If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Itanium up to 5,000 servers/quarter
IDC's 3rd quarter server stats are out, and it looks like Intel has
increased the number of Itanium servers it ships by almost 60% over Q2. In Q2 they sold a bit over 3,000 Itanium2 servers, and now for Q3 they are just shy of selling 5000 servers in the three-month timeframe. Reasonably growth I suppose, but still rather pathetic volume. For comparison, a total of 1.37 Million servers were sold, giving the Itanium2 a market share of less than half of a percent. AMD's Opteron did a tiny bit better, essentially doubling it's total server shipments from ~5000 in Q2 to a bit over 10,000 in Q3. They managed over three quarters of a percent of the total server volume. These numbers just count total servers of course, and not processors. It's likely that the Itanium shipped more processors/server on average than the Opteron did, so total units of processors shipped in the quarter were probably pretty close between the two chips. The Xeon looks like it's still reigning supreme though. No specific numbers, but it looks like over 75% of the server volume is made up of Xeon-based systems (92% of all servers were x86, I'm assuming mostly Xeons but with some P4's, AthlonXP/MP and Opterons thrown in the mix). Here are a few links: http://www.idc.com/getdoc.jhtml?cont...3_11_25_140001 http://www.bayarea.com/mld/mercuryne...ss/7397043.htm http://www.infoworld.com/article/03/....html?hardware ------------- Tony Hill hilla underscore 20 at yahoo dot ca |
#2
|
|||
|
|||
"Tony Hill" wrote in message
.com... The Xeon looks like it's still reigning supreme though. No specific numbers, but it looks like over 75% of the server volume is made up of Xeon-based systems (92% of all servers were x86, I'm assuming mostly Xeons but with some P4's, AthlonXP/MP and Opterons thrown in the mix). Hmm, one can only imagine the markups on all of the rest of the world of servers if 92% of the market were x86 by volume, but were 50% proprietary by revenue. That would mean that each percent of proprietary server volume accounts for 6.25% of proprietary revenue. But each additional percent of x86 server volume only accounts for 0.54% of x86 server revenue. That means the average markup on proprietary servers is 11.5 times the markup on x86 servers! Yousuf Khan |
#3
|
|||
|
|||
On Fri, 05 Dec 2003 00:24:22 GMT, "Yousuf Khan"
wrote: "Tony Hill" wrote in message t.com... The Xeon looks like it's still reigning supreme though. No specific numbers, but it looks like over 75% of the server volume is made up of Xeon-based systems (92% of all servers were x86, I'm assuming mostly Xeons but with some P4's, AthlonXP/MP and Opterons thrown in the mix). Hmm, one can only imagine the markups on all of the rest of the world of servers if 92% of the market were x86 by volume, but were 50% proprietary by revenue. That would mean that each percent of proprietary server volume accounts for 6.25% of proprietary revenue. But each additional percent of x86 server volume only accounts for 0.54% of x86 server revenue. That means the average markup on proprietary servers is 11.5 times the markup on x86 servers! One could quibble with your methodology (the difference in manufacturing costs not being irrelevant the way they often are in these discussions), the conclusion would come out the same no matter how you look at it. *Now* you know why Intel is willing to go through much agony with Itanium. I'm sure there are lots of people here who could echo the same thing: the big iron people know where the profits are, and they aren't in what Dell is selling. RM |
#4
|
|||
|
|||
Robert Myers wrote:
On Fri, 05 Dec 2003 00:24:22 GMT, "Yousuf Khan" wrote: "Tony Hill" wrote in message et.com... The Xeon looks like it's still reigning supreme though. No specific numbers, but it looks like over 75% of the server volume is made up of Xeon-based systems (92% of all servers were x86, I'm assuming mostly Xeons but with some P4's, AthlonXP/MP and Opterons thrown in the mix). Hmm, one can only imagine the markups on all of the rest of the world of servers if 92% of the market were x86 by volume, but were 50% proprietary by revenue. That would mean that each percent of proprietary server volume accounts for 6.25% of proprietary revenue. But each additional percent of x86 server volume only accounts for 0.54% of x86 server revenue. That means the average markup on proprietary servers is 11.5 times the markup on x86 servers! One could quibble with your methodology (the difference in manufacturing costs not being irrelevant the way they often are in these discussions), the conclusion would come out the same no matter how you look at it. *Now* you know why Intel is willing to go through much agony with Itanium. I'm sure there are lots of people here who could echo the same thing: the big iron people know where the profits are, and they aren't in what Dell is selling. There *are* good profits to be made in what Dell is selling - as Dell has proven over and over again. The fact that other companies can't make a profit from selling the same things as Dell says a lot about both Dell and those other companies. -- Reply to newsgroup only please. This e-mail account is real but effectively abandoned because of excessive spamming. |
#5
|
|||
|
|||
On Thu, 04 Dec 2003 19:04:40 -0600, Rob Stow
