[Canard PC Hardware] Intel prepares Ryzen's response behind the scenes

[Abwx]

The bulk of PC CPU sales do not happen in this price range.

Surely, but the bulk of the profits is made in this segment..

 

[swilli89]

The bulk of PC CPU sales do not happen in this price range.

Would you mind posting the source of that market data? I'd be interested to look.

 

[TheELF]

Hyperthreading i5's ? Yea no, I dont believe it . That would signal death of at least i3 and down

They are talking about one model i5 with maybe HT enabled,the cheapest i7 is 303 while the highest priced i5 at the moment is at 242 the i5 with HT will cost more then that so at the most we are talking about a $50 cheaper but also slower i7 ... that's not gonna kill nothing.
It's also not going to tempt anyone away from buying ryzen.

 

[Arachnotronic]

Would you mind posting the source of that market data? I'd be interested to look.

Take Intel's client computing group platform revenue and divide by third party market estimates of total PC unit sales shipped (times about 0.95 or so, assuming AMD has about 5% of the market). This gives an average price for CPU and PCH (counted as a platform) that's well south of $200.

My math is: $30.7B/(.26 billion PCs*0.95 share) ~= $124.

That's average for CPU and PCH combined, so CPU average prices are actually lower.

 

[DeletedMember377562]

I doubt that i5 will have SMT enabled . In their edit to the article they state that they were told by some sources it actually won't have SMT enabled which would fit perfectly in line with intel's current and all previous i5/i7 lineups.

They said they heard conflicting reports. Not that the i5 won't have SMT. Read the edit input again. Even Canard themselves they should know the fully story within the next week.

It would make no sense to call that part i5 since it would be i7 for all intents and purposes.

It also makes no sense for Intel to disable HT and call it an i5, when they can sell everything with 4c/8t. It does, however, make from a marketing position. They segment the market, as well as price the 4c/8t higher.

The same is the case here. Intel already sells both i5 and i7s with 2 cores and addtional 2 threads with HT for for laptops. Their only difference being their frequency and cache size. Just like here. But it hasn't stopped people for paying out more cash for i7s in their laptops (even though it also has meant more throttling), has it? The i7 part has higher clock speeds as well as higher cache. The i7 is a also a pretty popular brand name now -- just like Pentium is for those 40-50 year olds that still buy it.

Also, it's pretty much expected that i7 will move up to six cores in 2018, according a lot of sources.

Still, this will hurt their i7. So I agree with you there. But maybe it's a small sacrifice to pay, as they need to take this step to compete against Ryzen . Their current i7 sells for $320. Who is going to buy it, if AMD releases a 6c/12t, or maybe even 8c/16t, for the same price and IPC just below it?

All this is of course speculation. Canard themselves say Intel is only testing, not deciding anything. Higher frequency and HT are features that already exist in their current i5; they only need to activate it. So this can be seen as a "last ditch effort", after their price reduction (which some chips, in particuliar the 6900K, will need), if Ryzen turns out to be this great chip.

i7 7740K is a bit meh, only a 100Mhz bump in base clock - that is immaterial.

It's still better than nothing. According to Canard, GloFo has yield issues, causing Ryzen to have problems reaching 4 GHz and even more. So Intel is doing everything they can to beat Ryzen on single core performance (which is what's most important for games). Maybe it's only an addtional 100 Mhz from the 7700K, but it's still ~2%, and better than nothing. Not to mention it also looks better on paper for the average Joe.

 

[sm625]

My greatest hope for Ryzen is that it puts an end to this ridiculous HEDT segmentation which is really nothing more than massive price gouging. Check out the motherboard review on the anandtech homepage right now. Look at that price list! My god the people buying these things are flat out nutjobs. You can buy complete 8 core systems off lease for less than the price of half those motherboards, yet the IPC on those older systems is only 15% lower.

 

[John Carmack]

The rumored 7640K if it has HT would only be interesting if it isn't ridiculously overpriced like the 7350K.

 

[swilli89]

Take Intel's client computing group platform revenue and divide by third party market estimates of total PC unit sales shipped (times about 0.95 or so, assuming AMD has about 5% of the market). This gives an average price for CPU and PCH (counted as a platform) that's well south of $200.

My math is: $30.7B/(.26 billion PCs*0.95 share) ~= $124.

That's average for CPU and PCH combined, so CPU average prices are actually lower.

When you were responding to the question of "PC CPU Sales" were you including all pre-made OEM or just discrete retail CPU sales?

I'm positive the average CPU cost is far south of $200 with all beige box systems included, but I'd like to see data that suggests that retail CPU sales average anywhere near under $200.

 

[Vaporizer]

So Intel is using the centurion method by pushing TDP and clocks to be competitive. Seems disappointing...

 

[postmortemIA]

Yep, this smells like P3 1GHz fiasco. Ryzen must be game changer, otherwise I don't see why they would even bother.

