smooth

it does seem to have parallels to the rdram debacle but there are some important differences. first is that there are no changes to the drams on the board. just the addition of an interface chip.

second is that the dram companies don't have to pay rambus (more than they're paying already through lawsuit judgements.)

the problem with rambus (the main one at least) was that rambus didn't make drams so didn't need to "swap" patents with the dram manufacturers. rambus wanted money. the dram manufacturers usually avoid patent issues through cross licensing rather than monetary exchange.

fbdimm is coming at a cost of the interface chip but holds out the promise of memory tech indepence. it will be interesting to see how it works out.

too early to call in my view...

gb

Should be fun to see if AMD actually gets a quad out other than a sample.

gb

What sort of system does Clovertown plug into? Can it do a dual socket? If so that would make it 8 way capable and getting into pretty rarified territory volumewise.

I still don't understand the Dell/AMD thing if that is the case.

gb

Note that these are merely sorts on the existing product with price premiums. It's been my long experience (now retired) that in the desktop arena no one pays a premium for this kind of feature.

That makes this a purely hollow PR stunt with minimal real shipment impact.

gb

i listened to the meeting and revenue upside was unclear. i think by year end both viiv and vpro will be generating upside if not now.

both have intel software content not available any other way...and certainly not from "live."

both will also create a defensible position through a combination of hardware and software.

in the case of viiv it's arguable that it will be revenue upside. in the case of vpro it will be firewalling revenue behind a defensible brand and likely increasing corp mss.

gb

corbett's on the job!

ain't just a bucket a parts

gb

worthwhile viiv presentation:

http://www.corporate-ir.net/ireye/confLobby.zhtml?ticker=INTC&item_id=1219336

if the direct link doesn't work then go to the analyst day presentation and look for the morgan stanley presentation by kevin corbett.

gb

care to offer another interpretation?

gb

since it always makes sense to run fabs at full capacity and since amd was not "capacity constrained", what prevented them from fully utilizing capacity with their "cost advantage" to take more market share.

gb

not capacity constrained -> if we built more it wouldn't sell at current prices...

or said another way: why are we building more fabs???

gb

What happened???

From what I heard AMD had a good Q1. Down over $3 on 22M shares before noon today.

Everyone bailing?

gb

at every transition some folks view it as a performance thing. while it is that in part it is also a continuous cost reduction thing while serving an ever growing market.

it's the natural progression for intel.

gb

Mike did a lot of fine work at Intel and saw a shot to run a company. A shot he probably wouldn't get at Intel.

Why does everything have to be about technology?

gb

re: vista delay

i think most folks would rather get a box with vista loaded rather than wait to see if they box they buy beforehand will meet their expectations for performance after they go through an upgrade.

i personally hate upgrades. for our family it means no new computers this year.

another few months on this p3/850 won't kill me.

if a lot of people think like me it will clobber consumer sales this year.

gb

i wondered the same thing when i read it.

bios?

64bit linux?

drivers? <--my bet

gb

Re: SSE128

For those who use the Intel libraries I wouldn't be surprised to see them updated with optimizations for Conroe.

Could be a nice kick for things like Software Defined Radio where lots of DSP functions are done on the CPU.

gb

I'd be surprised if anything other than the smallest IT shops could "switch" all of their servers to AMD. I'm guessing most AMD gains came from adding to or replacing "some" existing Intel based servers rather than wholesale replacement.

That means that most shops would still be largely Intel based with a percentage of AMD.

Not much to keep them from switching back to buying Intel on the next go around.

gb

pure guessing here but i suspect there's a point of mss loss and a larger contribution from overall market volume decline. maybe a little asp as well.

if i'm right amd will benefit from the mss gain, but lose on asp and overall market volume sagging just like intel.

probably ugly for all...

gb

conroe will finally replace this old p3/850!

wife will get a merom and i'll get her "old" centrino for the ham shack!

gb

although intel supplies the chipsets, the way they're connected, what they're connected to and how the software is loaded and "keyed" will likely all play a part in whether the drm "promise" is met. since the last enabling layer is from msft i doubt they'll take chances with whom they allow to apply it.

given how quickly the dvd protection was broken (mostly due to a couple of sloppy companies who were allowed access to it...), the overall anxiety of the movie industry to this whole notion and intel's and msft's desire to get "something new" going successfully i think it's going to be managed very closely.

i don't the success of this will depend at all on the ability of mom and pop pc shops or home hackers to participate.

gb

I suspect it's a bit of DRM hatred per se and also a bit of thumping the drum for whitebox, linux and DIY crowds.

For the latter there's no way the DRM promise can policed even if agreed to.

gb

Charlie hates DRM. He also doesn't seem to understand that DRM is enabled by hardware but fulfilled by MSFT software. That fullfillment is a promise made by MSFT to Hollywood after determining that their software applied to appropriate hardware by those deemed trustworthy will meet the DRM promise. This excludes the DIY crowd.

