Gary,
Thanks for your post.
Re: "Intel's initial foray into the NIC and hub space is to enable faster communications (eliminate bottlenecks) so that the PC's processing power is being exercised to the fullest."
Good comment. The faster the piping/NIC, the more processing that can be done (i.e. if the processing doesn't exceed the capabilities of the networking, more chip sales.)
Maybe you mean netbog isn't all the fault of the NOS ;)
Re: "Furthermore Intel's success in these market is in large part due to their ability to bundle these devices onto the platform."
Interesting. Their sales are happening because of bundles - with what exactly? I'd be interested in discussing this more.
Re: "Intel has at no point in their history shown an ability to succeed in a systems business. Building networks is a whole lot different than building processors. "
Intel has built more than processors. They have built and currently build system communication products.
Re: "Intel certainly has the manpower"
humanpower
Re: "and money to compete in the networking market but doing so would mean spending less time on the core products - flash, embedded, central/network processing."
I agree.
Re: "Intel wants to be the building blocks to this market. "
I know Intel wants to play Switzerland and sell components to all the players, but my question isn't about what's in front of us, but what's going to happen in the way, way out future, when things become more commoditized on the low-end. How big is the future potential of the low-end commodity market?
Re: "Of course this leads to the next point - creating a standard platform for communications...thereby turning the LU's, NT's, and CSCO's of the world into DELL, CPQ, GTW type compaines. Well, in the PC/Server world this is a doable task, however today communications equipment is quite diverse. With Microsoft's help there is no-doubt that PC platforms will be used in some lower-end communications environements"
How big is the estimated future (not today's) market size for the low-end market?
Re: "Yes, CSCO could partner with INTC with INTC selling the platforms and CSCO selling the OS - where IOS would become the WIN of networking. "
So why don't they? If the IOS is bought at the same margin, why not?
Re: "LU and NT have a variety of operating systems. "
That's a big problem. This is where someone like a Cisco or MS would win out.
Re: "Microsoft could do it as well, but Microsoft hasn't got the capability within their OS or the perception of reliability to fill the market requirements. "
On the low-end, cost is very important and I suspect Microsoft is better set up to handle software costs for a low-end market on OEM sales than Cisco is, which doesn't appear to have a significant history selling software.
Re: "Do you think CSCO would sell the farm and get out of the network hardware business?"
If the crown jewels are in the IOS, then Cisco wouldn't be getting out of the network business. Didn't you say Cisco was really a software company? It would be a refocusing on developing IOS/software, meanwhile letting the low-end hardware be developed by an experienced high-volume manufacturing firm like Intel. The delivery mechanism of Cisco's IOS product to the customer changes.
Re: "What benefit would this provide to Cisco and to Cisco's customers."
Maybe higher volume sales? Lower costs?
Intel is great at high-volume manufacturing, Cisco is good at IOS. By letting each of the companies do what they are best at, I would think Cisco could benefit by piggybacking and leveraging Intel's great manufacturing strength. As long as the margin in the sale (of IOS bundle onto Intel hardware) is the same (or the profit is made up in significantly higher volume sales) as Intel's manufacturing prowess cranks out high-volume. Your thoughts?
Re: "Furthermore do you think INTC wants to deal with all that overhead, cost, and engineering???? "
Isn't that exactly what Intel is good at: low overhead, cost, engineering, manufacturing.
Re: "In the end INTC wants to sell components. They want to be the component supplier to PC users, to server users, to the internet, and now to networking companies."
Agreed this is their #1 goal, but I'm wondering about the future.
Re: "INTC knows nothing about selling, servicing and supporting a networking installation. the design, installation and support requirements are orthoginal to INTC's business model."
And Intel knew nothing about NICs/routers 2 years ago. Also, Intel has experience in selling communication systems which have these same types of parallel issues.
Re: "INTC could certainly build a team to accomplish such a thing however once again this would merely reduce the market for INTC networking components."
Not if the market were to eventually separate into a high-end and low-end (where there is no real distinguishable mid-range), and where the low-end is standardized and commoditized to the point where Intel could do this themselves and leverage their manufacturing strength to everyone's benefit, including Cisco's.
Re: "In the end it is counter productive for INTC to get more into networking - unless they intend to own the entire market "
I disagree. I'm not questioning the entire market. I'm not talking about the high-end, which is where I think Intel probably would have issues. I'm wondering about the market segments on the mid-range and low-end.
Re: "concede that carrier oriented vendors such as LU and NT will not purchase their silicon. "
I would think LU would be very happy if Intel squeezed Cisco out of the market and left LU to own the high-end? Speaking of the carrier market, do you have any thoughts on Csco's vs LU in the carrier market?
Re: "Assume your case...CSCO is gone - there is no one left to sell networking silicon to. "
For clarification and understanding on your statement above: what about LU?
Re: "Furthermore INTC has probably lost the PC processor war by then because of spending so much time focusing and buildng networking (they can't win at both) AMD, Cyrix, and CSCO are not slouch companies."
Your example is most interesting. AMD appears to almost be on their way out of the processor business as it seems they have a make-or-break situation, and Cyrix just exited the standalone processor business. I'm not sure you would want to compare Cisco to AMD and Cyrix in this regards.
Re: "LU makes their own silicon"
Could you please expand on this?
Re: "- The IXA provides foundation technologies. Most of what Cisco provides will stil be required."
I know there will exist a high-end market, but I bet things on the low-end can be commoditized over time. Some basic things can be absorbed into the OS. I'm wondering what parts can't be commoditized and how much of the mid-range market will erode and turn into the low-end market? i.e. where is the line drawn? I'm really keen on knowing how big the future potential market size is of a low-end commodity market and how these products would differ from the high-end products - i.e. what classes of feature sets will be continually developing and evolving (i.e. a description of aspects of the high-end that won't be commoditized).
Re: "To ask such a question tells me that you might want to consider looking ...at the evolution of technology period. Technology continues to evolve - creating new functions and uses for products.
Gary, I have some experience with the evolution of technology, having initiated new and successful products, having written vision documents for new products, and having managed the implementation of absorbing certain technology into Windows at Microsoft. I very much enjoy emerging markets. But, yes, in general (with one exception), Cisco routers are not my area of expertise. And that's the purpose of SI - to be a place where people with various strengths can come together and piece things together.
Re: "At no point in our short history has all functions been captured"
Agreed. However, not all functions are needed in the low-end markets where commodity occurs and cost rules. Where is the separation between the markets and how likely is commoditization and how big is the opportunity?
Unrelated, what are your thoughts about the Linux router? I imagine Cisco isn't very happy about this?
Amy J |
