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.
mmoy - re: dual - You've seen this haven't you?
Paul
http://www.iwill.net/product_2.asp?p_id=36
Anand Sun 2100z review
http://www.anandtech.com/systems/showdoc.aspx?i=2255
wbmw, perhaps you ought to try checking only 802.11b or leaving it at the "Show All" option.
5 products listed.
wbmw - re: "By the way, where is Cray on the next generation of custom ASIC to support HTT 2.0? Are they working in parallel with AMD to deliver a TTM solution? I think not."
Why would you think this would be so incredibly difficult?
Paul
mmoy - I haven't used clockgen much(CrystalCPUID more), but those are certainly options that can be helpful, though I don't know if the voltage control is as helpful with SOI K8s as it has been with K7s in the past.
The multiplier can be hugely helpful if the chipset doesn't lock the PCI or AGP clocks. If you push those too far, the system will be unstable, and if that happens before you've reached the max capability of your CPU, then you've left some performance on the table.
So try pushing the system and running Prime95 in "Torture Test" mode with the Priority Level (in the app) set to 10. That usually will quickly pull out any instabilities. Don't expect to do anything else with the system while it's running, though.
Though some like to run it for 24 hrs to ascertain "complete stability", usually if the system has any substantive instability, an error is going to come up within the first hour.
wbmw - Not that you may not be right, but that link doesn't prove it at all. It just lists companies who've attained certification in the past quarter.
"Companies obtaining Wi-Fi product certification in
the past quarter include:"
Cray begins shipping XT3s.
V
V
V
V
http://investors.cray.com/phoenix.zhtml?c=98390&p=irol-newsArticle&ID=634922&highlight=
MSDN W64 presentations
http://www.microsoft.com/seminar/events/series/msdn64bitwin.mspx
Smallpops - Thanks. eom
Does anyone know whose flash is being used in these?
http://www.bitmicro.com/press_news_releases_20041026.php
Keith - Yeah, Josh is a good guy and has come out with some very good stuff in the past. It's been a while since I've been by there, he had some serious family issues going on for a while, if I recall.
Paul
yb - Don't you think that a hotswappable memory feature might be something that HORUS type systems could eventually support? Might make the Opterons directly connected memory like a huge L3 cache, eh? I have no idea if that's workable, just daydreaming. But I do imagine it could be something that Bechtolsheim et al could be working on.
Paul
wbmw - Why on earth would AMD want to go to Broadcom for a Centrino-type wireless solution, when, if I recall correctly, they have a solution of their own that is at least as good as Centrino, if not better? And I can't imagine it would cost any more than what Broadcom or Intel charge for theirs. Sorry, I know Centrino has been very successful for Intel, but it's marketing BS. You can get a Pentium M system properly configured that's just as good for less $.
Paul
yb - There's still 10 years for the infrastructure to expand, and for the system to be made more inexpensively. As long as they don't do it for a loss, it's not likely to hurt AMD, and may well bring a lot of folks into the digital age who wouldn't have otherwise; and of those, some will certainly become more substantial AMD customers. Expanding the potential market can be a wise strategic move. It remains to be seen if it's executed effectively.
Paul
rlweitz & neye_eve - on nForce2 systems that are not true "SoundStorm" systems, they are missing several jacks, so they have to share/multitask some of them.
Paul
mmoy - that's a great idea!
NaS - Just to specify DARBES' reply a little, it had to do with provision of interenet access to the world - 50% by 2015.
neye_eve - light blue 1/8" jack = Line In
rupert - re E Stroligo - entirely possible and no.
Bet I know whose flash is used for BIOS! :)
Does anyone know how long WinCE usually takes to boot up on a HD-based system? Maybe they're using more/larger flash for a boot image?
Paul
smooth2o - Sorry, I figure the content of the post would make the smiley unnecessary! ;)
Paul
NaS - MCP was what they called their nForce "southbridge", though now that it's a single chip solution and there is no "northbridge" due to the ondie mem controller, I guess they carried it over to the single chip "chipset".
Paul
mas - good point.
Keith - Thanks, I was thinking that the 89W applied to the entire 130nm K8 family, but if the docs say otherwise....
Paul
Apparently they weren't planning on FX-55 on 130nm, since at 104W, it's the only FX that doesn't fit in that 89W envelope. Looks like they're beginning to push the limits of that process?
Paul
Intel Digital TV - killed by their need for fabs to make all those monster cache CPUs?
Paul
I-banker - I'd say the primary concern many might have is the ramp and yield of 90nm; but as of now, I don't think there's enough data to substantiate any great concern about that. So far, so good.
Paul
I_banker - Sorry, McCoy is an AMD exec who apparently sold some shares yesterday. Well, maybe it was exercised options and sold, I don't recall. I thought it was discussed here, but it might have been on SI.
Paul
Looks like McCoy should've waited at least another day to sell, eh?
aixman - Not exactly. All the lower grade Semprons are Socket A. They'll be around for a while yet.
Paul
aixman - It may be an expense issue, but I suspect it more likely had something to do with the different interface to shared memory. The IGP chipset was very successful, and they're still being sold.
Another anecdotal data point I'm on a machine with nF2 IGP right now(although I've recently added a discrete video card), and not only does it perform flawlessly, I can run it stably at 203/406MHz FSB/memory clock for any gaming or benchmarking,
http://service.futuremark.com/compare?2k1=7277759
and it's not even an Ultra, it's only technically rated for 166/333MHz using IGP.
Paul
aixman - nVidia already had the name for performance with K7 chipsets and IGP. IF they want to sell lots for lots of money it seems very odd for them to drop IGP, and therefore, the business sector. Unless they've decided to pass on the business desktop in favor of workstations.
Is there enough workstation volume for that to work? I don't know. It would be better for AMD if nVidia produced a solid IGP chipset for business desktops as well, but I imagine others will step into the void if there's demand.
Paul
For whatever reason, nVidia seems to be focussing on the performance market.
Paul
NaS - No on nF3 IGP. nF4? Who knows? Not yet. They seem to be ceding that market to Via & SiS for some reason.
Paul
Keith - They're probably right, but they may be looking at the wrong reason. The maturing of K8(Socket 939) is behind a lot of it.
Paul
rupert - Did you forget your Secret Decoder Ring Invisible Ink Exposer today? ;)
Paul
Ixse - Desktop -no. I believe there is a K8 mobile Sempron 3000+, though.
Excellent! M$ will not charge extra for dual core licenses!
Paul
http://www.theinquirer.net/?article=19165
CJ- Agreed. I thought had already made that point, just that Sanders' time was now past, but he set up what Hector is now moving forward in a different way.
Paul
dual core 90nm image
http://www.hkepc.com/hwdb/a644000fx55-3.htm
V
V
V
V
V