wrote: Robert Myers wrote: snip I'm sure there are lots of people here who could echo the same thing: the big iron people know where the profits are, and they aren't in what Dell is selling. There *are* good profits to be made in what Dell is selling - as Dell has proven over and over again. The fact that other companies can't make a profit from selling the same things as Dell says a lot about both Dell and those other companies. I'm not a big fan of HP, but even their PC division managed to eke out a profit last quarter. Don't know as I'd personally want to buy one of their PC's (or Dell's) for my own use, but they're managing to do a little better than break even on them. The rest of their business is going gangbusters, but I guess the people who buy stocks prefer Dell's particular flavor of baloney: HP at a (relatively) attractive P/E about half of Dell's stratospheric 38. The difference is that HP is a computer compnay and Dell is a screwdriver shop. Guess people think the future belongs to Walmart, even in high tech. RM |
#6
|
|||
|
|||
On Thu, 04 Dec 2003 21:44:38 -0500, Robert Myers
wrote: On Thu, 04 Dec 2003 19:04:40 -0600, Rob Stow wrote: Robert Myers wrote: snip I'm sure there are lots of people here who could echo the same thing: the big iron people know where the profits are, and they aren't in what Dell is selling. There *are* good profits to be made in what Dell is selling - as Dell has proven over and over again. The fact that other companies can't make a profit from selling the same things as Dell says a lot about both Dell and those other companies. Robert Myers wrote: blah blah blah When a distributor or OEM is running that close to the ragged edge, a mnufacturer can make or break a distributor or OEM. The fact that Dell runs such razor-thin profit margins and consistently makes a profit should at least make you wonder, and you just don't seem to want to go there. Discount stores, airlines, electronics stores, etc., in fact everybody selling commodities in very efficient markets usually depends on some very sophisticated (and probabilistic) pricing, inventorying, and selling models even to stay in the game, and it's inevitable that you're going to have some losing quarters. That's just the way the random number generator bounces. If the models work correctly, your expectation is to win on the average, but that doesn't mean you're going to be able to win all the time. That is, unless a manufacturer wants to stack the deck. Why would a manufacturer do that? Because they'd rather have one or two dominant suppliers out there in relatively cozy niches not driving prices down even further. The reseller and the OEM play a role in protecting a manufacturer's margins that most people just don't get. That's the kind of stuff lawsuits are made of. I got myself excused from a jury hearing just such a case because I was at the time engaged in making those kinds of price decisions every day. I had no incentive to do it, but all you would have to do is press down on the scale a little harder for one guy than for another, and manufacturers have enough flexibility that they can pick winners and losers. Some do. You, Felger, and Dean Kent could all be right. Maybe Michael Dell is a business genius the likes of which low-profit margin businesses almost never see. Sam Walton made a fortune that way, and if he had to do it by getting extra-special nice pricing, he had to do alot of talking to do it, because he had lots of suppliers to deal with. Michael Dell, on the other hand, is critically dependent on just one. RM |
#7
|
|||
|
|||
On Fri, 05 Dec 2003 00:51:02 -0500, Robert Myers wrote:
On Thu, 04 Dec 2003 21:44:38 -0500, Robert Myers wrote: On Thu, 04 Dec 2003 19:04:40 -0600, Rob Stow wrote: Robert Myers wrote: snip I'm sure there are lots of people here who could echo the same thing: the big iron people know where the profits are, and they aren't in what Dell is selling. There *are* good profits to be made in what Dell is selling - as Dell has proven over and over again. The fact that other companies can't make a profit from selling the same things as Dell says a lot about both Dell and those other companies. Robert Myers wrote: blah blah blah When a distributor or OEM is running that close to the ragged edge, a mnufacturer can make or break a distributor or OEM. The fact that Dell runs such razor-thin profit margins and consistently makes a profit should at least make you wonder, and you just don't seem to want to go there. Discount stores, airlines, electronics stores, etc., in fact everybody selling commodities in very efficient markets usually depends on some very sophisticated (and probabilistic) pricing, inventorying, and selling models even to stay in the game, and it's inevitable that you're going to have some losing quarters. That's just the way the random number generator bounces. If the models work correctly, your expectation is to win on the average, but that doesn't mean you're going to be able to win all the time. That is, unless a manufacturer wants to stack the deck. Why would a manufacturer do that? Because they'd rather have one or two dominant suppliers out there in relatively cozy niches not driving prices down even further. The reseller and the OEM play a role in protecting a manufacturer's margins that most people just don't get. That's the kind of stuff lawsuits are made of. I got myself excused from a jury hearing just such a case because I was at the time engaged in making those kinds of price decisions every day. I had no incentive to do it, but all you would have to do is press down on the scale a little harder for one guy than for another, and manufacturers have enough flexibility that they can pick winners and losers. Some do. You, Felger, and Dean Kent could all be right. Maybe Michael Dell is a business genius the likes of which low-profit margin businesses almost never see. Sam Walton made a fortune that way, and if he had to do it by getting extra-special nice pricing, he had to do alot of talking to do it, because he had lots of suppliers to deal with. Michael Dell, on the other hand, is critically dependent on just one. RM Just like every other player or would-be player in the market, his margins are critically dependent on his entire supply chain. Intel is a significant supplier, but hardly the only one, and there's more cost spread across everything other than the chipset and processor to manage... /daytripper |
