Confirmed - i9 9900k will have soldered IHS, no more toothpaste TIM

[skaertus]

One thing for sure is it'll be interesting to see what the future holds for both Intel and AMD. It's too early to predict the outcome and the past doesn't really play a role in the current situation of both companies.....Maybe the perfect storm for AMD?

With the delayed node, production shortages, decreasing margins, and increased pressure from AMD the future doesn't look as bright for Intel currently. Intel's going to have to make some wise decisions on what market(s) they want to defend at all costs and live with the outcome whatever it may be. I don't think even Intel has the current resources to fend off AMD on all fronts.

Yoda quote....“The fear of loss is a path to the Dark Side” he might have been on to something.

I agree. Intel has made some bad decisions which may lead to loss of market share. A few years ago, Intel was producing clear superior chips, and AMD was left in the dust. But then Intel chips did not advance as much as they should, and AMD found the doors opened for launching the Ryzen chips and gain market share, becoming relevant once again. Intel increased the number of cores to compete with its old 14nm chips, and the magic worked.

Now, Intel is delaying the launch of 10nm processors once again. It also did underestimate the demand, and is having to invest a billion dollars in production of chips based on old 14nm architecture. It is incredible that AMD gained market share last month and surpassed Intel.

Now, for next year, Intel is launching the very same processors packed in a different generation number, with slightly increased clock speeds. And it is launching a top of the line processor with more cores but also a higher price. Old tech repackaged. AMD, on the other hand, will launch its 7nm processor. Will Intel’s higher clock speeds and 8-core processor for a higher price be enough to counter AMD’s real advances in tech, and probably for a more reasonable price? Let’s see how this will play out. But I think Intel may be dropping the ball and playing wrong here.

Plus, ARM chips are putting pressure on Intel, as they are advancing fast and already run Windows. They will become faster at a fast pace, and may become a viable alternative to Intel in the near future.

Suddenly, Intel’s future does not look so bright. And I am not seeing Intel doing anything to reverse this.

 

[LTC8K6]

I agree. Intel has made some bad decisions which may lead to loss of market share. A few years ago, Intel was producing clear superior chips, and AMD was left in the dust. But then Intel chips did not advance as much as they should, and AMD found the doors opened for launching the Ryzen chips and gain market share, becoming relevant once again. Intel increased the number of cores to compete with its old 14nm chips, and the magic worked.

Now, Intel is delaying the launch of 10nm processors once again. It also did underestimate the demand, and is having to invest a billion dollars in production of chips based on old 14nm architecture. It is incredible that AMD gained market share last month and surpassed Intel.

Now, for next year, Intel is launching the very same processors packed in a different generation number, with slightly increased clock speeds. And it is launching a top of the line processor with more cores but also a higher price. Old tech repackaged. AMD, on the other hand, will launch its 7nm processor. Will Intel’s higher clock speeds and 8-core processor for a higher price be enough to counter AMD’s real advances in tech, and probably for a more reasonable price? Let’s see how this will play out. But I think Intel may be dropping the ball and playing wrong here.

Plus, ARM chips are putting pressure on Intel, as they are advancing fast and already run Windows. They will become faster at a fast pace, and may become a viable alternative to Intel in the near future.

Suddenly, Intel’s future does not look so bright. And I am not seeing Intel doing anything to reverse this.

I think that is largely a FUD post.

Intel's 10nm node was not as good as it's 14nm++ node. It made no sense to go with a 10nm node that was not going to offer any benefits.

Intel's 14nm+++ node should compete okay.

Plus we have some hardware fixes for Meltdown/Spectre being incorporated into Cascade Lake and Whiskey Lake.

Intel's 10nm++ node will be along when it's ready.

 

[moonbogg]

This thing is looking more and more like the "Intel i9 buyer's remorse edition" with Zen 2 right around the corner. 8/16 @ 4.5ghz for $350 or whatever fair price they ask would make me regret going with a 9900K for nearly $600! Six hundred! SIX HUNDRED!!!!!!! Oh boy Intel is BACK baby! SIX HUNDRED!!!!!!!!! Anything to get as close as possible to that hilarious $1000 price tag they miss so much. They miss those $700 6/12 chips and $1200 8/16 chips so bad it hurts. Maybe $600 for 8/16 is the best they can do.

