Why is 8950hk config upgrade SO much more than 8750h?

[eltouristo]

I'm not asking a general stupid question for people to say 'that's just how much it cost' or 'it's worth it'. I mean, does anyone have the actual reason? Does it really COST the (bga MB) suppliers that much more to get the chip from Intel vs the 8750h? I am talking about what is generally around $600 more JUST for that upgrade! But you see the internet published Intel price difference between the two chips is 'only' about $200 So...THAT is either wrong or there is something else going on. I still sorta think there is an ACTUAL 'legitimate' reason, like it does actually cost close to that much more for vendors to get it in their hands. BUT, I know it could be 'other reasons'. It seems to be consistent across all vendors...but they could just all be forced to pay up somehow from a single supplier...idk. SOMEWHERE in the supply chain it seems to be that ($600) much more...but where? and why? Anyone know? If I knew that, I might not feel quite so UTTERLY frustrated and refusing to considering to muster almost 50% more for my base config vs 8750h. Though I still probably would never pay an extra $600 for that. I would gladly pay $200-300, maybe $400 at the outside, but $600 is just like *** ?!' I can't have what I want/need because.. screw u and your *** up pricing ! It's seems just weird...like they DON"T want people to get the better chip/mb, or some weird 'catch 22' where they are afraid to have adequate supply of them from fear of less demand from slightly higher price, which ends up making price dramatically higher? R/D and production costs of MB etc are, or should be, all be the same or very nearly so. Yeah I have no idea. Just pissed really. I think it's an interesting question I've not yet seen any clarity on. I sorta doubt there is some 'conspiracy' to 'gouge' those that 'have more money than sense' but, in the laptop industry, well...let's just say there are probably too few suppliers for everyone's good, idk. Still I more strongly suspect there is a 'legitimate' reason, I just have no idea what it is and am very curious about that.

 

[ShadyHamster]

You pay a higher premium for the unlocked SKU, it's slightly higher factory clocked and probably higher binned

 

[eltouristo]

You pay a higher premium for the unlocked SKU, it's slightly higher factory clocked and probably higher binned

Yes I know 'all the usual answers'. I understand all the features and aspects. The question is not why does it cost more, it's why does it cost about $600 more, instead of a difference closer to the pricing published by Intel, which is about a $200 difference.

 

[Barty1884]

No, it won't cost the manufacturers anywhere close to +$600.

Take a look at ark:
https://ark.intel.com/products/134903/Intel-Core-i9-8950HK-Processor-12M-Cache-up-to-4_80-GHz
https://ark.intel.com/products/134906/Intel-Core-i7-8750H-Processor-9M-Cache-up-to-4_10-GHz

The recommend increase to *customers* is <$200 on the chip. Mfg's will probably be paying a $100 premium or something.


But the "why" aspect is much greater. Overclockable, requires more complex cooling solutions = more R&D = more 'cost' passed onto the consumer.

On top of that, if people will pay a $600 premium all else being equal (haven't substantiated that number myself), then why wouldn't they?

 

[eltouristo]

No, it won't cost the manufacturers anywhere close to +$600.

Take a look at ark:
https://ark.intel.com/products/134903/Intel-Core-i9-8950HK-Processor-12M-Cache-up-to-4_80-GHz
https://ark.intel.com/products/134906/Intel-Core-i7-8750H-Processor-9M-Cache-up-to-4_10-GHz

The recommend increase to *customers* is <$200 on the chip. Mfg's will probably be paying a $100 premium or something.


But the "why" aspect is much greater. Overclockable, requires more complex cooling solutions = more R&D = more 'cost' passed onto the consumer.

On top of that, if people will pay a $600 premium all else being equal (haven't substantiated that number myself), then why wouldn't they?

Thanks, but this is all still 'not real answers'. These are 'excuses' we make to try to explain something. This not necessarily 'extra' costs, both chips require cooling, and it's not clear that the 8950hk even uses different power chips or different coolers etc. Even if it did, that is not 'more' it is just 'different'. As to the similarly common 'pseudoanswer', 'if people will pay...then why wouldn't they?' no offense but this is counterproductive rationalization. It's like saying 'if you can gouge, then why not? Of course some people will pay the extra $600, and apparently the selling chain has 'decided' that price is producing 'enough demand'. Obviously less people are going to pay that much (vs a 'reasonable' increase0, and maybe some, like me, are frustrated and curious. Actual 'shortages' can 'drive' prices up but that is a 'illegitimate quirk of capitalism' that happens of course with all sorts of things...when the cost to produce is dramatically lower than price, because in essence buyers 'bid' up the price simply by purchase despite increases in pricing that being 'arbitrarily' placed on the product/service. (like the recent video card pricing, for example, which there was no mystery about). But this seems different from that, because it seems more 'stable/consistent' of an increase...and that leads to a murkier view of the cause of the increase, hence the intrigue. Sometimes price is affected by low fab yeilds, but again, it seems there's no news of anything like here. Idk, that's why I inquire...about the 'actual' reason.

