Intel stole back 4% market share in Q1

pxc

Extremely [H]
Joined
Oct 22, 2000
Messages
33,063
according to iSuppli: http://www.tgdaily.com/content/view/31666/118/

Based on preliminary findings, iSuppli vice president Dale Ford estimates that AMD’s overall market share in the microprocessor industry may have declined 4.7 percentage points from 15.7% in Q4 2006 to 11.0% in Q1 2007. The analyst believes that Intel was able to regain market shares in the same time frame: “On a preliminary basis, I would estimate that Intel increased its market share in microprocessors from 75.7% in Q4 to 79.5% in Q1,” Ford told TG Daily.

During that time Intel was still selling mostly Pentium 4s, Celerons and Pentium Ds. Ouch. I guess Intel sold so many that it has ramped up Core 2 a whole quarter ahead of schedule. I really can't wait for the E2xxx Conroes.
 
If those numbers hold up, that is an interesting shift and make me wonder what the Q2 numbers will look like.
 
Something else in that quote that I think is interesting, Intel only gained 3.8% market share while AMD lost 4.7%. Who gained the other 0.9%? VIA?
 
Something else in that quote that I think is interesting, Intel only gained 3.8% market share while AMD lost 4.7%. Who gained the other 0.9%? VIA?
They're just preliminary estimates. Split the difference and you'll probably get a closer estimate for each.
 
I wonder if AMD lost the lead in the retail market share. Or if it's still winning?
 
I wonder if AMD lost the lead in the retail market share. Or if it's still winning?
AMD went from 81.5% at the beginning of 2006 to:

link in the first post said:
Current Analysis reported last week that AMD recently lost its dominance in the U.S. desktop retail market to Intel. According to the market research firm, AMD held a 43% share in Q1, down from about 54% in Q4 2006 and 66% in Q3 2006.
 
ouch. If AMD is lying about Barcelona being 50% faster than Conroe, then they are done.
 
pxc said:
They're just preliminary estimates. Split the difference and you'll probably get a closer estimate for each.

Via lost big last year. It wouldn't surprise me if they did gain back 0.9%. I read recently about a deal between HP and Via.
 
ouch. If AMD is lying about Barcelona being 50% faster than Conroe, then they are done.

Sadly, I think they are exaggerating about it, probably to gain investors back. These guys better start pulling out their buckets, they are a sinking ship.
 
Sadly, I think they are exaggerating about it, probably to gain investors back. These guys better start pulling out their buckets, they are a sinking ship.

Sinking ship? Worse....they are a submarine....and Intel is firing MK 48s at them. (Plural.)

I'm not being a *!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!; it's simple macroeconomics, microprocessor-style. Intel launches Kentsfield; AMD tries to respond with processor price cuts. Now here comes the second salvo of MK48s: beefier low-end Conroes and Kentsfield II - both with lower price tags. (AMD had already said that their own quad-core response won't ship until 1Q '08 at the earliest.) The *third* salvo is coming in three months (this one is strictly price cuts, except for the death of E6600/E6700; however, both the killed Conroes are being directly replaced by Kentsfield IIs at the old Conroe pricing, if not cheaper).

Kentsfield II: the multi-locked version of the original Kentsfield (remember, Kentsfield launched as an unlocked processor)
 
Sinking ship? Worse....they are a submarine....and Intel is firing MK 48s at them. (Plural.)

I'm not being a *!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!; it's simple macroeconomics, microprocessor-style. Intel launches Kentsfield; AMD tries to respond with processor price cuts. Now here comes the second salvo of MK48s: beefier low-end Conroes and Kentsfield II - both with lower price tags. (AMD had already said that their own quad-core response won't ship until 1Q '08 at the earliest.) The *third* salvo is coming in three months (this one is strictly price cuts, except for the death of E6600/E6700; however, both the killed Conroes are being directly replaced by Kentsfield IIs at the old Conroe pricing, if not cheaper).

Kentsfield II: the multi-locked version of the original Kentsfield (remember, Kentsfield launched as an unlocked processor)

Quad Core is still priced too expensive for the majority of the market even at the new $530US price point with mainstream Q6600.

For the moment AMD is competitive performance wise up to the E6600, but any Dual Core above that no.

AMD's Agena and Kuma desktop competitors to the Kentsfield and Conroe, won't arrive till Q4 2007, so they will be coming in hot, basically into a very competitive atmosphere.
We will have to see, but with Dual Core at $74 to $266 and $266 Quad's at that time, AMD isn't going to have that much leeway with pricing unless they can beat the E6850 and QX6850 by then.
 
