• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

Associate
Joined
30 May 2016
Posts
620
I know its frowned on by some round these parts but when you take a look at some actual FACTS (intels financial reports) you can see that Intel's profit margins on their CPU line is likely to be in the region of 25% or less of their pre tax prices so I think some people are in la la land when they talk of intel's cpus costing 'double' what they should....

That sounds about right. Intel produces at cost X and sells wholesale at price 1.25*X. Then the wholesaler adds on another 10% so it's 1.1*1.25*X. Then the retailer adds another 30% and it goes to 1.1*1.25*1.3*X which is about 1.8X and then there's VAT/Sales tax.

So when you get a 6900K for £1000, that's actually retailing at £840 pounds (the rest is VAT) and was bought at £650 from some wholesaler who got a large batch at around £590 from Intel who build the chip at a cost of around £470.

EDIT: My point is, people hear it costs £500 to build and see it selling for £1000, then make the wrong assumption that Intel pockets a profit of £500 per chip. That's just wrong.

However, think of the same path for Zen: if it is indeed that cheap to produce (some say it's half the cost of the 6900K) it'd make a huge difference because the margins are applied on top of each other (compounding). Using the same percentages a £250 chip should retail at about £450 plus VAT (i.e. around £550 total price).
 
Last edited:
Soldato
Joined
27 Jan 2009
Posts
6,572
I don't care why the prices have gone up because I have no control over them.

Well that's about the only thing you have said recently that makes any sense.

Man also has no control over the sun. Didn't stop people enquiring as to why the sun rose every morning on one horizon and set every evening on another. If you want to be ignorant that's your choice just don't whine when people point out that your uniformed opinions may not be based in reality

You've been shown your wrong time and time again. Calling my explanations 'spam' isn't going to cut it if you have an argument make it but have something to back up your claims. Yes Intel make huge profits but nothing like the amounts you think as a % of the cost of their products.

Intel's top end consumer CPU pricing remains stable in $$$'s since at least Sandybridge. Costs for Intel have gone up massively due to R+D spend in that time..

the 7700k has a US MSRP of $339 - $350 ......

Its simply not true that Intel have ramped up the price of their I7 CPU's over previous gens in recent years.

The price we pay fluctuates due to the £/$ exchange rate and (especially with Skylake) shortages in the retail channels (in was candidly admitted on these forums last year by a member of ocuk staff that retailers in general were making much larger percentages than normal on Skylake chips as the demand was far outstripping the supply so retailers could charge a bigger premium.)

I have previously demonstrated that Intel have NOT ramped up their pricing in recent years... Its just not true!

Check out the launch bulk prices (i.e. what Intel sell on to retailers OEM’s etc) for the previous ‘top end’ i7 consumer socket four core/ eight thread CPU’s over the past four years


http://hexus.net/tech/reviews/cpu/85...-14nm-skylake/

Launch 1ku prices

6700k $350 - August 2015
5775c $366
4790k $339
4770k $339
3770k $313
2700k $332 - October 2011


Allowing for inflation (http://www.usinflationcalculator.com/ ) from 2011 to 2015 plugging the 2700k value in gives an inflation adjusted price of…………………….


Drum roll


$351.20!!!!!!!!!!!!!!!!!

The 6700k is a 14nm CPU so where is the premium Intel charged for the CPU new over the previous gen Broadwell (which they asked more for! - probably due to the iGPU) and over the gen before that, Haswell, where they asked for a whole $11 dollars less.

Haswell to Skylake is 22nm to 14nm with a whole new CPU design to cost for

Haswell-E to Broadwell-E is the same design shrunk from 22nm to 14nm i.e. probably cheaper to deal with then a new CPU design

Oh and if you factor in inflation for the 4770k (march 2013) to 2015 the price goes from 339 to 346
so basically adjusted for Inflation Intel's 4c/8t top end cpu pricing has remained pretty much unchanged despite spending 'billions' in the mean time to develop new designs on smaller processes.......




Current Intel recommended pricing for the 6700k ...... $350!

http://ark.intel.com/products/88195/Intel-Core-i7-6700K-Processor-8M-Cache-up-to-4_20-GHz


Adjusted for inflation in $ the 6700K is not more expensive then your 2600k at launch ($317 in January of 2011) Allowing for inflation = circa $350 in 2016 money!

http://hexus.net/tech/reviews/cpu/2...ge-core-i5-2500k-core-i7-2600k-review/?page=3

Intel's R + D spend has Skyrocketed of late as they have tried to keep shrinking the processes used to make their CPU's

From 2005 to 2015 they went from being 13th to 3rd in the list of top total R+D spend worldwide across all companies with the R + D spend increasing in that time from $5.145bn for the four quarters in 2005 to 12.128bn for the four quarters in 2015

source: https://ycharts.com/companies/INTC/r_and_d_expense

If Intel had just increased the R + D in line with Inflation in US $'s they would have spent about $6.25bn so they not far of DOUBLED there R + D spend in REAL TERMS in that period.