 

[Barty1884]

I'm not sure what you're looking for here? We are not Intel, or any of their partners who are implementing the HK into a product.
At best, you'll get speculation/educated guesses.

Even if you could get a direct "answer" from any given manufacturer, you'd have to quantify them as "excuses" too, if that's your rationale.

1. The chips will be better binned - less production yields of these 'top' tier binned chips

2. Whether the cooling solution is "more" or simply "different" doesn't change the fact that there's added R&D involved, resulting in an increase to the end customer.

3. Prices of these chips are not independently regulated or controlled - So yes, "if you can gouge, then why not?".
Yes, there may be less interested parties at a product with what appears to be a disproportionate markup but...
1000 buyers of a product at +$600 increase makes substantially more business sense than finding 3000 buyers at +$200/product.
Less product/materials/labor required for 1000 or any given item vs 3000.
The production market is well beyond the economies of scale where they NEED to sell more units.

4. Arguably, there is a 'shortage' as to be an 8950HK, there's going to be (relatively) minimal production yields. Anything that doesn't make the grade as an HK can be an H. Ark's listing is a "recommended" increase too - so manufacturer's have free-reign to price it at whatever they feel is viable in the market.

 

[USAFRet]

Basic answer - "Whatever the market will bear", on top of the extra manufacturing cost.

 

[HEXiT]

the simple answer is they charge as much as they think you will pay.
higher end chips cost a the same to produce as midrange as far as materials. but theres also dev costs and added qc costs and a cost associated with how many samples per wafer they can produce.

but at the end of the day they charge what the customer is willing to pay and if its a business part it will often be x2 or more for a similar performing part to an enthusiast grade part. just because a business is likely to have more money to spend on its hardware.

 

[eltouristo]

The reason I'm asking this not really for a 'speculative discussion' about all the 'possible reasons'. I keep trying to emphasize that. I don't expect someone to have the answer, but that is why I ask, because we don't have an answer. I'll go back to the graphics card analogy. We knew 'the answer' in that case. We don't know 'the answer' in this case. So again, I'm asking if any knows the answer, not for a pointless rehashing of all possible forces that can 'affect' pricing. (I don't mean to pointelessly argue, but I say 'different' on purpose. This is not necessarily 'added' R&D, is it 'different'. The fallacious 'added' argument could be just as easily applied to the two chips in the opposite way. The are just 'different' they are not necessarily 'added', like any different model). I'm asking 'just in case' someone has experience in such things in reality, or actually has some connection to the vendor assembly/pricing forces in this particular case. Yes I could inquire with vendors directly too. Also, I'm not even sure Intel has anything to do with the price gap. I guess I could attempt to ask them too. But the problem with asking such things to such parties directly it they have a culture of sort of 'no comment' type answers, but there can be 'unofficial' individuals etc..that can share certain insights. Though asking vendors might be slightly more fruitful, if you happen to come across the 'right' person.

 

[Barty1884]

Based on the argument that an HK vs H is not "added", merely "different", then by logic, Intel does have something to do with it. Their MSRP is higher - when you'd be (essentially) arguing that they have no additional cost into the 'higher end' product, no?

While I'd love to be wrong, and for a vendor/mfg or someone with experience there to come along and give us a definitive "this is what X product costs more than Y product"....... I can;t see that ever happening.

You could inquire with vendors directly, absolutely. I wouldn't even expect a reply though - let alone one that actually goes into details.


Nobody is going to tell you "why" outright - at least not in any official capacity.
You'll get answers as have been posted above (except no vendor is going to tell you "because we can").

 

[rgd1101]

because it a business, and business do care about profit. If you or someone are willing to paid $600 for 5% better performance they do it.

 

[eltouristo]

Thanks to all for participating. But again, I wasn't asking for a speculative discussion. I also repeatedly referenced that Intel does charge more, and the reason for that is known. What is not known is why there is such a large difference in the cost of the final product, not the $200 difference in the chips. The costs to the supply chain, of the finish motherboards. Those boards are not necessarily any more costly to deliver. They may be, but that is part of the question. There may be detailed teardowns of both models somewhere. We would need someone to go over the power chips and the heatsinks/fans. (and any other differences) IF those were 'upgraded' that would explain some SMALL part of the premium, yeah.

 

[USAFRet]

Thanks to all for participating. But again, I wasn't asking for a speculative discussion. I also repeatedly referenced that Intel does charge more, and the reason for that is known. What is not known is why there is such a large difference in the cost of the final product, not the $200 difference in the chips. The costs to the supply chain, of the finish motherboards. Those boards are not necessarily any more costly to deliver. They may be, but that is part of the question. There may be detailed teardowns of both models somewhere. We would need someone to go over the power chips and the heatsinks/fans. (and any other differences) IF those were 'upgraded' that would explain some SMALL part of the premium, yeah.