It also means that opensource is almost certainly ruled out since its pretty hard to keep an "open secret." "open-DRM" would be an interesting challenge.

But ViiV has other aspects to it as well which require that appropriate hardware be qualified like chassis, PS, graphics etc. But it really isn't ViiV unless it's all put together by someone MSFT trusts.

Actually he may understand all of this but it doesn't help his ranting cause.


edit: just got a email back from charlie who appears to understand all of the above with the oneliner: "Which is one of the many reasons why it is evil."

gb

all of this is compounded by the fact that intel has been more successful with integrated graphics in both desktop and especially notebook than anyone would have believed. that consumes silicon.

gb

If you look at what was done with Centrino it was tight integration of the platform that Intel provided to the OEMs as well as the mountains of behind the scenes testing with third party routers etc that went on to ensure that the brand promise was met. Combined with a big marketing campaign with hotspot providers it made it possible to tell your notebook customers that if they bought one it would really work and without those profit robbing customer support calls.

I expect the same thing from Viiv: stuff that really works rather than a lab experiment by teenage fanbois geek hobbyists. This time instead of hotspots you'll get content. To the extent that content is delivered by Intel websites it will no doubt only download to Viiv branded machines.

Impossible to duplicate without the connections that Intel has cultivated and requires a scale that can't be matched by the competition no matter what motley band they assemble around a hollow slogan. I'm sure Hector is aware how screwed he is and will no doubt add this to his lawsuit.

gb

On ViiV's chances:

During my pre-retirement days at Intel I watched a lot of programs promoted by various organizations get their chance at bat. Very few get the corporate muscle behind them and they only get that muscle when they have shown they have done all the work to set up for a win. Previous HTPC efforts were not in this class; not even close.

This one looks like it will out-do Centrino based on what I've seen so far. And Centrino itself has a nice run ahead of it as well.

And there won't be any useful piggybacking by AMD on it. I can't believe they actually tried that crap. They turn from piggybacking to squealing like a pig before the year is out.

gb

from the news reports it looks like viiv is off to a good start.

i expect intel will layer on new partners on a monthly basis to keep a rolling thunder.

i sure wish the cable card 2.0 spec were nailed down. combined with vista that would make me slap plastic for a htpc for sure.

this is going to be fun!

gb

my guess is that it is a lot easier to make "promises" to content providers when you are pretty sure of all the pieces.

corraling a bunch of independent chipset providers and trying to make the same claims would be quite difficult.

gb

Seagate buying Maxtor

http://news.com.com/Seagate+to+buy+storage+rival+Maxtor/2100-1015_3-6003822.html?tag=nefd.top

gb

i think this was for the lowend core stuff they got through one of their acquisitions that was intended for embedded control. i think they realized that they'll never have the critical mass to do anything with it so they just pawned it off for some cash and perhaps royalties. good luck collecting those...

gb

deleted

louis burns

gb

i wonder how this affects strataflash? i thought that was an intel exclusive.

perhaps alan81 might respond.

gb

i remember the internal discussions about the bandwidth requirements of the "netburst arch." it was acknowledged that it was bandwidth hungry but due to ooo and extensive buffering not too sensitive to latency. it was supposed to be a perfect match to the abilities of rambus.

i'm not qualified to debate it one way or the other but we know rambus didn't make it in the market.

i wonder what the performance would have been with rambus?

btw, it was also acknowledged that rambus/p4p performance advantage wouldn't show until above 3ghz. unfortunately by the time 3ghz was achieved rambus was toast.

gb

actually intel has always had sales coverage on apple.

the apple/intel warming was due to apple not any previous lack of interest on intel's part.

ceo change was irrelevant.

gb

ROFLMAO!

gb

actually their loan covenants prevent either dividends or stock buy backs.

gb

ial and systems group know what they're doing. btx is a good example of well thought out system design.

gb

this happens every time intel makes a shift. most of the mb suppliers don't want to make the change for a long list of reasons.

crap hits the fan for about 3 months, digitimes prints untold inches of tales of woe, inq repeats it like gospel, droids chant joyfully...then it all passes until the next time.

anyone remember the 430tx or 440bx? same ole, same ole...

gb

in my experience digitimes reports of chipset shortages usually are mb vendors wanting more of the old stuff when intel is pushing the new stuff.

happens every time...

gb

i think apple went with intc *for now* because they can't afford any additional risk in this transition. it will be difficult enough to cross over to x86 without amd's financial uncertainty.

further intel has resources to mitigate the risks in transition with their tech support in hw and sw.

roadmap and balance sheet...i suspect both weigh equally.

gb