 

[LTC8K6]

This thing is looking more and more like the "Intel i9 buyer's remorse edition" with Zen 2 right around the corner. 8/16 @ 4.5ghz for $350 or whatever fair price they ask would make me regret going with a 9900K for nearly $600! Six hundred! SIX HUNDRED!!!!!!! Oh boy Intel is BACK baby! SIX HUNDRED!!!!!!!!! Anything to get as close as possible to that hilarious $1000 price tag they miss so much. They miss those $700 6/12 chips and $1200 8/16 chips so bad it hurts. Maybe $600 for 8/16 is the best they can do.

Intel costing more than AMD has never really been a problem, though. Not even much of a problem when Ryzen arrived.

I don't think anyone is going to be a 9900K buyer at ~$600 except those few dedicated firsters who just have to have the latest thing..

If Intel prices the 9900K at ~$600, they are just plain stupid, imo.

Such an RCP might even shock me into swapping AMD and Intel around in my priority list.

 

[DrMrLordX]

Intel's 10nm node was not as good as it's 14nm++ node.

Understatement of the year.

Seriously, though, 10nm/10nm+/10nm++++++ were essential to Intel's continued survival for numerous reasons. Intel's failure to bring that node to market is pretty extraordinary. We can lay most of Intel's current woes at the feet of that development.

 

[skaertus]

I think that is largely a FUD post.

Intel's 10nm node was not as good as it's 14nm++ node. It made no sense to go with a 10nm node that was not going to offer any benefits.

Intel's 14nm+++ node should compete okay.

Plus we have some hardware fixes for Meltdown/Spectre being incorporated into Cascade Lake and Whiskey Lake.

Intel's 10nm++ node will be along when it's ready.

Intel’s 10nm chips are not ready yet, and the market is worried at this delay.

In August, Goldman Sachs downgraded Intel shares, recommending investors to sell them (instead of keeping them) due to the struggles in getting to the 10nm process.

And it was even before Intel had supply problems with the 14nm process. Intel shares are volatile at best, which means investors are insecure about its future.

AMD shares, on the other hand, have skyrocketed this year. At the expense of Intel, of course.

While you and other enthusiasts may be OK with Intel releasing the i9-9900K at a higher price point this year, and little more than that, investors are definitely not as happy. Intel is underperforming in many ways this year, and this is mainly due to poor management, bad decisions, and lack of past investment.

 

[jpiniero]

Will Intel’s higher clock speeds and 8-core processor for a higher price be enough to counter AMD’s real advances in tech, and probably for a more reasonable price?

As long as it's faster in games, yes.

 

[Topweasel]

Understatement of the year.

Seriously, though, 10nm/10nm+/10nm++++++ were essential to Intel's continued survival for numerous reasons. Intel's failure to bring that node to market is pretty extraordinary. We can lay most of Intel's current woes at the feet of that development.

Exactly. Hell think of it this way. Coffee lake was never supposed to exist. 14nm 6 core consumer CPU's wasn't supposed to happen. The 8c version certainly waay out there in the never going to happen land. The fact that they exist knowing CPU development time really should tell us everything we need to know about how big of an issue 10nm development has been for them.

 

[skaertus]

Exactly. Hell think of it this way. Coffee lake was never supposed to exist. 14nm 6 core consumer CPU's wasn't supposed to happen. The 8c version certainly waay out there in the never going to happen land. The fact that they exist knowing CPU development time really should tell us everything we need to know about how big of an issue 10nm development has been for them.

Yes. Intel 8th and 9th gen are just gimmicks, beefed-up 14nm processors to keep consumers interested and counter AMD’s advances while Intel figures out what to do with the 10nm architecture that is not working.

 

[maddie]

Intel costing more than AMD has never really been a problem, though. Not even much of a problem when Ryzen arrived.

I don't think anyone is going to be a 9900K buyer at ~$600 except those few dedicated firsters who just have to have the latest thing..