The exact "why's" cannot be known to us out here.
That is closely guarded corporate info. Just like with any other product.
Why does Brand X, Model K cost more than Brand Y, Model Z? Even if outwardly...they look and act somewhat but not 100%, the same.

Why does an Acrua cost more than a Honda? 4 wheels, engine, interior. Same parent company. All the same, right?

 

[rgd1101]

Because they can. We do B2B at work, we have high, mid, low end products, we charge (cost to price) more on high end, because we can, and the same as the dealers, they charge more on high end products, because they target different customer.
The low end is for those only low at price.
Mid for those who know what they are looking for
High is just because they have $ to burn.

 

[Barty1884]

Thanks to all for participating. But again, I wasn't asking for a speculative discussion.

While you may not think that you were asking for speculative discussion, there's no avenue to get an official "yes" or "no" so in essence, by asking on a public forum, with no ties to any vendor, you were.

Does it really COST the (bga MB) suppliers that much more to get the chip from Intel vs the 8750h?

To get the chip? No. The raw additional cost to vendors like Dell/MSI et al will likely be comfortably <$100 vs an H.

Those boards are not necessarily any more costly to deliver. They may be, but that is part of the question.

In raw materials, no. mosfets, HSF etc are cheap. You might be looking at $1-2 in additional raw materials on any given board.

IF those were 'upgraded' that would explain some SMALL part of the premium, yeah.

Not necessarily small. You could argue one way or another - you'll argue "small", vendors will argue it's more substantial.

To play devil's advocate here:
Let's assume that incorporating an 8950HK involved an entirely new implementation (it's not, but whatever).
The R&D involved, let's call it $400k.
The added cost of the chip = $100/system
The added raw materials = $1-2/system, max.

So *insert vendor here* would look at that and say (hypothetically) ok, we can sell 3000 units at +$250.
= $750,000 income vs $706,000 "cost" = $44,000 "profit"

OR
1000 units at +$600
= $600,000 income vs $502,000 "cost" = $98,000 "profit"

Either way, the R&D is mostly complete for future implementations (making the additional cost/unit in future lower).
So they could either have their R&D further forward, with a greater market share but lesser profit.
Or a lesser market share (therefore, a larger pool of people to sell the next gen of products to) & greater profit.


Of course, all speculation - but that's all you're going to get, unfortunately. No vendor is letting anybody into their trade secrets re: pricing (let alone "some person" on a public forum)

 

[USAFRet]

How to know the true whys and hows of CPU pricing:

1. Get a job at Intel. Engineering and/or R&D.
2. Work your way up to Manager position for a particular line
3. Cross over into Marketing
4. Work your way up to Manager of that dept
5. Cross over into Sales
6. You may be able to start at the Manager level
7. Take your experience at Intel, and jump ship to Asus/Dell/Gigabyte
8. By now, you should be just short of the VP level. You'll have on your desk, a graph showing the profit difference a $15 retail price will gain or lose.
9. Come back here, violate your NDA, and tell us exactly why chip X costs more than chip Y.

Of course, this will take a decade or two.

 

[eltouristo]

Thanks but all these answers ignore the context I provided for the question. The example is that we know why video cards got overpriced by some or most sellers, though not all. It was no mystery, nor are such things 'secrets'. And also I've repeated pointed to the publish Intel pricing, but poster keep talking about 'knowing the cost of cpu's'. No this is not clearly about that, since we have that published pricing. Such mysteries are not exactly 'strictly enforced' secrets. No one would get fired from anywhere if they had the answer and said it as a rumor, it's not a 'trade secret'. It is just a mystery, not a secret. So I guess no one around here has heard anything, that's fine. Thanks anyway : )

 

[USAFRet]

You asked:
"I mean, does anyone have the actual reason?"

The "actual reason" is buried and massaged over dozens of layers between multiple different billion dollar companies.

 

[Barty1884]

You're comparing apples & oranges.

GPU prices were not increased by vendors (MSI et al) - there was a slight ($20-$30) bump in MSRP due to increased cost of raw materials (memory).
GPU prices increased as a result of supply/demand and retailers/third party sellers taking advantage.
MSI/EVGA et al, did not make any more money beyond their existing agreements with retailers for their "cut".


Similarly, Intel's pricing that is available, is a "suggested" price.
https://ark.intel.com/products/126684/Intel-Core-i7-8700K-Processor-12M-Cache-up-to-4_70-GHz
i7-8700K @ $359-$370
For sale: $350 or a little less:
https://www.amazon.com/dp/B07598VZR8/?tag=pcpapi-20
https://www.superbiiz.com/detail.php?name=I7-8700KBX&c=CJ


While you're right, nobody is going to get fired for disclosing the answer...... the 'company line' answer.
R&D, increased cooling, smaller yields etc.

