There's a big difference between studying and assigning a full development team...
And then there's the time-to-market...for example, if Intel were to release an HT capable chip next year, they would have had to have started development at least 3 years ago.
The question is, how could they have known the specs on cHT 3.0 back then?
zsdersw
Lifer
- Oct 29, 2003
- 10,505
- 2
- 0
They're shooting ahead into the future with CSI.. so if they want in on HyperTransport instead, they can shoot ahead into the future there as well. And come on.. you don't think Intel's engineers can forsee what the future improvements to HyperTransport will include?
The point is that Intel wouldn't necessarily be behind if they chose to adopt HyperTransport.
The only thing that worries me about CSI is the delays...it was supposed to have been out 05/06, and now they are expecting 2009. I know they shuffled their development teams around so that they could get C2D out quicker, but that is quite a long delay.
As to foreseeing the future of HT, I don't think even AMD's engineers can do that (and they invented it!). I imagaine that if Intel does assign a full development team to it, they can certainly get within 6 months of AMD.
But remember that 5 year timeframe...Intel must first make the decision to go HT, or go through the rather large expense of running an HT paralell development team.
I don't think anyone here thinks that Intel will reign supreme on the PC market forever.
the question here is will AMD regain the performance crown.
AMD could find themselves in real trouble soon. If their market share falls precipitously in the next year or so because conroe is super successful, and K8L is delayed or shows up with unimpressive performance, combined with the huge debt AMD will carry from the purchase of ATI (it's something like 2 BILLION in loans in addition to cash and stock), then AMD could go under (or more likely be bought out by someone else [like apple or some other company that wanted to get in processor manufacturing]). In that case, it would be another company that eventually make a fantastic product and reclaimed the performance crown from Intel.
having invented all of those fantastic speculations, i think that is all very very unlikely. AMD has a pretty comfortably entrenched position in the PC market (they are up to like 15% now), and it would take a huge string of business disasters to ruin the company, but it certainly COULD happen.
The answer is simple and undenaible.
A loud, resounding Yes.
Of course they will.. AMD does better with performance chips historically than Intel ever has. The K6-2 might have kinda sucked, but it wasnt THAT bad and offered a very good price..
From the slotA Athlon to the current X2s and up till the Conroe launch.. AMD was pretty much superior the whole time in the performance realm.
To think they wont take it back again is insane. And I'll bet against you.. by putting my money into AMD stock ANYDAY.
Those boys at AMD are NOT dead. And they arent any less of a threat now than they were before.
As an aside, I dont see Conroe being as successful as Intel probably thinks it will be simply because they have a marginally faster product out now after years of being in disgrace.
Its a great deal for us, as we get cheaper chips.
A loud, resounding Yes.
Of course they will.. AMD does better with performance chips historically than Intel ever has. The K6-2 might have kinda sucked, but it wasnt THAT bad and offered a very good price..
From the slotA Athlon to the current X2s and up till the Conroe launch.. AMD was pretty much superior the whole time in the performance realm.
To think they wont take it back again is insane. And I'll bet against you.. by putting my money into AMD stock ANYDAY.
Those boys at AMD are NOT dead. And they arent any less of a threat now than they were before.
As an aside, I dont see Conroe being as successful as Intel probably thinks it will be simply because they have a marginally faster product out now after years of being in disgrace.
Its a great deal for us, as we get cheaper chips.
zsdersw
Lifer
- Oct 29, 2003
- 10,505
- 2
- 0
The future of HT isn't all that much of a mystery, especially for companies like AMD and Intel. With the kind of knowledge about the technology that both companies likely have, it's not too hard to figure out what areas there are to improve upon and what improvements are likely to work and which ones aren't.
Of course Intel would have to make the decision to develop HT.. but that was never the point, nor was it in dispute.
Anything COULD happen, but the odds are infinitesmally small (about the same as there being a sudden flaw found in C2D and Intel needs to recall the whole line for 2 years...).
AMD is still gaining marketshare, and will continue to do so across all lines through at LEAST the rest of this year (remember that Dell is boosting AMD's sales by 16-17% starting this quarter).
Ummm...AMD was at ~23% last quarter, and analysts expect them to hit 30% by years end...
- Aug 22, 2001
- 29,452
- 24,133
- 146
Monopolies are constantly having their chops busted by goverments, hence Intc needs AMD to a certain extent. That alone will likely ensure AMD's survival. While the evidence is weak, one could speculate that many of the current happenings, including DEll and AMD finally doing business, is more than just a matter of product and supply? </this morning's conspiracy theory>
in about 6mos AMD will regain, then intc will regain. its the revolving door of procs, been happening since ~1999
I'd say 6 months is too little time. While AMD COULD match the X6800 in 6 months (if the 65nm shrink tames the K8's power consumption, and that's still a bit doubtful), Intel can always increase its clockspeeds. Yes, the mainstream line would, most likely, be limited to an E6800 (because of power consumption) but an X7000 (3.46GHz) would be more than AMD can deal with without a significant core revision (aka K8L, and I doubt this one will match Core's INT performance anyway).
Agreed. Intel can increase the clockspeed or FSB (or both) if it feels threatened. If 3.46GHz is a stretch (I doubt it though, considering current overclocks) it can do a 3.33GHz/1333FSB part instead for similar levels of performance. It would take a 4GHz+ K8 to match that, and I just don't see it happening.