If Intel prices the 9900K at ~$600, they are just plain stupid, imo.

Such an RCP might even shock me into swapping AMD and Intel around in my priority list.

Is hell experiencing an ice age?

 

[skaertus]

As long as it's faster in games, yes.

Perhaps for a very small percentage of gamers and enthusiasts who care only for the best performing desktop processors in the next year, and do not really care about escalating prices, nor mid-range or laptop processors, nor how the future will look like a few years from now even for the high-end.

 

[jpiniero]

Perhaps for a very small percentage of gamers and enthusiasts who care only for the best performing desktop processors in the next year

The 8700K's been the best seller at Amazon more or less since release. Coffee Lake's been a huge success for the most part.

The enthusiast market in the scheme of things is tiny, but the desktop outside of gaming is pretty much dead anyway.

 

[Topweasel]

Yes. Intel 8th and 9th gen are just gimmicks, beefed-up 14nm processors to keep consumers interested and counter AMD’s advances while Intel figures out what to do with the 10nm architecture that is not working.

Yeah my point was a bit deeper. At minimum these had to be in the works in 2014 and 2015 respectively and possibly as far back as 2012 and 2013 respectively to be out now. Which means that this 10nm issue isn't just a delay, but something they have been dealing with forever and reports that we are getting make it sound like 10nm is still in the shape they dealing with that made them start work on these CPU's back then.

 

[LTC8K6]

Is hell experiencing an ice age?

I have a friend I build computers for who likes to look in Device Manager and see a whole lot of cores listed...
He doesn't know jack about the computer itself, though.

 

[moonbogg]

Exactly. Hell think of it this way. Coffee lake was never supposed to exist. 14nm 6 core consumer CPU's wasn't supposed to happen. The 8c version certainly waay out there in the never going to happen land. The fact that they exist knowing CPU development time really should tell us everything we need to know about how big of an issue 10nm development has been for them.

I remember hearing many claims that 8700K was in the pipeline all along and was planned long before release. They said it was coming anyways regardless of outside influences. Funny how I haven't heard a single comment of a similar nature regarding Intel's totally spontaneous and purely self willed 8 core lineup.

 

[PeterScott]

I remember hearing many claims that 8700K was in the pipeline all along and was planned long before release. They said it was coming anyways regardless of outside influences. Funny how I haven't heard a single comment of a similar nature regarding Intel's totally spontaneous and purely self willed 8 core lineup.

Actually the claims I remember is that 6-8 cores were in the pipeline for a long time at 10nm. IIRC, there was a job/resume posting about it, even before Ryzen was news.

But the then 10nm failed, and 6-8 cores on 14nm were more of reaction to the failure of 10nm to show up on time.

 

[skaertus]

The 8700K's been the best seller at Amazon more or less since release. Coffee Lake's been a huge success for the most part.

The enthusiast market in the scheme of things is tiny, but the desktop outside of gaming is pretty much dead anyway.

I am not really sure. In the corporate market, there is still lots of demand for cheap desktops.

 

[Topweasel]

I remember hearing many claims that 8700K was in the pipeline all along and was planned long before release. They said it was coming anyways regardless of outside influences. Funny how I haven't heard a single comment of a similar nature regarding Intel's totally spontaneous and purely self willed 8 core lineup.

Because people can talk themselves into anything and people have short term memories. Coffee Lake showed up decently early on the roadmap. Late 2016 or so. By that point Intel had basically killed off 2 10nm items from it's roadmap. So it was easy to see that Intel was confident at the time that they weren't making the deadline on 10nm pretty early and CoffeeLake would be its response to that. The really emergency is this "willed" 8c chip. I think even as recently as last year Intel had this design worked out in its back pocket but really really didn't want to use it. But yeah Coffeelake if you try to go back any earlier than 17 was obviously added post 10nm worry. Admitted existence early enough to look like a normal release. Because they knew when they started work that Coffeelake was absolutely going to have to be a thing on the market. But CoffeeLake refresh doesn't have that luxury. But both exist as result of outside factors, that being 10nm and for the refresh 10nm and Ryzen (if Ryzen bombed, Intel would have ate the development cost just to keep die size smaller).