Oh yeah it is impressive that Intel was able to gain 4.7% I believe points back in terms of revenue share reducing AMD down to 11% in terms of revenue share.

We have no figures on unit marketshare for the moment.
 
It was actually 6%: http://yahoo.reuters.com/news/artic...4-24_19-36-06_N24434675&type=comktNews&rpc=44

;)

This should be unit share figures:

Intel's share of the $30 billion market for x86 processors that power most personal computers was 80.5 percent in the first quarter, according to Mercury Research, a market tracking firm whose data are closely watched by the industry.

That represented a gain of more than 6 percentage points from the 74.4 percent Intel had in the fourth quarter.
...
JP Morgan also cited Mercury data as showing that Intel gained 8 percentage points of share in the desktop segment, about 4 points of share in notebooks, and 7 points in the market for server computers.
 

Wee go Intel. :D

" JP Morgan also cited Mercury data as showing that Intel gained 8 percentage points of share in the desktop segment, about 4 points of share in notebooks, and 7 points in the market for server computers."

So going from AMD's Q4 2006 Marketshares.

Desktops: 29.1% Up from LQ
Laptops: 19.4% Up from LQ
Servers: 22.3% Down from LQ

Overall 25.3%

A 6.1% Share gain at the expense of AMD would bring AMD back down to 19.2% overall, with AMD down to 21.1% in Desktops, 15.4% in Notebooks, and 15.3% in Servers give or take.

Overall very impressive for Intel.
 
CPU's users need to pray that AMD can counter. Or down the road it could be like it was a few years back with high priced Intel processors without competition.

Price wars are good, but if a company is wiped out, that is bad for consumers.
 
Stupac said:
CPU's users need to pray that AMD can counter. Or down the road it could be like it was a few years back with high priced Intel processors without competition.

Price wars are good, but if a company is wiped out, that is bad for consumers.

Pray for corporations? How absurd and superficial. I can do the whole prayer + Ford, Pepsi-Cola, milk, gas, lottery, and cheese thing. I'm going to spare us a rant, though. Why don't you just pray you "make it down the road", OK?:rolleyes:
 
Pray for corporations? How absurd and superficial. I can do the whole prayer + Ford, Pepsi-Cola, milk, gas, lottery, and cheese thing. I'm going to spare us a rant, though. Why don't you just pray you "make it down the road", OK?:rolleyes:

Guess your not a capitalist or a consumer.;) Maybe a literalist?
 
Quad Core is still priced too expensive for the majority of the market even at the new $530US price point with mainstream Q6600.

For the moment AMD is competitive performance wise up to the E6600, but any Dual Core above that no.

AMD's Agena and Kuma desktop competitors to the Kentsfield and Conroe, won't arrive till Q4 2007, so they will be coming in hot, basically into a very competitive atmosphere.
We will have to see, but with Dual Core at $74 to $266 and $266 Quad's at that time, AMD isn't going to have that much leeway with pricing unless they can beat the E6850 and QX6850 by then.

QFT

Agreed; however, that price will only last three months ($530 for Q6600). AMD's immediate problem is that Intel has refused to sit still.

1. In three months, Q6600 will completely replace (pretty much) E6600, and a new Q6700 will hit the market, replacing the E6700 (and at the E6700's pricetag). Note that neither Agena or Kuma will have shipped yet.

2. E6320 and E6420 are available right now, and in quantity. The price spread compared to their 2+2 MB predecessors is no better than $5USD. Within three months, I actually expect the E6x00 Conroes to vanish from retail.

3. Intel's processor marketshare is only *helped* by the current chipset price war between Intel and nVidia. (The system-board price advantage AMD used to have is now gone completely, including the value non-onboard graphics area, which was AMD's biggest strength.) If you look at the midrange, you have two Intel chipsets (965 and 975X) and three nVidia chipsets (650i Ultra, 680i LT, and 680i Ultra). Note that all five chipsets have systemboards available in quantity, and in most cases under $200USD (even for 680i Ultra or 975X-based systemboards).

All this is either current or will happen with certainty in three months (as stated earlier, before Agena or Kuma even ship). When Agena and Kuma finally do ship in Q4 07 (barring delays), they will be facing an *established* E6320/E6420 (4+4 MB Conroe), Q6600/Q6700 (Kentsfield II) and QX6850 (desktop Clovertown).

