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.
EP - for heaven's sake! Has someone been p**ing in your Cheerios lately? What's with the constant ragging on everything involving AMD and those related to it? Lighten up!
Paul
Just goes to show what you get when someone tries to speak authoritatively with only a surface knowledge of the subject. You could say that on the surface there seem to be many truths, but when you dig deeper into the specifics you find that you can't take things at face value. I've rarely seen such a collection of "facts" gathered into such a ball of "wrongness".
Paul
SemiconEng, I'm not saying this problem exists or not, but speaking theoretically, if the problem is in the strained silicon part, would that be considered a process or design problem, or would that depend on exactly what the issue was?
Paul
neye_eve - OT - I don't know if you're still working on the PC, but if you are, it would be very helpful to know the details/specs on both 1) all the new parts and 2) all the old parts you are using to swap, as well as the PSU specs. If you want to PM me feel free. Oops, I just thought about the fact that if you're an ihub freeloader like me, you can't. If so, feel free to e-mail me. Don't know for sure if I can help, but I'll try.
Paul
avatar - 1)NIH 2)Intel has none that will work there yet.
Paul
SemiconEng(&chipguy) - so are you saying the problem is in the design or in the process?
Paul
As we creep closer to it, how much of a psychological effect do you think it will have on the markets when the DOW and NASDAQ clear 10000 and 2000 again?
Paul
DDB, they certainly seem to like it. But substantial money from AMD? I don't know about that. I think they just like to be "on the winning team". They loved Athlon when it was wiping up the floor with P3(and for a while P4), and ragged on Intel. When P4 took the lead, they loved it and ragged on AMD. Now the tables are turning again. Tom's is kind of like a power CPU "groupie".
Paul
Ed S. may have something about the Prescott delay.
Paul
http://www.overclockers.com/tips00480/
PS. If you look at the front page vrzone.com story, rather than the page Ed linked to, you'll see the basis for calling it a Celeron.
EP, thanks, I'll have to read it later.eom
More on that SiS chipset(755/760). Notice the silkscreening under the "northbridge". Local frame buffer RAM chips, I'll wager. Looks like 755 and 760 must be pin compatible other than the video parts. If they use fast enough RAM, it could be interesting. If they also allow a discrete card to use the same interface and RAM, it could get really interesting.
DDB, Tom's getting slow in his old age. Ace's did a review of this system almost a month ago(9/25/03).
http://www.aceshardware.com/read.jsp?id=60000268
Gofbum, it'll be the "same" die with sections disabled.
Looks like SiS will be the first to do what some had postulated might happen(local video frame buffer) for K8 boards. Also used internal AGP interface at 3.2GB/s to match nicely with DDR400.
Paul
http://www.sis.com/products/chipsets/integrated/athlon64/760.htm
http://www.sis.com/products/chipsets/ultraagp.htm
EP - the same reason that you so frequently do. Taking what information we do know and hear, and trying to make an educated guess at the real situation.
Paul
EP, of course he won't - he doesn't know and neither do you.
EP - Yes I know what you're saying; after all, they'd come from the same wafer. Just playing a little "devil's advocate".
Paul
EP - LOL! However, your examples are not functional differences, which disabled links and controllers would be.
Paul
EP, I guess it would depend on how you define different and on how they disable the links and channel(permanent or reversible). If it's permanent, then once they're disabled they are essentially different dice.
Paul
block of 122,100 shares just bought.
Wow! AMD almost as active as Intel. eom
Wow! Way better than I'd feared, if not what I might have hoped for(pie in the sky). Q4 should be interesting!
Paul
chipguy, unambiguous as far as which processor it corresponds to, yes; not at all clear as to exactly what that figure represents.
Paul
So, what do we know about Randy Allen?
From that news.com K9 article:
"Chips based on the K9 architecture will likely be released--at least in sample quantities--by the second half of 2005, Weber said. The project is being overseen by AMD engineer Randy Allen."
His title on the website is Vice President, Design Engineering, Computation Products Group. Does he have a solid record?
Paul
chipguy, your statment should shed a whole new light on the discussion about the 89W Opteron figure that some here argued so vehemently could not possibly be a platform figure, but that it had to be the figure for the fastest CPU at release.
Paul
Intel's news bringing strong opening?($14+)eom
Any thoughts, if AMD's CPU ASPs are dropping, why the prices on pricewatch have been steady to up almost across the board recently for XPs? Since this is pricewatch/bizrate pricing, not AMD, that would seem to run counter to the whole supply and demand dynamic if their ASPs are dropping. Or have they become production-limited on XP due to K8 switchover? Did they have a monster sale to OEM(s) that dropped ASPs some and cleared out inventory, which left less product for pricewatch type outlets?
