Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
pfosse - If I recall, it has to do with the focus of their purpose.
http://www.hpcwire.com/hpc/767114.html
The following IBM System x and BladeCenter products were introduced on Tuesday:
1. BladeCenter LS41 - Enterprise class scalable 2-way to 4-way blade; targeted for ERP, data marts, data warehouses, databases and HPC clusters.
2. BladeCenter LS21 - Enterprise class 2-way blade optimized for performance computing; targeted for financial services, scientific, high performance computing, databases.
3. System x3755 - For mid-market, large enterprise customers, targeted for scientific computing, such as weather simulations and crash test analysis.
4. System x3655 - Business performance server, targeted for database/ERP, business intelligence, IPTV and Video on Demand applications.
5.
System x3455 - High performance compute node, targeted for scientific and technical computing, database and Linux clusters.
wbmw - As if that's some surprise? AMD always goes down after the CC. It has held so far above the recent low, so it's looking pretty good to me. Tomorrow should be interesting, and I'll be on a bus to Atlanta / New Orleans to build some houses with Habitat for Humanity / World Changers for a week, so I won't have a clue what happens. :(
cg - You expected them to keep it open after killing Geode?
Keith,
it dropped about $0.25 in the last 5 mins of after-hours.
"Odd" - Indeed it is.
But Conroe and Woodcrest share the same shared cache architecture, don't they? It seems to be the main commonality.
CJ - How about the shared cache? If it were a MB/chipset problem, I'd think it would've showed up with Netburst chips, too.
Jules - Link?
Ah, found it:
http://www.dailytech.com/article.aspx?newsid=3169
BUGGI - Thanks for the word!
BUGGI - WOW! Very interesting!
SSE seems like a stretch to compare to what AMD has planned, more like their 3DNow!. The TCP/IP offload is more like it. But Intel is not about to open up to other firms co-processors like AMD is, and I haven't gotten the impression yet that they have anywhere near the kinds of plans for on-die coprocessors that AMD has. Maybe I've just missed the reports.
Tenchu - re: Mas, 32-bit Xeon was nowhere in the 64-bit space, PAE is not a serious solution.
"LOL, I guess no Xeon server ever shipped with more than 2 GB of memory before Opteron came around."
Oh, please, that's like saying "I guess no transAtlantic flights existed before Boeing introduced the 707."
Sure they existed, but, relatively, they were very slow and ineffective. A kludge.
sgolds, Tenchu - And AMD is doing what Intel would never do(or is at least supremely reluctant to do) in the realm of co-processors, both slot-type and eventually, on-die.
spaarky - Sorry, the chipset dog don't hunt. Much ado about nothing.
wbmw - That's not long.
wbmw - Because Intel won't hold a position of superiority for long. :)
wbmw - Don't worry, there'll be more, better opportunities. :)
DoU - Stability? Is that what you call frantically whacking off body parts ballast and fire sales?
wbmw - I didn't say that it would be up for the next 6 months necessarily, but it wouldn't greatly surprise me. There's more to stock success than having the highest-performing product. Y'all should remember all too well - just look at Intel's performance during the first year of Opteron's existence.
wbmw - re: holding - Yes indeedy.
I'm afraid the delusional ones are the ones who don't think AMD has an upside. :)
And STILL up 200% from when I got in. :)
chipguy - Which AMD products?
Gotcha, though I'm surprised Dell isn't cranking on weekends.
wbmw-??
7/13/2006 minus 12 days looks to be July 1st to me, not today.
cg - re: "Great call! "
We'll see, won't we?
Never occured to you that maybe Intel's had its day, eh? Not that I think it's going to disappear, but I think its dominant position is eroding fast, not likely to recur.
The sour grapes and condescension from the Intel'ers here are quite amusing considering their stated attitudes of Intel's superiority, and their continual droning about AMD's imminent defeat, if not demise.
It's a different world now. Get used to it. :)
Doug - Makes sense to me.
Doug - Looks more like 27 July.
http://www.theinquirer.net/?article=32518
wbmw - Didn't occur to you that it might be because AMD has a lot of demand from someone(or several someones) for primarily 2x 512K chips? Opterons will still be available with 1M, as will socket 939 X2s(for a while).
spokeshave - If you're still checking in here, have you heard any more about the ORNL project?
spokeshave - If you're still checking in here, have you heard any more about the ORNL project?
wbmw - Right, Dirk is a real neophyte like Luke was...
Not.
chipguy - re: "It suggests that AMD couldn't get cost
and power of its own Turion chip down far enough to sell
those to AMD instead of an externally sourced design that
Intel killed off once already. It is really kind of embarrassing
for AMD if you think about it. It also makes you wonder why
MS didn't take the design directly to UMC and cut out the
middle man. "
Not quite. Here's why:
http://www.reghardware.co.uk/2006/06/06/amd_to_sell_transmeta_efficeon/
...The programme is broader than the PIC, but clearly AMD reckons it's more cost-effective to offer chips based on Transmeta's FlexGo-ready technology than build FlexGo support into its own CPUs.
FlexGo equips PC suppliers with a software and hardware framework to support pre-pay and subscription computer usage models based on the way mobile phones are offered around the world. Instead of buying a machine outright, users can "borrow" the PC and buy processing time through pre-pay cards or a monthly subscription. The idea is to lower the cost of PC ownership and get more of the developing world and customers in emerging markets online.
mmoy - I missed that you were talking laptops. eom
CJ - Wouldn't it require entirely new paths for the pads, since the interface is 240 pins vs. 184 for DDR1? And that would mean new masks, at least at some levels?
gollem - I have to admit it sounds pretty unlikely; I don't think the DDR2 mem controller is backward-compatible to a DDR interface. Now, next time around, with DDR3, that may happen...
mmoy - Not much of a pain at all, happens all the time. They're always updating BIOS firmware to support new processors; some do better than others, but they generally want to be seen as supporting the latest possible stuff....
wbmw - Or it could be that AMD is doing the hard work now, and that it'll pay off with a more effective, longer-lived 65nm process. Kind of like their 90nm one.