 

[inf64]

Found something interesting on reddit :


It looks like intel is doing some training and damage control before Ryzen launches. Note the mentioning of top 1.7K$ 10 core SKU and slight shifting of goal posts in last paragraph where they mention other products aside from CPUs.

 

[Sven_eng]

I really wanna take that quiz lol.

 

[Crumpet]

That's hilarious.

 

[NTMBK]

Found something interesting on reddit :


It looks like intel is doing some training and damage control before Ryzen launches. Note the mentioning of top 1.7K$ 10 core SKU and slight shifting of goal posts in last paragraph where they mention other products aside from CPUs.

"Core count doesn't matter! But also we have more cores."

 

[NTMBK]

So Intel is using the centurion method by pushing TDP and clocks to be competitive. Seems disappointing...

Hey, clock bumps were just standard operating procedure back when AMD and Intel were actually competing. Both companies were pushing updated versions of their chips with clock bumps on a regular basis. Just look at how many different version of the Phenom II X4 came out- between January 2009 and May 2011 AMD pushed the clock speed all the way from 3GHz up to 3.7GHz, with the 940, 945, 955, 965, 970, 975, and finally 980.

 

[raghu78]

I think Intel realizes that AMD has a very competitive product which has not been the case for a decade. So I am not surprised to see them a bit concerned. One thing is sure AMD is going to take market share. If Intel wants to hurt AMD they have to hurt themselves first by cutting prices and cratering profits. I think 2018 is going to get more interesting as we will have no tock from Intel till 2019. Skylake-X vs Zen+ is going to be a mouth watering prospect.

 

[Arachnotronic]

I think Intel realizes that AMD has a very competitive product which has not been the case for a decade. So I am not surprised to see them a bit concerned. One thing is sure AMD is going to take market share. If Intel wants to hurt AMD they have to hurt themselves first by cutting prices and cratering profits. I think 2018 is going to get more interesting as we will have no tock from Intel till 2019. Skylake-X vs Zen+ is going to be a mouth watering prospect.

So losing market share wouldn't hurt profits but cutting prices to defend market share would?

Also Skylake-X is coming in August, so it's really Zen vs Skylake-X/Kaby Lake-X.

 

[raghu78]

So losing market share wouldn't hurt profits but cutting prices to defend market share would?

Also Skylake-X is coming in August, so it's really Zen vs Skylake-X/Kaby Lake-X.

Its a damned if you do damned if you don't situation for Intel. There is no way for Intel to expect the same profits and market share when there is a competitive alternative. dude Skylake-X is realistically 6 months away. In fact Skylake-X is probably going to be the HEDT lineup from Intel for atleast 24 months. With Coffeelake for desktops and notebooks being at 14nm in 2018 I do not see Intel moving to 10nm for HEDT anytime soon. Now tell me if you have a Skylake-X in Sep/Oct 2017 and a Zen+ in late Q1 2018 and the vast majority of Skylake-X lifetime spent competing Zen+ would you call Zen+ as the real competition for Skylake-X or Zen.

Arachnotronic how much ever you defend Intel nothing is going to hide the fact that Intel became complacent. There is nothing interesting from them till their next tock Icelake probably in mid-2019. If Cannonlake and 10nm were in good shape why is Intel releasing Coffeelake on 14nm in 2018. Intel has a massive hole in their roadmap for roughly 2 years and AMD is going to drive some serious gains. AMD has tocks lined up for 2018-2020 according to their CTO. Intel is walking into a perfect storm.

 

[Vaporizer]

No. It depends on price for Zen if the competition is 7700 or skylake x.

 

[Arachnotronic]

Found something interesting on reddit :


It looks like intel is doing some training and damage control before Ryzen launches. Note the mentioning of top 1.7K$ 10 core SKU and slight shifting of goal posts in last paragraph where they mention other products aside from CPUs.

Typical marketing BS designed to be accessible to kids who are working at Best Buy on summer break, hah.

Also, weird that they reference Iris and Iris Pro considering they recently killed both in favor of just one high-end gfx brand -- Iris Plus.

As far as the "shifting of the goal posts" I don't think so. Remember, this info is for people trying to sell complete systems to computer novices (you don't go asking the sales rep for help if you are a DIY enthusiast), and so all of those complementary technologies that Intel is promoting here are helpful in moving product as they are marketable.

 

[dfk7677]

As far as I understand Kaby Lake-X is going to be a higher clocked 7700K with higher TDP. I still believe that AMD will put their 6C processor against that. I also find it difficult to believe Intel will reach a 10% clock increase with Kaby Lake-X.

As far as the HEDT parts (Skylake-X), I think AMD will only target the 6C (7800K?), with their top model.

Intel will still have more ST performance but AMD will offer 2 more cores, at bit lower price (that is my own assumption).

In my opinion Intel will lose big time in the HEDT segment, unless they drastically lower prices (and hurt their margins). The 7800K (which will suffer the most) offers the lower margin , as it has a die fitting 10C (which will be larger than the die for Ryzen).