In what way is that good for AMD?
 
Stupac said:
Guess your not a capitalist or a consumer.;) Maybe a literalist?

LOL! What are you Homey da Clown? Ok, whatever homey.

Sure, AMD is feeling the sting. They've had to drop their prices and face lowered demand. As if no one could see it coming a year ago.

For AMD to stop making processors, management needs to be insane, shareholders need to be insane, OEM's and other partners need to be insane, and last but not least, someone must be insane enough to buy them out then take them out of business.

Gee willy!

As if running low on funds(having only several hundred million dollars in cash), and owing a couple billion dollars somehow prevents them from making job cuts, selling off equipment, finance/refinance, or making a major change in direction; they can trim the fat and reconstitute.

They've seen worse and worse is yet to come. They might take a while to recover or they might never get back to where they were before, or they might come out on top. But, barring some Enron-like incident, they're not going to stop making cpus for the forseeable future.
 
QFT

Agreed; however, that price will only last three months ($530 for Q6600). AMD's immediate problem is that Intel has refused to sit still.

1. In three months, Q6600 will completely replace (pretty much) E6600, and a new Q6700 will hit the market, replacing the E6700 (and at the E6700's pricetag). Note that neither Agena or Kuma will have shipped yet.

2. E6320 and E6420 are available right now, and in quantity. The price spread compared to their 2+2 MB predecessors is no better than $5USD. Within three months, I actually expect the E6x00 Conroes to vanish from retail.

3. Intel's processor marketshare is only *helped* by the current chipset price war between Intel and nVidia. (The system-board price advantage AMD used to have is now gone completely, including the value non-onboard graphics area, which was AMD's biggest strength.) If you look at the midrange, you have two Intel chipsets (965 and 975X) and three nVidia chipsets (650i Ultra, 680i LT, and 680i Ultra). Note that all five chipsets have systemboards available in quantity, and in most cases under $200USD (even for 680i Ultra or 975X-based systemboards).

All this is either current or will happen with certainty in three months (as stated earlier, before Agena or Kuma even ship). When Agena and Kuma finally do ship in Q4 07 (barring delays), they will be facing an *established* E6320/E6420 (4+4 MB Conroe), Q6600/Q6700 (Kentsfield II) and QX6850 (desktop Clovertown).

In what way is that good for AMD?

I don't think I ever implied Intel's price cuts were good for AMD. :)

1. Your also not considering the new 1.33GHZ FSB Conroe's coming out in Q3 to replace the current batch of Conroe processors, E6320 and E6420 don't have any official pricing information in Q3. You haven't factored the April 22nd price cuts have you E6700 is already down to $316 now, so the Q6700 will be replacing the Q6600 at it's price point today.

2. No there will be a few new models of Conroe coming out to address the gap between $133 Allendales and $266 Kentsfields. The new models are as follows E6540, E6550, E6750, and E6850.

They will be prices at Unknown, $163, $183 and $266.

Look at this page to view the upcoming models, made largely by yours truly.
http://en.wikipedia.org/wiki/List_of_Future_Intel_Core_2_microprocessors

So in Q3 after all the introduction your lineup looks like this.

Celeron 4xx, Pentium E, Core 2 Duo E4x00, Core 2 Duo E6xx0, Core 2 Quads E6x00, and finally Core 2 Extreme QX6850.

The pricing structure is as follows:
$39-$59, $74-$84, $113-$133, $163-$266, $266-$530, and $999 or higher.

Like AMD they give you a choice higher clock Dual Core vs lower clocked Quad. look at the Athlon 64 4000+ vs the Athlon 64x2 3600+ for something skin to that, except Single to Dual.

3. By Q3, the 1066FSB Conroe will be on their way out as they aren't having a price drop replaced by 1333FSB Conroe no less then 2.33GHZ and up to 3.00GHZ.

You will have the E4400 and E4500 picking up the slack for the E6320 and E6420. As in most situations they will likely be faster stock.
 
I don't think I ever implied Intel's price cuts were good for AMD. :)

1. Your also not considering the new 1.33GHZ FSB Conroe's coming out in Q3 to replace the current batch of Conroe processors, E6320 and E6420 don't have any official pricing information in Q3. You haven't factored the April 22nd price cuts have you E6700 is already down to $316 now, so the Q6700 will be replacing the Q6600 at it's price point today.

2. No there will be a few new models of Conroe coming out to address the gap between $133 Allendales and $266 Kentsfields. The new models are as follows E6540, E6550, E6750, and E6850.