You can bet, if somebody came out and announced the real reason: "because people will pay it", they'd be fired on the spot.

Sure, they could 'leak' that reason out..... but then it's just a rumor - and no more 'confirmed' than anything written in this thread thus far!!

 

[rgd1101]

There a list MSRP price, then there a actual price retailer get their, and then each retailer add whatever they think they can get it for. It just business.
Because that what capitalism do.

 

[eltouristo]

things stated here so far are actually 'off topic' and are also in the form of 'examples'. That's all ok, but of course
they are not actual reasons. "whatever they can get" is not, for example, a reason. Or at least, the examples made so are are more like arguments that there is no such thing as an actual reason (or other things just a bit off topic also). So, to try to state question another way: An examples of actual answers would be ' "because this vendor has to pay this supplier this much for this MB". or "this supplier has to pay Intel this much for the cpu", or "because the cpu's can't be bought for the price Intel publishes for them". Then we would want to know why for example the supplier charged that, but we would have an actual fragment of an actual answer. I hope people might being to see what is meant by 'actual answer'. It means something specific. It's doesn't necessarily mean something 'special' or 'secret'. So there is no just cause to assume that. Just things that are specific. Generalized (non specific) related talk/speculation is not really informative, because while it is good to know in general at first, after looking into such things for a while in you life you see that generalized talk is not new, doesn't add anything, because those are all existing, known, general considerations. It's all like the 'normal background set of general considerations'. So what I've meant is, does anyone know actual reason(s)?

 

[USAFRet]

"because this vendor has to pay this supplier this much for this MB"
"this supplier has to pay Intel this much for the cpu"
"because the cpu's can't be bought for the price Intel publishes for them"

And those exact numbers are closely held corporate numbers. You will not find them published anywhere.

 

[eltouristo]

"because this vendor has to pay this supplier this much for this MB"
"this supplier has to pay Intel this much for the cpu"
"because the cpu's can't be bought for the price Intel publishes for them"

And those exact numbers are closely held corporate numbers. You will not find them published anywhere.

that not really exactly so. 'Closely held' is not what they really are always, or even ever, but especially not often. They are just contracts. In fact sometimes contracts are quite public. Unpublished is definitely true. But if someone had that knowledge and talked about it loose terms and hints but still gave some ball park guidance that would not get anyone in trouble. It perfect reasonable to think that if you just happen to talk to right person on the phone at one of the Asian vendors or something, that they would pretty much give the actual reason. That's not farfetched, but it's not necessarily likely either. It just is what it is. It's not really easy to get those guys on the phone but sometimes you can. I might try to call Sager, for example, I've had a decent experience talking with them about Clevo etc.

 

[Barty1884]

Nobody you'll talk to at Sager/Clevo etc are even going to know the supply chain cost, let alone be willing to share it with you.
Those are decisions/information that's at a much higher level than anybody answering the phone.

Sure, somewhere along the supply chain you might be able to find out that X component costs Y amount to a given vendor.... sure, with a bit of luck & some loose lips on a given individual.

What is that going to get you though? Ok, so a MB/Chassis etc etc costs X. Unless you're putting the same call in to suppliers of every component, and having the same luck with a given individual at every factory, you're still no further forward.

"because this vendor has to pay this supplier this much for this MB"
"this supplier has to pay Intel this much for the cpu"
"because the cpu's can't be bought for the price Intel publishes for them"

And those exact numbers are closely held corporate numbers. You will not find them published anywhere.

that not really exactly so.

Except, it is.
MB suppliers negotiate their prices with vendors individually. If a vendor discloses how much they pay for a given MB, they're in breach of contract -- and they won't be dealing with that supplier any longer, at least not at that price.

How much a supplier has to pay Intel for a given chip isn't going to be knowledge the vendor is even likely privy to.
Their interest is how much the MB (etc), including the CPU, will cost them. Sure, the supplier can claim Intel are charging more to justify the cost to the vendor, but how would you validate how truthful that is?

Say you ask MSI. They tell you "the supplier told us they have to pay X for the Intel CPU". If the supplier is outright lying about that fact, what MSI tell you is then incorrect.


Nobody knows what price Intel *charge* for the chip is. So it's not even a matter of being available at Intel's listed price.
Intel list no such price. They list a "Recommended Customer Price". There's no requirement to stick to that, and there's no direct link between a customer price & a supplier cost.

 

[eltouristo]

Making assumptions about the nature of a thing without knowing the nature of it. We can't know how 'complicated' the answer is, so we can't know it's 'impossibly complicated'. It seems likely to be either the Intel deal being more than is 'recommended', or who is making the MB increasing the price vs 8750h MB, or a combination of those.