

You're just bitter because in 2016 you'll be sitting on an expensive and slow 4 core Intel CPU, while others will be using a cheaper and faster 8 core AMD CPU.

 

[Shivansps]

ok i got this from the last VGA driver on a asrock h170 motherboard, its newer than the one that is on Intel site.

; SKL HW
iSKLULTGT1 = "Intel(R) HD Graphics 510"
iSKLULTGT15 = "Intel(R) HD Graphics 510"
iSKLULTGT2 = "Intel(R) HD Graphics 520"
iSKLULTGT2f = "Intel(R) HD Graphics 520"
iSKLULXGT1 = "Intel(R) HD Graphics"
iSKLULXGT2 = "Intel(R) HD Graphics 515"
iSKLDTGT1 = "Intel(R) HD Graphics 510"
iSKLDTGT2 = "Intel(R) HD Graphics 530"
iSKLHaloGT2 = "Intel(R) HD Graphics 530"

Since Intel only keep the strings for released products on non beta drivers, i guess we can say that the G4500 actually has the GT2 and no a GT1.5, and that the GT1.5 exist on a U model under the HD510 name, that probably means te 4405U is a GT1.5