Even since 2011 when the Sandybridge 2600K came out there spend has increased massively with the four quarter of 2011 coming in at a little over $6.43bn so the bulk of the increase is since Sandybridge came out!

The cost reductions for using a smaller process don't come close to compensating for this with the cost of a 300mm diameter CPU wafer being around $5,000

A 6700K die is around 122.4mm2...

so some quick math's shows that a wafer has a size of around 70,685mm2 (pie x radius squared)

that's around 570 dies worth of space on a wafer... now wafers are round and CPU die's are square or rectangular so there's some wastage so lets say its actually 500 dies per $5,000 wafer that's $10 per wafer so even if the die's were twice the size on the wafer the cost for the wafer part would only be around $20!

That's not going to recoup the massive R + D costs incurred for going to ever smaller processes

The Intel Pentium Processor Extreme Edition 955 launched at the end of 2005 had a dies size of 162mm2 for a CPU that cost $999 at launch.

That would be enough for around 436 dies with no wastage so say around 400 with wastage from a $5,000 wafer that's a whole $12.5 dollars per wafer a saving of $2.5 from the reduction in die size for the cost of the wafer from the P4 to the 6700k!

'Prices have gone up and costs have gone down'

No they haven't! What a load of nonsense

*It should be noted that some of the die's on a wafer will be found to be defective in some way and will either have to be discarded or used for cheaper products (i.e only using some cores) but my examples give a good idea of the actual costs for the wafer size and the 'savings' made by reducing the manufacturing process size
 
Last edited:
Soldato
Joined
26 May 2009
Posts
22,101
Just because some things have long, referenced and detailled explanation's =/= spam

When you repeat the same fanboy tripe and it gets refuted by someone pretty much every time then it is pretty much == spam.

It's a fact that Intel have been deliberately holding back progression due to having no competition while at the same time price gouging their own customers for years.
 
Soldato
Joined
26 May 2009
Posts
22,101
I'm not siding with anyone. I'm talking about the former part where you said Intel are deliberately holding back progress.

Ahh right, well the facts kinda speak for themselves but to sum up, since the launch of Sandy Bridge in 2011 Intel have been averaging a 5% performance increase per generation (or in other words **** all), the general consensus for why this is happening is that because they are/were ahead of AMD by such a margin they saw no need to bother with proper upgrades, they could save on or redirect R&D and without any alternative customers would be forced to pay through the nose for their incremental upgrades. As a result progress has slowed to a crawl and profiteering has seen extortionate pricing.

To put it in perspective the i7-2600K (Sandy Bridge) and i7-7700K (Kaby Lake) were launched 6 years apart, the performance increase is lower than between the i7-2600K and the previous generation's i7-920 (Nehalem) launched 2 years earlier.
 
Soldato
Joined
18 May 2010
Posts
22,420
Location
London
.... Who cares about stock. Turbo is where the grunt comes... This is not stock-clockers forums.

A very blinkered outlook. Stock is the product you buy and is the base measure that your money is stacked against.

Over clocking is another ball park for I imagine a small subset of customers relative to the entire customer base of CPUs and comes with no guarantees.

I would argue it's ALL about the stock out of the box experience as that's what your buying and what comes certified as a guarantee.

Anything else is silicone lottery and icing on a cake.
 
Last edited:
Soldato
Joined
5 Sep 2011
Posts
12,827
Location
Surrey
Ahh right, well the facts kinda speak for themselves but to sum up, since the launch of Sandy Bridge in 2011 Intel have been averaging a 5% performance increase per generation (or in other words **** all), the general consensus for why this is happening is that because they are/were ahead of AMD by such a margin they saw no need to bother with proper upgrades, they could save on or redirect R&D and without any alternative customers would be forced to pay through the nose for their incremental upgrades. As a result progress has slowed to a crawl and profiteering has seen extortionate pricing.

To put it in perspective the i7-2600K (Sandy Bridge) and i7-7700K (Kaby Lake) were launched 6 years apart, the performance increase is lower than between the i7-2600K and the previous generation's i7-920 (Nehalem) launched 2 years earlier.


That's purely circumstance. The only people that could confirm what you are implying are likely on 6 figure salaries.

Not that I want to entertain this, but Intel stuck with Netburst for some years when Athlon was still competitive. If there's any real evidence to the contrary that shows Intel are deliberately under achieving then that would be quite a media storm.
 
Associate
Joined
6 Nov 2005
Posts
2,418
I kinda wish we knew more about the actual turbo performance. Other than AMD telling us its a thing we've seen nothing, and it seems like it could be the make or break of this chip.

Doubly frustrating to have someone who does actually know, at least a little more than the rest, trolling :p
 
Back
Top Bottom