They will be prices at Unknown, $163, $183 and $266.

Look at this page to view the upcoming models, made largely by yours truly.
http://en.wikipedia.org/wiki/List_of_Future_Intel_Core_2_microprocessors

So in Q3 after all the introduction your lineup looks like this.

Celeron 4xx, Pentium E, Core 2 Duo E4x00, Core 2 Duo E6xx0, Core 2 Quads E6x00, and finally Core 2 Extreme QX6850.

The pricing structure is as follows:
$39-$59, $74-$84, $113-$133, $163-$266, $266-$530, and $999 or higher.

Like AMD they give you a choice higher clock Dual Core vs lower clocked Quad. look at the Athlon 64 4000+ vs the Athlon 64x2 3600+ for something skin to that, except Single to Dual.

3. By Q3, the 1066FSB Conroe will be on their way out as they aren't having a price drop replaced by 1333FSB Conroe no less then 2.33GHZ and up to 3.00GHZ.

You will have the E4400 and E4500 picking up the slack for the E6320 and E6420. As in most situations they will likely be faster stock.


1. Not factored in because they are largely irrelevant (except for the E65x0 series, which will slot between the E64x0 and Q6600) in the greater marketplace, except where they directly replace 1066 MHz FSB Conroes, and other than than that, there is no impact from the E63x0/64X0 at all, except for falling prices.

2. I pointed out that Q6700 will in fact pretty much replace E6700 (not Q6600, which drops to $266 in Q3). E6320/E6420 are new in the current quarter, and have no price cuts scheduled in Q3 (their lower-cache E6300/E6400 cousins are going to die, however). So, above Allendale, you will have E6320/E6420/E6540 (Q3 launch)/E6550 (Q4, will likely replace E6540)/Q6600 (available now, price drop in Q3)/Q6700 (Q3 launch)/QX6850 (replacing both X6800 and QX6700, possibly not until Q4 to let current stock shrink). It's possible we could see Q6650/Q6750 (both in Q4) depending on yield outcomes.


3. Notice that E66x0 is missing; that is deliberate. The E65x0 will basically replace the E66x0 series; except for current stock of X6800, this will be the ceiling for dual-core desktop CPUs (all series above this will be either Kentsfield or Clovertown; both of which are quad-core). So far, there has been no real shortage of quad-core CPUs (either Kentsfield or Clovertown); this means there is little or nothing preventing Intel from taking the faster increased-cache Conroes and making perfectly good (if multiplier-locked) Kentsfields out of them (especially if they clock fast enough). Remember, E66x0 and E67x0 are both going away; however, QX6850 is shipping today. What do you do with product that doesn't bin as fast, if you're Intel? Simple; lock the multiplier and turn them into perfectly usable Kentsfield II (or even Clovertown II) processors in the Q66x0/Q67x0 series.

Intel apparently is *very* serious about taking quad-core mainstream; Allendale has basically become a Celeron replacement in the value segment, and could find itself replaced with Conroe in six months. Conroe has already moved from the midrange to the lower-end (albeit at faster FSB speeds) and with the death of E66X0, Conroe's own days in the midrange are numbered. Q6600 is Kentsfield for the mainstream (though the initial debut pricing belies that); however, future Kentsfield variations (based on larger-cache higher-FSB dual-core Conroes) will further increase the conundrum facing enthusiasts now (and both midrange and eventually mainstream users in the next six months), especially with Q66x0/67x0 (and of course, QX68x0) in the future.

"Quad-core now, or wait for the apps?"
 
1. Not factored in because they are largely irrelevant (except for the E65x0 series, which will slot between the E64x0 and Q6600) in the greater marketplace, except where they directly replace 1066 MHz FSB Conroes, and other than than that, there is no impact from the E63x0/64X0 at all, except for falling prices.

Maybe for you, but for many Dual Core is already enough, and for certain workloads more clockspeed is more useful then Quad. So E6750 and E6850 have their place. Nothing you told me here I don't already know. Not all people are going to go for Quad. E6750 and E6850 are quite relevant as they help bridge the gap at $183 and $266. You can't jump from $163 straight to Kentsfield's $266, you need intermediate products.

Assuming Intel is still pushing E63x0 and E64x0, I still believe the Allendale E4400 and E4500 would render them redundant, unless you absolutely need Virtualization, or are running a very cache dependent app, as they have higher clockspeed which is where the majority of the performance comes from.

