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.
yb - Better look again. To get to 3.6, they boosted Vcore to 1.952V. They do look to have gotten to 3.5 with no boost in Vcore, though. :)
Paul
PS. They haven't even run the benchmarks @ 3.6 yet, they're waiting for a MB they like more.
Tenchu - Why do you think AMD has shifted focus to servers? Precisely because they can bring performance that does matter to the table. The "rules" are changing in that arena.
I'd have to beg to differ with you on AMD's winning mindshare and marketshare in that arena. It's doing very well and will do much better in the future. Majority status anytime soon? No, but they're going to make money like they never have before!
Paul
Golfbum - I'd think it'd be more the OEM rather than a bank that'd be concerned with that.
Paul
Cray dual supercomp wins.
V
V
V
V
V
V
http://news.com.com/Cray+signs+two+supercomputing+customers/2110-7337_3-5392983.html?tag=nefd.hed
I_banker - we can only remove, not edit. eom
I_banker - One of the problems is that it has taken so long for IPF to "mature" that much of the evidence is probably long gone. Internet caching/archiving wasn't nearly so sophisticated back then as it is now.
Paul
fpg - very cute! ROFL!! ;)
Paul
wbmw - I'd have to agree. Though I'm sure Intel has given away some IPF CPUs, Intel's entire culture is against taking a loss(unless they believe they can do it short term to annihilate opposition and capture a market - but I don't think you can describe much about IPF as "short-term"). I'd have to have hard evidence to believe they're giving away very many.
Paul
chipguy - re: "Good news for AMD but Sun employees should take
cold comfort because thousands (more) of them will
have to get the boot and hundreds of millions per
year in R&D will have to be slashed to make Sun's
new pseudo-commodity hardware business model
work.
"
Not if Sun sells enough of their new systems. Time will tell. I suspect the R&D isn't necessarily being slashed, but re-directed.
Keith - It appears AMD did their homework with their mem controller and the memory makers are taking advantage.
Paul
Celeron D vs. Sempron Comparison
http://www.xbitlabs.com/articles/cpu/display/sempron.html
New ATi Linux driver.
http://www.theinquirer.net/?article=18816
chipguy - perhaps you might want to offer the basis for your side of the issue as well, if you're going to challenge them for making baseless claims. "I suspect" isn't any better of a justification than what they offered.
Paul
l_banker- the SI and ihub boards have over time tended to work in a synergistic manner; frequently discussions on one bleed over to the other, and result in fuller discussion/understanding. There are a couple of reasons it comes about, some of which Keith mentioned. There should eventually come to be less of it now, with the mutual free accounts, except in the extreme cases like Doug. For now, just enjoy the benefits or skip the posts, they're usually clearly marked in the first words of the post on both boards. ;)
Paul
PS. BTW, is your username I_banker(as in Investment banker) or lower case "L" banker? The displayed font makes it impossible to tell.
Grimes - I don't think they are, either, but it can hardly be re-iterated too often(unless it's allowed to become a meaningless, habitual mantra within the company).
Paul
Interesting perspective from Arron Rouse at the Inquirer.
http://www.theinquirer.net/?article=18733
I think the more important thing for AMD to establish is their ability to reliably put out large volumes of their superior product, but greater performance would be nice, too. Complacency is not to be allowed, that's for sure.
Paul
chipguy, maybe I was imprecise. I was referring to the temperature, not the leakage, which was also being discussed.
Paul
chipguy, Wouldn't you think that the wide margin in clock rate between Banias/Dothan/Itanium and P4 would have something to do with that?
Indeed, those gamers knowledgeable enough to discern throttling are very likely to find it unacceptable to allow their CPU to get hot enough for it to be more than a rare issue.
About the only folks I can think of who would regularly run into it are folks like me who are involved in distributed computing projects like UD/Grid Cancer Research, Folding@Home, SETI, BOINC, etc., because many of us run our systems at 100% CPU utilization 24/7. If our systems were to throttle under high temp conditions, the performance drop would become apparent,though it might take a few days to verify. Mine generally runs 24/7 at 53C±1C CPU temp, some run theirs even hotter. Now Prescotts probably run considerably "warmer". ;)
When gaming, my system always runs cooler, because UD only runs idle CPU cycles, so any app with higher than idle priority gets its needs fulfilled first. The only things that run it hotter that I know of would be something like BurnK7.
