re:I think there will be two angles that will help AMD get into corporate space:
Lots of good, logical (hopeful) opinions here Chuck. If Intel just stands where they are, this could very well come to pass (20% chance). If Intel responds (or worse, pre-empts) in their usual manner, AMD's foray into this space is doomed (80% chance).---RJ
One is performance - I think AMD has a clear edge here and there is always a segment that will buy performance. And, Intel has a good job of training the corporate (and consumer) guys that they should buy the highest performance grade available to get maximum mileage out of the system. So, that helps
---The performance sell has always been a tough one in IT since they are usually using last years' applications only seem to buy high or midrange out of fear, but it has generally worked. If K7 has a performance advantage today, it is not that clear, and by the time OEMs think of offering K7 business SKUs, will there be one at all?-RJ
Second is alternatives - there are a lot of buyers out there who would like to see competition in the supply chain. And there are also people who support underdogs (these are the people who love to hate the market leaders and buy Apple, Netscape, AMD, Linux, etc.)
---Very true, but in my experience, those people don't thrive in F1000. Corporate IT is much more conservative and tries to drive in the middle of the road as much as possible. Apple has been essentially purged from the corporate desktop, Netscape is essentially on life support, AMD nonexistent. Linux on the other hand is getting some good looks, especially in the web server space, not in the backoffice/LoB apps. This is because of Microsoft's eternal delays on Win2000. AMD has a better chance of eating in the small business space as long as the 800lb gorilla doesn't go looking for bananas there---RJ
The qual and platform issues I think are minor irritants and are more FUD than reality. The key is OEM support. IMHO, OEMs didn't use K6s in corporate space because AMD was selling K6s solely on price. OEMs hate this because corporate segment is where they want to sell the highest ASP stuff. Asking them to sell K6-2 products based on price is like asking them to cut their own throats. (by the way, from what I gather some OEMs currently hate Intel for bringing in Celerons at a low price without performance penalty and are encouraging AMD and Intel do as much as they can to differentiate products consumer and business segments - AMD has an easier task than Intel on this count) But, with K7 it is different. That alone will make some OEMs adopt. Once OEMs adopt - corporate IS will buy.
---On this one I disagree wholeheartedly. Qual and platform issues are large in most companies. They are expensive, time consuming and a large irritant to IT - they have been very happy for Intel to delay the 440BX-820 migration as long as possible. Just speed bumping from PII-300 to 400 to 450 to PIII 500 to 550 to 600 on the same BX platform has been a breeze for them, but when the platfrom changes, it is very different. Why didn't IBM and CPQ offer K6 SKU's to business? - because their internal customer surveys told them that customers didn't want them because of several risk factors - compatibility, TCO, stable (long term and reliable) supply being tops. OEMs probably hate Intel for a variety of reasons, but Compaq/AMD started the price/margin war and Intel responded. Paradoxically, Intel did such a good job of smearing the Celeron brand in it's early days that IT still believes that it is an inferior product. Even today, Celerons are usually quoted by OEMs to IT as a desperate attempt to win on price if all else fails.---RJ
As for the server space, Athlon Ultra is going to be a very compelling product. Most of the business here is in single and dual processor systems and AMD was wise to do the chipsets for this segment inhouse. Compaq/Alpha are also going to play a major role in this segment. I say - just watch - you will be surprised how quickly they will ramp in this segment.
---Again, an order of magnitude more difficult to crack than the corporate desktop. Lots more time, lots more money in validation, etc. Why would IT buy one of these if they already have multiple safe choices from Intel, not to mention Sun, HP, etc. I'm hard pressed to believe that K7 will have a performance advantage over Cascades or especially Merced/Foster by the time this becomes an issue.---RJ
Finally, can AMD sell several million K7's by Q1/Q2 of '00? I believe they can make them, I'm just not convinced they can sell them. If everything goes perfectly for them technically, they still have to get past the somewhat predictable response they are going to get from the 800lb gorilla that will protect its bananas.---RJ
Good topic Chuck. I believe that my Ali's Tourette's Syndrome is clearing up since we've had a civilized exchange of opinions.
Rick |