Paul
http://www.anandtech.com/guides/showdoc.html?i=1903
Keith, re: falling flash ASPs
Doesn´t bode well, ASPs seem to continue to fall.
Didn't they raise prices in the 2nd quarter? If so, falling prices could just mean coming back to reality, which hopefully wouldn't affect AMD much if they took advantage of the artificially high prices Intel tried to pull off and gained marketshare.
Paul
Keith, the deal makes a lot more sense in that light. eom
Packard Bell A64 notebook specs. See if the INQ got it right this time!
http://www.theinquirer.net/?article=12113
Golfbum - taken at face value, I'd agree it seems unwise, but maybe there's more going on than meets the eye. Only The Shadow knows....
Paul
wbmw, that's Intel's fault for "introducing" it so long before availability. Can you recall any other Intel chip introduced and launched with such a long gap between the two? I seem to recall AMD being raked over the coals for the 2800+ - the only difference was they either lied or were mistaken about when they would be able to intro that chip. Intel's EE intro/launch strikes me as perhaps even more cynical.
Paul
Keith, if there's really a shortage, at least the impact should be minimized. Sure doesn't seem like Intel's approach to the flash market has been very well thought out lately.
Paul
No A64 Ferrari, INQ jumped to conclusions. AXP2500+. Still quite a nice little "hot rod".
Paul
http://www.psc-sb.de/psc-sb/acer_ferrari3000.html
Little GTO? ("Ferrari")
A64 speed demon notebook?
Paul
http://www.theinquirer.net/?article=12079
sgolds, re 90nm
(Look for comments on the 90nm conversion to be perhaps the most important piece of information from the CC next week.)
Agreed!
Paul
Dan3, Perhaps I should ask, where did the 1.8GHz figure come into play?
Paul
yb, I was being ironic, I have no expectation of Dell coming on board any time soon. :)
Paul
There will be at least one really good 940 board for FX before A64 FX goes the way of the dinosaur.
Paul
http://www.anandtech.com/mb/showdoc.html?i=1901&p=12
Final Words
The Gigabyte K8NNXP-940 is an impressive motherboard. The performance it exhibits at standard settings in our benchmarks is simply outstanding. The early impressions of the Athlon64 FX51 were that it was not much faster than Athlon64. Perhaps we should recast this to say that the early Athlon64 FX motherboards were not much faster than the Athlon64 motherboards running Athlon64 processors. In almost every benchmark, the Gigabyte K8NNXP-940 with an Athlon64 FX51 tops our fastest Pentium 4 motherboard — the Asus P4C800-E running a 3.2GHz processor. Even the usual P4 stronghold of Content Creation sees the Athlon64 FX51 and 3.2GHz P4 in a dead heat. In almost every other benchmark, the Athlon64 FX51 is a clear winner — over both the Pentium 4 3.2GHz and the Athlon64 3200+. The only exception here is our Media Encoding benchmark, which tends to favor Intel processors and will soon be replaced with an updated benchmark.
The features, layout, and included accessories make the Gigabyte K8NNXP-940 a stellar home for a new FX CPU. Gigabyte has done an outstanding job in the engineering of this Socket 940 motherboard, and in standard testing, it is in every way a winning combination with the Athlon64 FX51. The addition of Multipliers to the release K8NNXP motherboard just sweetens the picture that much more. With the high cost of the FX51, Enthusiasts will need some persuading to pay the price for the FX processor. AMD and Asus were smart to make it clear that the FX51 could be unlocked in the BIOS, unlike regular, cheaper, single-channel Athlon64 chips. Gigabyte becomes the second manufacturer to add the ability to manipulate FX ratios in the BIOS.
With this high praise, we do believe Gigabyte has more work to do on the BIOS of the K8NNXP-940. It is certainly fast and full of options in the release F1 version, but it is disappointing to find that the new CPU ratios do not work reliably, and the FSB often exhibits some strange behaviors — resetting itself to lower values on boot and we don't just mean default 200. We had two cases where settings of 220 came back at boot as 209.5, and the 12 multiplier only worked at 200 FSB setting.
Despite the issues with the immature BIOS, we still highly recommend the Gigabyte K8NNXP-940. It can take the Enthusiast to new performance heights as it is, and we fully expect even better overclocking with some of the BIOS kinks worked out. It performs better than any Socket 940 board that we have tested so far.
We have been told that we will see at least a couple of new and revised Athlon64 FX motherboards in the near future. They may prove to be even better, in particular, an update of the VIA chipset that is said to have new features for that chipset. However, until we test those Socket 940 boards and perhaps find one that performs even better, the Gigabyte K8NNXP-940 is king of the hill.
Shock! Amazement! Dell won't use Opteron! ;)
Paul
http://www.theinquirer.net/?article=12030