Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : How high will Microsoft fly? -- Ignore unavailable to you. Want to Upgrade?

To: SunSpot who wrote (39037)	3/5/2000 12:16:00 PM
From: John Carragher	Respond to of 74651

3/4/00 Consensus Estimate Alert! MICROSOFT CORP (MSFT) EPS estimate up $0.01 to $0.44 for quarter ending 6/00. Earnings Details: Stock Qtr End # Est Change Curr Est Prev High Low Yr Ago MSFT 6/00 20 +$0.01 $0.44 $0.43 $0.46 $0.42 $0.40

To: SunSpot who wrote (39037)	3/5/2000 2:41:00 PM
From: rudedog	Read Replies (2) \| Respond to of 74651

SunSpot - The "EtherNet" problem you are describing is not an issue of full duplex or half duplex - in a CSMA/CS environment, duplexing is a nit. The problem the Cisco guys were outlining is related to stack overhead in the drivers, not the duplex setting. In terms of wire protocol, all of these OS variants run "at full speed" - i.e. at the wire speed. Collisions reduce effective bandwidth to between 60% and 85% of pipe bandwidth depending on traffic... most folks agree that if the pipe is more than 60% utilized it needs to be expanded. MSFT does have more stack overhead in the drivers than Linux and most Unix variants but none of them do a good job. MSFT averages 40,000 instructions to move a packet of data, Linux about 25,000... whereas a VI based NIC will execute only about 2000 instructions to get the job done. Next generation I/O subsystems like infiniband which implement VI pretty well can reduce that overhead even more. So IMO every OS out there "sucks" when it comes to NIC driver overhead, a pox on them all... when 10MB ether net was the norm, designers could afford to be sloppy. 100MB ethernet made it harder but faster processors came to the rescue. But gigabit Ethernet and infiniband will require much lower overhead, fewer context switches and in general keep the processor out of the game. database technology is mainly generating earnings by selling consulting hours and training. It's about 50/50 for Oracle. MSFT does not make that consulting revenue directly, but others do... and also MS SQL requires less administrative support. From a customer standpoint that's a good thing. re: The difference between MS SQL and Oracle is, that when the customer wants quicker responses than 2 minutes for a customer lookup, the Oracle guys can do something about it That statement is just not true. Oracle has not been the leader in query performance for a long time - SQL server, DB2 and even Informix (remember them??) are faster. If you think otherwise you are a victim of Oracle hype. When the original Sybase engine first came out in 1986 or so, it was about 10 times as fast on the same hardware as Oracle. It took Oracle nearly 10 years to get back into the same performance range. In the past, the advantage that Oracle has had is that it runs on bigger hardware, so in absolute performance terms one could get better performance on the biggest problems, but at 5X the price. Now Oracle is #2 in absolute performance as well, and still at a huge price premium. Take a look at tpc.org MS SQL holds the top 2 positions for absolute performance. DB2 is not in the top 10... this is not price performance, it is the raw ability to serve up the data the fastest. So I would have to say that unless you are among those who don't believe that TPC means anything, your statement about query performance is just wrong. Now take a peek at the top 10 price-performance numbers - tpc.org There is NO ONE BUT MSFT SQL SERVER in the top 10. SQL server still has a ways to go in terms of some features that enterprise customers need, but in terms of performance and cost it has blown away Oracle, DB2 and everyone else in the game. Password protection is as much a function of the administration policy as it is the OS capability. As far as other encryption, are you saying that the 128 bit encryption that MS uses is somehow different than what everyone else uses? Encryption standards are not an area where a company can innovate... if they did, cross platform encryption would not work too well. MSFT just reads the manual like everyone else.