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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Joe NYC who wrote (195227)	4/27/2006 12:00:29 AM
From: bobs10	Read Replies (1) \| Respond to of 275872

you... I think winning DARPA alone could lead to 5 bagger from current levels. Me(The other bob).. Yeah, I've owned Cray since last November, and I’ve added some a couple of times since then. Stock price moves around so if your patient you might be able to pick up a bargain. Definitely think DARPA is what could propel Cray into the stratosphere. Anyway, the company looks better than I can remember in a very long time. One other thing I like is that it fits nicely into my overall scheme of extensible X86 eventually taking over the mainframe market. Just because you and dougie(tecate term of endearment) no longer have a stake in AMD doesn't mean well be hearing less of you? Sorry to hear you’re out of AMD, but that's what makes a market. Personally, I feel like I'm in act III of a Wagnerian opera and I really want to hear the fat lady sing out in a lot of pain. In act IV we still have to find out if the king’s bastard son NGA is really legitimate and whether the white knight (RevF) saves the princess. I’ve paid for my seat and I’m not leaving early just to maybe beat the traffic.

To: Joe NYC who wrote (195227)	4/27/2006 7:20:27 AM
From: Rink	Read Replies (1) \| Respond to of 275872

re: Is Cray shipping (as far as current systems) anything other than Opteron based systems? They still have vector based X1E: cray.com BTW, the filing problem the NASDAQ made an issue about has been solved: "The Nasdaq Stock Market stating that the filing delinquency, for which Cray had requested a hearing, has been cured and the hearing has been cancelled as moot. " I also think that socket 939/940 FPGA's will be a problem for XD1 sales (as it commoditizes FPGA solutions, like is also happening to some extent HTX-based connectivity, it'll become less easy for Cray to differentiate their solution). In some aspects the market is moving quite a bit faster than Cray can (their design cycles are based on main frames while everyone else's cycles are based on mid range systems). Cray is reasonably save in high end, but a bit vulnarable in mid range and that situation is not going to change soon. Still winning DARPA funding is quite an event for Cray. re: XT3 currently uses some custom IBM connectivity parts (which I seem to recall were a source of some problems earlier). I wonder if Cray is going to dump these for pure HT based connectivity, potentially with some custom logic like Horus. Good question. Theoretically there are three possibilities: 1) More HT links per processor. As you pointed out previously the HT3.0 presentation suggests 'more interconnect links', see slide 7, here: hypertransport.org One of the questions is how much more. You need a minimum of 7 HT links for a 3d mesh (actually 6 for the mesh + 1 for the IO). That's 4 more than current situation. Sure that would make life easier for them as they don't have to pay IBM to upgrade the custom communication chip, but it's rather hard to predict that 2008 Opteron will really have 7 links. 2) Upgrade IBM communciations chip (this one already has 6 comm-to-comm chip links + 1 HT link). I don't know. The IBM communciation chip works and it might not require that much more money to upgrade it. 3) HT switch somewhat like you said, but I think it'll need to have way more HT ports that you mentioned. BTW, name of that DEC/Alpha data communications team wasn't API, was it? Regards, Rink