2. I pointed out that Q6700 will in fact pretty much replace E6700 (not Q6600, which drops to $266 in Q3). E6320/E6420 are new in the current quarter, and have no price cuts scheduled in Q3 (their lower-cache E6300/E6400 cousins are going to die, however). So, above Allendale, you will have E6320/E6420/E6540 (Q3 launch)/E6550 (Q4, will likely replace E6540)/Q6600 (available now, price drop in Q3)/Q6700 (Q3 launch)/QX6850 (replacing both X6800 and QX6700, possibly not until Q4 to let current stock shrink). It's possible we could see Q6650/Q6750 (both in Q4) depending on yield outcomes.

The Q6700 doesn't get introduced till Q3, but like I have to re-iterate again, it's is coming in at $530, it is replacing the Q6600 because that was the last CPU at $530 prior to the Q6700 launch, the Q6600 gets a price bump down to $266. It is hardly replacing the E6700 as at that time the E6700 is only $316 and not a good deal as it is a discontinued processor and didn't receive any price drops the E6750 is only $183 and offers the same or greater performance.

On April 22nd E6600 got replaced by E6700, E6700 got replaced by Q6600, and there will be nothing new at $851.

All 1.33GHZ FSB models are launching in Q3, and I don't really see the point of launching 1.33GHZFSB Quad in the mainstream till Q1 2008 with Yorkfields.

3. Notice that E66x0 is missing; that is deliberate. The E65x0 will basically replace the E66x0 series; except for current stock of X6800, this will be the ceiling for dual-core desktop CPUs (all series above this will be either Kentsfield or Clovertown; both of which are quad-core). So far, there has been no real shortage of quad-core CPUs (either Kentsfield or Clovertown); this means there is little or nothing preventing Intel from taking the faster increased-cache Conroes and making perfectly good (if multiplier-locked) Kentsfields out of them (especially if they clock fast enough). Remember, E66x0 and E67x0 are both going away; however, QX6850 is shipping today. What do you do with product that doesn't bin as fast, if you're Intel? Simple; lock the multiplier and turn them into perfectly usable Kentsfield II (or even Clovertown II) processors in the Q66x0/Q67x0 series.

The E6650 was renamed to E6550 because Conroe still derives the majority of it's performance from it's clockspeed. Since Conroe doesn't support half Multipliers E6650 can't exist as at 2.33GHZ/1333 the performance is inferior to E6600 2.4GHZ/1066. So naming it E6650 is misleading. Your forgetting E6750 and E6850, they still exist for those who want better Single Threaded performance on the cheap. They have their purpose. The E6850 allows transition to Quad, for the same price you can get a lower clocked Quad at the same price. So in the end you sacrifice Single Threaded performance for Multi Threaded. Tit for Tat and since you have 2 processors to choose from the user is able to select the processor he or she feels gives here the best performance for their needs.

E6600 and E6700 are leaving to be replaced by faster models at cheaper prices, ala the E6750 and E6850. There is no longer a need for X6800 unless you want an unlocked multiplier.

I believe the QX6850 returns Intel to having a differentiating factor from the Extreme to the Mainstream then just an unlocked multiplier. For the moment I see Kentsfield remaining on 1.06FSB.

Intel apparently is *very* serious about taking quad-core mainstream; Allendale has basically become a Celeron replacement in the value segment, and could find itself replaced with Conroe in six months. Conroe has already moved from the midrange to the lower-end (albeit at faster FSB speeds) and with the death of E66X0, Conroe's own days in the midrange are numbered. Q6600 is Kentsfield for the mainstream (though the initial debut pricing belies that); however, future Kentsfield variations (based on larger-cache higher-FSB dual-core Conroes) will further increase the conundrum facing enthusiasts now (and both midrange and eventually mainstream users in the next six months), especially with Q66x0/67x0 (and of course, QX68x0) in the future.

"Quad-core now, or wait for the apps?"

$266 is still not really mainstream more like performance-mainstream, you need prices to fall down to mid $150 for them to really be selling like hot cakes. Allendale still remains a little higher then Celeron, basically a Pentium 4 HT and Pentium D replacement. We need something as cheap as Pentium D 805 was when it debuted, this will depend on AMD's Quad Core performance in Q4.

You mean Yorkfield which is based on having 2 Wolfdale dies as 1 MCM, which has a 2x6Mb Cache structure, Q6600 already has the max cache available on 65nm.

To the last statement, depends on the customer.
 
Back
Top