#8
|
|||
|
|||
On Fri, 05 Dec 2003 00:24:22 GMT, "Yousuf Khan"
wrote: "Tony Hill" wrote in message t.com... The Xeon looks like it's still reigning supreme though. No specific numbers, but it looks like over 75% of the server volume is made up of Xeon-based systems (92% of all servers were x86, I'm assuming mostly Xeons but with some P4's, AthlonXP/MP and Opterons thrown in the mix). Hmm, one can only imagine the markups on all of the rest of the world of servers if 92% of the market were x86 by volume, but were 50% proprietary by revenue. That would mean that each percent of proprietary server volume accounts for 6.25% of proprietary revenue. But each additional percent of x86 server volume only accounts for 0.54% of x86 server revenue. That means the average markup on proprietary servers is 11.5 times the markup on x86 servers! Yup, though I'm not sure that the term "markup" is exactly accurate here. The x86 servers tend to be MUCH smaller servers. People don't buy too many 1-4 processor UltraSparc or Power4 servers. If they're spending the resources to get an UltraSparc or Power4, chances are that they're going big. If all they need is a small server, x86 it is. A 2 processor server with 2GB of memory and a pair of SCSI drives is not going to cost nearly as much as a 32-processor server with half a terabyte of memory and a few dozens drives, no matter what processor you use to power that server. ------------- Tony Hill hilla underscore 20 at yahoo dot ca |
#9
|
|||
|
|||
On Fri, 05 Dec 2003 06:00:24 GMT, daytripper
wrote: You, Felger, and Dean Kent could all be right. Maybe Michael Dell is a business genius the likes of which low-profit margin businesses almost never see. Sam Walton made a fortune that way, and if he had to do it by getting extra-special nice pricing, he had to do alot of talking to do it, because he had lots of suppliers to deal with. Michael Dell, on the other hand, is critically dependent on just one. Just like every other player or would-be player in the market, his margins are critically dependent on his entire supply chain. Intel is a significant supplier, but hardly the only one, and there's more cost spread across everything other than the chipset and processor to manage... You are correct, but Robert's statement isn't wrong at the same time. Sam Walton (Walmart dude for those that don't know him) is critically dependant on having a LOT of products. People shop at Wal-Mart because they know that they can get (most of) the brands that they want for less. If Wal-Mart could only get a few brands here and there, they would be just another bargain-bin store for the el-cheapo brands. Dell, on the other hand, doesn't need to worry if they can't get any hard drives from Seagate. No one is going to care if Dell switches to Western Digital and Maxtor drives and doesn't carry Seagate (just an example here, I've actually got no idea what kind of HDs Dell uses). They are critically dependant on being able to get supply in several different markets, but it doesn't matter too much which supplier(s) they happen to get for that market. However, when it comes to processors and motherboard chipsets, Dell is a one-horse pony. When Intel falters in their ability to supply chips, Dell can hurt bad. We've seen this happen already a few years back, Dell claimed that tight supply in Intel processors cost them $300 million in lost sales for Q4 of '99. What's more, Dell managed to weather the storm better than some others because Intel was apparently supplying them at the expense of all other OEMs (there was quite a bit of grumblings about this, particularly from Gateway). Dell also ran into some problems in the same time frame when Intel shot themselves in the foot on the chipset market. When Intel decided that it was to be RDRAM or nothing, everyone else jumped shipped and started offering VIA chipsets. Dell stuck to their guns and pushed RDRAM. ------------- Tony Hill hilla underscore 20 at yahoo dot ca |
#10
|
|||
|
|||
"Robert Myers" wrote in message
... You, Felger, and Dean Kent could all be right. Maybe Michael Dell is a business genius the likes of which low-profit margin businesses almost never see. Sam Walton made a fortune that way, and if he had to do it by getting extra-special nice pricing, he had to do alot of talking to do it, because he had lots of suppliers to deal with. Michael Dell, on the other hand, is critically dependent on just one. Speaking of Walmart, remember that story about the woman who claimed that she was trampled by a mad horde of people rushing to buy a $39 DVD player at a Walmart? Well, it turns out this same woman has sued many times before and she's gotten thousands of dollars from various Walmarts in the past. Yousuf Khan |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
M$ not supporting Itanium for next gen OS | General Schvantzkoph | AMD x86-64 Processors | 0 | November 20th 04 02:00 AM |
New Itanium chips cost just $744 | Yousuf Khan | General | 343 | November 13th 03 09:58 PM |
Itanium performance | [email protected] | General | 2 | November 4th 03 06:16 AM |
Supercomputer interconnect technologies, Opteron & Itanium | Yousuf Khan | General | 4 | August 29th 03 12:47 PM |
Chess software benchmarks for Itanium and Opteron? | totojepast | Overclocking AMD Processors | 0 | June 23rd 03 08:39 PM |