 

[Topweasel]

Actually the claims I remember is that 6-8 cores were in the pipeline for a long time at 10nm. IIRC, there was a job/resume posting about it, even before Ryzen was news.

But the then 10nm failed, and 6-8 cores on 14nm were more of reaction to the failure of 10nm to show up on time.

Certainly 10nm 6 and eventually 8 cores (I think that was going to be like 4 2c, or 2 4c dies in emib). Most of that was because they stopped development of Intel graphics and they needed it larger for packaging reasons. But Intel really really didn't want to increase die size at all if they could avoid it.

 

[Zucker2k]

I remember hearing many claims that 8700K was in the pipeline all along and was planned long before release. They said it was coming anyways regardless of outside influences. Funny how I haven't heard a single comment of a similar nature regarding Intel's totally spontaneous and purely self willed 8 core lineup.

Core count progress is inevitable; it just may not be by your time-table. If Ryzen quad could've taken on Kabylake quad, AMD would've been more than happy to ditch octocore chips on the desktop. That's just common sense. You should be glad this 9900k is coming because it means potentially more cores from your favorite manufacturer.

 

[skaertus]

Core count progress is inevitable; it just may not be by your time-table. If Ryzen quad could've taken on Kabylake quad, AMD would've been more than happy to ditch octocore chips on the desktop. That's just common sense. You should be glad this 9900k is coming because it means potentially more cores from your favorite manufacturer.

Yes, this is true, although Intel will charge a premium for it. The premium price seems inevitable at this point, as Intel should not even be expecting to sell a lot of 9th generation chips for the time being, given the shortage of 14nm chips. But it does very little to counter the progress of alternatives such as AMD. And Intel should be really scared of ARM chips, which are advancing at a very fast pace. The Snapdragon 845 is at least 25% faster than the Snapdragon 835, and Apple's A12 Bionic is about 20% faster than the previous generation (and all of them for the same price of the previous generation). Intel's chips are certainly faster, but I wonder for how long if Intel keeps delivering very small increases and charging a premium for every bit of them.

 

[beginner99]

This thing is looking more and more like the "Intel i9 buyer's remorse edition" with Zen 2 right around the corner. 8/16 @ 4.5ghz for $350 or whatever fair price they ask would make me regret going with a 9900K for nearly $600!

It won't be $600, at least not MSRP (probably $479). Really question is availability (very low) which might jack up prices that high. Anyway if it's $479 and AMD can compete, the 3700x won't be $350 but more like $450.

 

[coercitiv]

Anyway if it's $479 and AMD can compete, the 3700x won't be $350 but more like $450.

I doubt that. They might introduce a 3800X but 3700X will be the 9700K designated competitor.

Personally I think the i9 branding is a mistake. They are diluting the strength of a brand that was built in over a decade. Nvidia marketing got this right: if you're going to gouge, do it with your strong product branding, not some new invention.

 

[skaertus]

I doubt that. They might introduce a 3800X but 3700X will be the 9700K designated competitor.

Personally I think the i9 branding is a mistake. They are diluting the strength of a brand that was built in over a decade. Nvidia marketing got this right: if you're going to gouge, do it with your strong product branding, not some new invention.

In fact, AMD is launching a Ryzen Threadripper 2920X with 12 cores and 24 threads for USD 649 (https://www.anandtech.com/show/1344...4core-threadripper-2-cpus-coming-late-october ).

This may be a contender to the i9-9900K. And, at this price point, it may mean that AMD is expecting the i9-9900K to cost in the USD 550-600 range.

 

[jpiniero]

Personally I think the i9 branding is a mistake. They are diluting the strength of a brand that was built in over a decade. Nvidia marketing got this right: if you're going to gouge, do it with your strong product branding, not some new invention.

They wanted to charge $480 for the top model. I imagine Intel expects most to buy the 9700K, but will try to reel in a decent % to get the extra 100 bucks.

Branding wise, I imagine Intel will be switching to the metals eventually.