It's K8L or nothing for AMD.
you are right. I was pulling that 15% out of my head without double checking the numbers, and I would rather be wrong and too low than wrong and too high.
in any event, a higher market share just improves my argument, AMD is no longer a fringe competitor hanging on to the market by a thread (a la Apple in the mid to late 90's). they are now pretty firmly established as a major player. but i don't think AMD is sailing along without worries. If K8L is delayed or unsuccessful, AND the ATI merger turns into a huge cash sink for them, times could get very rough. i think that is pretty unlikely, and by far the most likely scenario is that AMD stays steady or slightly increases their share over the next 12 months (lose some ground to conroe, gain some ground from Dell etc etc) and then introduces K8L which makes things interesting in the CPU market, even if it doesn't take the performance crown back outright.
AMD's last meaningful gain in unit market share % was Q4'05 (this year it's remained approx. flat).
ATI could be a huge ball and chain around AMD's ankle. It's costing $2.5B and that's on top of $5B AMD needs to pour into capital spending. (Dresden 65nm conversions, NY Fab, etc). I also wonder how committed ATI's employees will be since they've essentially been cashed out.
AMD needs to be a platform company and ATI could help get them there but a merger of this magnitude could compromise the focus they need to regain CPU leadership and to improve their balance sheet. They could eventually come out much stronger or it could drain them dry. (mega mergers are usually not that successful) It remains to be seen if AMD will continue to let ATI embrace crossfire on Intel architecture. Either they support Intel or a major source of revenue goes "poof".
In addition Intel's new C2D gives them room to fight at the medium and low end yet that's the only place AMD plays in desktops these days. AMD cash cow is servers but that's also eroding. A price war will really hurt AMD and they don't even have the benefits of 65nm until late '06/early '07.
I have a lot of confidence in Hector. However he has some major challenges.
ATI could be a huge ball and chain around AMD's ankle. It's costing $2.5B and that's on top of $5B AMD needs to pour into capital spending. (Dresden 65nm conversions, NY Fab, etc). I also wonder how committed ATI's employees will be since they've essentially been cashed out.
AMD needs to be a platform company and ATI could help get them there but a merger of this magnitude could compromise the focus they need to regain CPU leadership and to improve their balance sheet. They could eventually come out much stronger or it could drain them dry. (mega mergers are usually not that successful) It remains to be seen if AMD will continue to let ATI embrace crossfire on Intel architecture. Either they support Intel or a major source of revenue goes "poof".
In addition Intel's new C2D gives them room to fight at the medium and low end yet that's the only place AMD plays in desktops these days. AMD cash cow is servers but that's also eroding. A price war will really hurt AMD and they don't even have the benefits of 65nm until late '06/early '07.
I have a lot of confidence in Hector. However he has some major challenges.
LOL I get a kick out of Intel supporters getting their panties in a bunch.
Its a great chip.. but just cuz you are hyped on it like crack, and now use one.. doesnt make it the choice of the Gods for CPUs. Theres a thing called price/performance that decides all and that hasnt really been laid out yet with relatively poor Conroe availability and probably not having seen the extent of the AMD pricecuts.
Basically now, Intel has something that isnt a disgrace is all.
Now, whether you go AMD X2 or C2D.. you are getting a decent chip.
Kudos Intel. Bout time much? I mean.. thanks, I think. But it only took 5 years.
3 years, my friend. Yeah, it's kind of disappointing that the 4MB Core 2s are so rare (excepting the X6800... thanks Intel) but things should get better... eventually. I will say that the difference is more than just marginal but Intel is aiming this chip to the mid/high-end (most likely to boost its battered ASPs) so AMD still has a lot going for it.
When AMD is beating Intel by 10%, Intel is a 'disgrace'.
But when Intel beats AMD by 20%, it is 'marginally faster'.
Admit it Crusader, Conroe beats A64 in far more benchmarks than it ever did against P4, and at larger margins.
I therefore hand you the 'Sharikou Award'. Hehe nice work spinmeister.
AMD (for the short to medium term) have pretty much returned to the latter Athlon XP days, are only competitive on the budget front, as you move into the mid range it would be foolish to get an X2 over Conroe, and I don't think 65nm will do all that much to change that.
But when Intel beats AMD by 20%, it is 'marginally faster'.
Admit it Crusader, Conroe beats A64 in far more benchmarks than it ever did against P4, and at larger margins.
I therefore hand you the 'Sharikou Award'. Hehe nice work spinmeister.
AMD (for the short to medium term) have pretty much returned to the latter Athlon XP days, are only competitive on the budget front, as you move into the mid range it would be foolish to get an X2 over Conroe, and I don't think 65nm will do all that much to change that.
zsdersw
Lifer
- Oct 29, 2003
- 10,505
- 2
- 0
That's fine, but when he (and others) say "well, it took Intel long enough.. 5 years" they have to realize that you're not gonna get a totally new chip design in significantly less than 5 years.. from either Intel or AMD.
If you want new chip designs in less than 5 years, hey.. go right ahead and try to do it yourself.
What a great idea, let me start up my fab =)
Seriously, though, AMD should have something far enough along in development to be able to counter Conroe in the next year-and-a-half to two years. Process technologies is what AMD needs to focus on since that is what held it back during the K8's reign.
Screw microchips - did you know that a rat's brain can be wired up to a parallel port and trained to use a flight simulator?
RESPECT.
I for one welcome our new rat/F16 overlords.
RESPECT.
I for one welcome our new rat/F16 overlords.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 20K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 12K
-
Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)- Started by Vattila
- Replies: 13K
-
Facebook
Twitter