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 : Mylex -- Ignore unavailable to you. Want to Upgrade?

To: E_K_S who wrote (1651)	7/28/1998 10:36:00 PM
From: Baker	Read Replies (2) \| Respond to of 1820

Eric, according to Mylex, the SG&A was higher this quarter due to a $400K legal settlement and about $200K in other legal costs. Of course, the $400K is a non-recurring charge but, according to a questions that was asked, Coleen indicated that future legal costs will "continue to be material." But not as high as they were this quarter (though there is still a long running legal issue with a Dr. ?) Other points from the conference call: There were a total of 47 positions eliminated. The 27 they originally announced and another 20 from the NP&L area. Mylex has taken back 3 accounts from AMI (but they would not disclose who they were - HP definitely was not one). As it relates to HP, since HP did set the TPCC record with the 8 way Mylex product, HP has to offer the product within six months. Mylex has never stopped their dialogue with HP and Al describe this dialogue as being as "vigorous as always." Mylex currently has product in the labs of AMI customers. The gross margins are much better on the new products (good for both Mylex and the customers as Mylex can provide them better pricing and higher margins for themselves). Expected savings from the work force reduction and restructuring will be about $2.3M per quarter. Pacific Region sales did increase to 18% which is up from previous quarter (this is good). Mylex expects 5% revenue growth over Q2 each in Q3 and Q4. This is taking into consideration the slowness in Asia and a typically weak Q3 in Europe. Mylex purchased another 500K shares during Q2 bringing their total to 1.3M. Purchase Orders for Digital are now coming from Compaq and they have seen no slow down. Digital has been using Mylex DAC960PU Ultra SCSI for the last few years and the DAC PJ is in qualification which they to ship in Q3. Al was asked to speculate about when (if ever) Mylex would begin to see a slow down given the Compaq/Digital merger. Al said it was strictly speculation on his part as they have no information but if there were any slow down/loss it would probably be several quarters away. Probably the highlight of the conference call for me was that according to Al, they expect, given the cost savings they've instituted that MYLEX WILL BE OUT OF THE RED Q3. Though he wasn't asked a follow up, I'm guessing this will be before the one time charges related to the restructuring though I could be wrong. Mylex feels they are well positioned to move forward as they have a group of brand new products which are just emerging. They can provide a full range of solutions to OEMS on all levels of serves from low price to high performance. Neither AMI or Adaptec can do this. Hopefully, I've accurately stated the points. Thanks.

To: E_K_S who wrote (1651)	9/4/1998 12:20:00 PM
From: AJ Berger	Respond to of 1820

Mylex getting into 133mhz PCIX chip sets? Big PC makers split with Intel on server design. How big is the breach? By Lisa DiCarlo 09/03/98 09:50:00 PM Three rival computer makers are plotting to dull what they see as Intel Corp.'s increasing control over the design of NT servers. IBM (NYSE:IBM), Hewlett-Packard Co. (NYSE:HWP) and Compaq Computer Corp. (NYSE:CPQ) have jointly and surreptitiously developed an enhanced PCI specification for servers that may provide up to a sixfold increase in I/O performance on NT servers, sources said. The PCIX specification, code-named Project 1, is only the first in a series of planned steps to be taken by the group, in its collective effort to assume a more prominent role in X86-based server development, according to sources. Doing Intel one better PCIX takes the existing and aging PCI specification to new heights. PCI, which was developed by Intel (Nasdaq:INTC) in 1993, runs at a paltry 66MHz and transfers data between the CPU and peripherals at 132MB per second. As a result, current PCI-based I/O is a bottleneck for processing multiple applications. CIX will support bus speeds of up 133MHz, and it will transfer up to 1GB per second between the CPU and the peripherals. The new PCIX chip sets and third-party add-in components will be backward-compatible with the older PCI architecture. PCIX "fills the gap [between] current PCI and the next-generation I/O coming with Merced," said one executive involved in the development of the new spec. An announcement of PCIX should be forthcoming within 60 days, sources said, with servers and third-party products available in the first half of next year. Beyond the immediate performance improvements of PCIX, the symbolism of a cartel making architectural changes to the PC server without consulting Intel is significant. "The time is ripe for a grass-roots type of backlash," said the executive of one of the companies involved in PCIX. "We're trying to create an environment where Intel is an equal player in the technology, not the controller." Indeed, Intel's control has been a silent bone of contention among vendors with the know-how to overcome Wintel's scalability, availability and performance limitations. The rub has become particularly severe as Microsoft Corp. (Nasdaq:MSFT) and Intel try to move their architectures into the glass house. Those close to the group of PCIX vendors said the trio is determined not to make commodities of midrange and high-end NT servers, especially when they can differentiate their products from vendors, such as Dell Computer Corp. (Nasdaq:DELL), that buy most of their technology directly from Intel. "We want to be more involved in creating and setting the standards, but sometimes you get stuck with what they're offering," said an executive with a second company involved in the development. "Today's PCI is an example of that." The only option for new companies? For new companies coming into the Intel-compatible systems arena, breaking away from Intel's designs may be the only way to succeed. "Systems-oriented companies really become restricted the more Intel puts under the covers," said Gordon Bridge, CEO of Paradigm Computer Systems Inc., a Philadelphia startup developing eight-way X86 servers that are not based on Intel's chips, chip sets or motherboards. An Intel spokesman in Santa Clara, Calif., said the company's Enterprise Server Group is evaluating the PCIX spec but couldn't say when it would determine its level of support. Regarding PCI system design, the spokesman said Intel "has a long history of playing within the open standards of PCI." Sources said Intel received the proposed PCI specification from the trio in late August. The PCI Special Interest Group is also mulling the spec over and is expected to approve it as a standard. With or without Intel The sources also said Intel is expected to endorse PCIX and eventually build support for it into IA-32 chip sets. But even if it doesn't, IBM Microelectronics and Mylex Corp. will build supporting chip sets, they said. "It would give [third parties like Mylex] a competitive edge over Intel," said one of the executives. 3Com Corp. (Nasdaq:COMS) and Adaptec Inc. have also committed to building products based on the technology, and Microsoft has endorsed it, sources said. Despite Intel's broad system development efforts, there are cases where it has deferred. For example, the company will support Compaq's HotPlug PCI in the 450NX server chip set, due late this year or early next. HotPlug PCI lets administrators swap out PCI peripherals without having to take the system offline. Estimated pricing for systems based on the advanced I/O specification was not available, but one of the companies involved said it would be "self-defeating" to charge much of a premium over similarly configured Intel-based servers.