Paul
Keith - VeriSign is an important one. eom
Doesn't sound much like a downturn from here:
Most Taiwan first-tier mobo makers running at full capacity
http://www.digitimes.com/mobos/a20040927A1001.html
Emma Wang, Taipei; Steve Shen, DigiTimes.com [Monday 27 September 2004]
Taiwan-based first-tier motherboard makers, including Asustek Computer, Gigabyte Technology and Micro-Star International (MSI), are currently running at full capacity, while Elitegroup Computer Systems (ECS) is operating at more than 90% of capacity, according to sources at the makers.
Buoyed by increased orders from OEMs and the channel, the four makers said...
Paul
wbmw - I have to admit that, though technically correct if you include the text of the article, the headline is somewhat misleading. Selling papers, I guess?
Paul
wbmw - Perhaps their probe was being hit by the fan(s)? ;)
Paul
Over here in North Carolina, it's beautiful(mostly sunny and 78F/25C) in advance of the rains soon to come(tomorrow) from Hurricane Jeanne, which will likely only be a "storm" by the time it gets here over land.
Paul
conductor - Sorry, we aren't around 24/7; it takes a little while sometimes. With your permission, I will excise your responses to that little sub-thread as well to clean up the mess.
Paul
kpf - the Max die temp change may just mean that right now they're pushing 90nm closer to its top clock than 130nm,which is not unreasonable with a new process. This will change with time as the process matures.
Paul
Yuri, if no other steps are taken to speed or ease the transfer of the heat out of the chip, then, yes the temperature will tend to be higher for a given power/heat output.
Think of the water that can flow out of a small hole in a bucket, with the water representing the heat to be dissipated and the hole in the bucket representing the surface area of the die. IF you decrease the size of the hole, the water drains out of the bucket more slowly, so it stays full(hot)longer. And if there's a constant inflow of water(heat production), then the greater the rate of water(heat) flowing out of the bucket, the less water(heat) will remain in the bucket. In other words the chip will be cooler. This assumes, course that the rest of the cooling system is equivalent.
Here's a link to one of the cool things going on in the interface field:
http://www.rntfoil.com/
Where I find its possible use to be particularly compelling is in the die surface to heat spreader interface.
Paul
UnD - re: "That is a huge difference in temperatures!"
I don't know that it's such a huge surprise. There is similar power/heat output having to be dissipated through a much smaller die surface. I hope as they develop 90nm and especially 65 and beyond, they develop/perfect vastly superior interfaces between the core and the heat spreader.
Paul
Yuri - greg s was the moderator on the Intel forum here at ihub, and has tended to be rather inflammatory when he visits this forum. He is behaving now, which is very welcome. Look at his profile and go back a ways in his posts and you'll see what we mean.
Paul
BTW, Yuri, In case no one else has done it yet, welcome to the forum! :)
Paul
chipguy - Not if AMD feels what they're getting is a good deal for what they're paying/giving.
Paul
mmoy, I'm not sure, it's new and I haven't researched it, but I thought it might be of interest to y'all who have K8s to pursue.
Paul
A64 CPU multiplier utility.
http://www.digit-life.com/news.html?108546
sgolds - Thanks! eom
sgolds - Website? Link, please? ;) eom
BUGGI, NaS - newegg usually lists the process size in the product description.
Keith - re BB A64 desktops
Something very odd going on. 2 HP models come up in search, but the pages no longer exist.
V
V
V
V
http://www.bestbuy.com/site/olspage.jsp?id=1083714106418&skuId=6714503&type=product
http://www.bestbuy.com/site/olspage.jsp?id=1083715205863&skuId=6720471&type=product
All this while they are actually marketing the Athlon 64!
http://www.bestbuy.com/olspage.jsp?id=pcmcat18700050010&type=category
http://www.bestbuy.com/olspage.jsp?id=pcmcat22300050001&type=category
http://www.bestbuy.com/olspage.jsp?id=cat12077&categoryRep=cat01000&guideID=1062562440283&am...
Paul
chipguy - Perhaps IBM is just arrogant enough to think that, regardless what AMD has said about staying within the socket power envelope with dual-core, that since IBM has been toasting power limits with their dual-core, then of course it follows that AMD will have to.
Paul
Keith - Fortunately, ATi is about to come out with the RS480 chipset which has a nice DX9 core, and since the mem controller is now on the CPU, it pretty much removes my reservations about ATi chipsets. Apparently it has been less problematic than the Intel-supporting RS400, too.
Paul
Niagara at JS blog.
Paul
V
V
V
http://blogs.sun.com/roller/page/jonathan/20040910#the_difference_between_humans_and