Estimated CPU production cost

[X-Files]

Is there any major profit margin left with the actual CPU prices war between AMD and Intel? Well, I still think so after reading this acticle: http://www.ciber-runa.net/guide/ChipCosts.html

What do you guys think?

 

[m25]

Well, just consider this; the cheapest K8 available, the Socket 754 Sempron 2600+ is sold for $27 and considering all the channels it goes through, I'd give it a cost estimate no thicker than $12. Now, considering that an athlon 64 has more cache we could get it to a cost of $16 and this is for ALL the singlecore athlons, so, the higher the frequency they sell it at, the larget the profit. Following this simple but effective logics, you come to the conclusion that even a 6000+ has a cost of $32; of course, the binning here takes some extra money, but certainly they get a lot more.

 

[YO_KID37]

Excellent Article Find, Thank you For the Post. This is a Load to read but, have bookmarked it and am going through it Part by Part. Everything seems very much to what most of us need to know on calculating Costs of our CPU's.

 

[halbhh]

Great article, and pretty interesting. Kudos for digging this up. One of the arguments made in another thread suggested that the total costs ("costs of sales") was the way to guage cpu costs, but that doesn't think about overhead in a good way, IMO. I ran a business a long time, and an important idea is what I call the "marginal" cost of something. How much would one more unit of the product cost to make, in manufacturing costs. This is meaningful in certain important ways for a business. I'm glad to see these estimates.

 

[m25]

Yes, but I got pretty near with a much shorter reasoning :lol:

 

[qcmadness]

Great find

 

[halbhh]

Yes, I liked that insight too.