I think that Intel wishes they had their first non HEDT 6C processor as 7700K instead of a 4C now and not in one year.

 

[raghu78]

I think that Intel wishes they had their first non HEDT 6C processor as 7700K instead of a 4C now and not in one year.

This is very true. Intel's response to Zen should have been Coffeelake in 2017. But this is an arrogant Intel which grew complacent after a decade of domination. Unfortunately for them Coffeelake will have an even more difficult task against Zen+ in 2018. Intel's margins are done for no matter what they do.

 

[Arachnotronic]

Its a damned if you do damned if you don't situation for Intel.

That's your opinion, please don't state it as fact.

There is no way for Intel to expect the same profits and market share when there is a competitive alternative.

Define "competitive"?

dude Skylake-X is realistically 6 months away. In fact Skylake-X is probably going to be the HEDT lineup from Intel for atleast 24 months. With Coffeelake for desktops and notebooks being at 14nm in 2018 I do not see Intel moving to 10nm for HEDT anytime soon. Now tell me if you have a Skylake-X in Sep/Oct 2017 and a Zen+ in late Q1 2018 and the vast majority of Skylake-X lifetime spent competing Zen+ would you call Zen+ as the real competition for Skylake-X or Zen.

Huh. So you assume the absolute worst case for Intel ("Skylake-X is HEDT for 24 months"), the best case for AMD ("Zen+ based Summit Ridge successor coming in 1Q 2018) and then conclude that Intel's competitive position will be poor. I think you might want to rethink your assumptions.

Arachnotronic how much ever you defend Intel nothing is going to hide the fact that Intel became complacent.

Only in your opinion. If you look at what has actually been happening, Intel has been putting out new processors for notebooks, servers, and desktops at a regular pace. They did screw up by not bringing Broadwell to desktop, that was a case of a poor decision that came back to haunt them (though they recovered by getting Skylake out when it was supposed to come out -- more or less), but to say that Intel has been complacent when their R&D spending is at all time highs and they are taking actions to make sure to put out new products each year (14nm+/Kaby Lake is a good example) is just unfair.

I urge you to look at the situation with a little more objectivity.

There is nothing interesting from them till their next tock Icelake probably in mid-2019. If Cannonlake and 10nm were in good shape why is Intel releasing Coffeelake on 14nm in 2018. Intel has a massive hole in their roadmap for roughly 2 years and AMD is going to drive some serious gains. AMD has tocks lined up for 2018-2020 according to their CTO. Intel is walking into a perfect storm.

OK, so first of all, I never argued that 10nm is in good shape, I actually think it's in pretty poor shape right now. But let's talk about Coffee Lake. If Coffee Lake is similar to what we saw with Kaby Lake, then we should see Intel improve the performance of its transistors again (14nm++) and the leaks show that they are planning a 6 core die. How is this "complacent" if they are adding more cores, tweaking the process, and probably tweaking the physical design to potentially improve maximum frequency, too?

It seems to me that they are going to deliver a product that will drive performance in desktops and high performance notebooks pretty nicely up. Working to build better products each year does not signal complacency.

As far as AMD "tocks" you are again introducing assumptions that I don't think are necessarily given. In AMD's mind Bristol Ridge was a "tock". Do you know what Bristol Ridge was? An optimization of Carrizo, in the exact same way that Kaby Lake was an optimization of Sky Lake. They improved the transistor performance (working with GloFo on this, obviously) and it is likely that they had to redo the physical design to support the enhanced 28nm process.

I also see you are now referring to Ice Lake as mid-2019 rather than late 2019 as I have seen you post in the past -- an improvement, but I think once again you are being overly pessimistic about Intel's product launches. Intel seems to be executing to a "one product family every year" cadence. Kaby Lake in early 2017, Coffee Lake in early 2018. Why would Ice Lake be mid-2019? Based on the release cadence, early 2019 seems more likely and late 2018 if they want to be super aggressive.

 

[Atari2600]

That's your opinion, please don't state it as fact.

How can it be anything else?

Unless Zen performance is actually much worse than has been released/leaked to date, then AMD will increase their market share.

Assuming that market share increase is beyond any overall CPU sales increase, then Intel are losing sales. At that point they can either:

1 - continue with pricing structure as is and accept reduced sales with lower overall revenue and from that, lower profits.

2 - lower SKU prices to keeps sales up but accept lower profits per sale (and I suppose, also reduced revenue due to the lower sale price).


How can they be anything but damned if they do and damned if they don't. That's not an opinion, you could I guess say its conjecture based on expected Zen performance, but then Zen would have to be a unmitigated disaster for the status quo to continue.

 

[Arachnotronic]

How can it be anything else?

So Intel has no ability to respond with new products? New SKUs? The topic of this thread is about the new SKUs that Intel is supposedly launching, and then when Skylake-X/Kaby Lake-X come out, Intel has quite a few options available to it.