backtop

Hardware AMD Sheds Additional Phenom Details
Gabriel Ikram (Blog) - October 15, 2007 10:48 AM
Print E-mail del.icio.us 70 comment(s) - last by raven3x7.. on Oct 21 at 4:43 PM
AMD reveals the model names and operating frequencies of its quad-core Phenom desktop processors

AMD guidance revealed to DailyTech the model names and clock frequencies of its forthcoming quad-core Phenom desktop processors.

AMD plans to debut its initial quad-core Phenom processors in late November: the Phenom 9600 and 9500. The 9600 will come clocked at 2.4GHz and the 9500 will have a slightly lower operating frequency of 2.2GHz.

Both Socket AM2+ processors feature a 512KB of L2 cache for each core and a shared 2MB pool of L3 cache. In addition, both processors will operate on a 3.6GHz HyperTransport bus and will have an 89 Watt Thermal Design Power envelope. 

In December the slightly higher-end Phenom 9700 will become AMD's third desktop Barcelona-derivative. All technical aspects of the processor will be the same as those for the 9600 and 9500, save for a few improvements. The processor will run at a slightly higher operating frequency of 2.6GHz. The downside of these improvements is a higher 125-Watt Thermal Design Power envelope.

An additional Phenom 9xxx desktop processor is on the horizon and is expected to make its debut in Q2 2008. Although the official clock speed has not been set for this new processor, AMD guidance suggests that it will not exceed 3GHz.  The company has demonstrated its 3GHz Phenom processors since July 2007.

Whereas Intel usually releases its high-end chips first, AMD seems to have a completely reverse policy. AMD will be launching its high-end Phenom desktop processors after the initial launch of lower-end Phenom desktop models.  AMD's high-end Phenom FX chips will appear on the market in Q1 and Q2 of 2008.

In Q1 2008 the Phenom FX-82 will be the first Phenom FX chip to make its debut. The chip will feature a clock speed equal to or greater than 2.6GHz . Like all of the other Phenom models, it will also come with 512KB of L2 cache for each of its four cores and a shared 2MB L3 cache. The Thermal Design Power of the chip is yet to be decided.

An additional Phenom FX-8x model will make its way to market in Q2 2008. AMD has yet to release additional details of this new processor.

All quad-core Phenom desktop processor are manufactured on a 65nm node.

This week AMD announced additonal price cuts and new chips for its Athlon brand of processors. These are expected to be the last Athlon price cuts before the Phenom launch.

In September AMD launched its newest Opteron server processors. The company has also confirmed that it's next-generation graphics processor, RV670, and its next-generation desktop chipset, RD790, will launch at the same time as Phenom.

Comments     Threshold


This article is over a month old, voting and posting comments is disabled
Will wait for a stepping or two.
By DesertCat on 10/15/2007 12:06:42 PM , Rating: 3
As someone who bought an AM2 board with hopes of upgrading at some point, I still figure I'll wait for this to sort itself out. This is for a variety of reasons

1) I need to see how these processors actually perform on an AM2 board as opposed to the AM2+ boards with higher hypertransport speeds. I mostly hear that it shouldn't make that much difference since we aren't close to saturating HT2. I want to see the benchmarks.

2) With time and additional steppings, I imagine these processors will become more affordable 3-6 months after their release. Most things seem to suggest that they had to spin a new stepping to get to the highest clocks. Another stepping may allow them to get more speed still. The net effect is that the fastest Phenoms at release may just become mid-high parts after a stepping allows them to raise the speeds. This will also tend to drive prices down on 2.6 GHz part since it will no longer be a flagship part (especially after the FX processors come out).




RE: Will wait for a stepping or two.
By KristopherKubicki (blog) on 10/15/2007 12:12:27 PM , Rating: 2
Well the funny thing is even with the B2 stepping, there's no 3.0 ghz chips -- even though the company has been showing off 3.0 ghz chips for months.

I have no idea what kind of PR move that is.


RE: Will wait for a stepping or two.
By Kuroyama on 10/15/2007 2:54:01 PM , Rating: 2
Well, it's a bit of an extreme case, but I recall back in 2002 when Intel showed Anandtech a 10GHz air cooled 32bit ALU. I'm sure both cases are just cherry picking a few good procs off the assembly line:

http://www.anandtech.com/showdoc.aspx?i=1584&p=5


By deeznuts on 10/15/2007 2:59:03 PM , Rating: 3
That was a tech demo, of just one part of a CPU. I believe he was referencing AMD showing off machines built with a 3ghz chip.


RE: Will wait for a stepping or two.
By defter on 10/15/2007 3:50:28 PM , Rating: 2
They still managed to ship a CPU with 64bit ALUs running at 7.6GHz...


By crystal clear on 10/15/2007 4:13:58 PM , Rating: 1
quote:
I have no idea what kind of PR move that is.


Its not PR rather plain marketing stratergy-anyway at the high end level they nothing worth the match to compete with Intel.

Intel Preps 45nm Quad-core Desktop Launch
http://www.dailytech.com/article.aspx?newsid=9070

Do they have a marketing manager to formulate/direct/implement their marketing statergies ???

AMD prefers to concentrate on the low end market because thats what is left for them to operate in.

Even in mainstream markets-low prices & low profit margins plays havoc with AMD.

AMD is responding to market feedbacks from dealers/resellers etc that guide its product launch statergy & prices.


RE: Will wait for a stepping or two.
By softwiz on 10/15/2007 12:15:33 PM , Rating: 1
quote:
As someone who bought an AM2 board with hopes of upgrading at some point, I still figure I'll wait for this to sort itself out.


You got plenty of time, no rush. It's not like AMD is going to abdandon / discontinue AM2 anytime soon such as when AM4 socket comes out leaving you to choose between AM3 / AM4. At least, I don't think they will. ;) (kidding)


RE: Will wait for a stepping or two.
By NARC4457 on 10/15/2007 12:47:14 PM , Rating: 5
Right, because they didn't do that at all with the 939 socket.

/Bitter


RE: Will wait for a stepping or two.
By Gul Westfale on 10/15/2007 1:23:18 PM , Rating: 1
yeah i'm on a 939, too... where are my phenoms, AMD?

on another note, doesn't it seem a bit strange that they are already at 89W and 125W with these chips, even though they are now using a more forgiving method of rating power consumption and even though these are new models?


RE: Will wait for a stepping or two.
By Locutus465 on 10/15/2007 1:48:01 PM , Rating: 4
Apparently S939 Opterons (still in the channel) are drop in replacements for Athlon 64X2s... So we're not completely abandond.... We S939 owners are just forced to live in the broom cubord and fight for scraps :P


RE: Will wait for a stepping or two.
By dgingeri on 10/15/2007 3:02:09 PM , Rating: 1
quote:
Apparently S939 Opterons (still in the channel) are drop in replacements for Athlon 64X2s... So we're not completely abandond.... We S939 owners are just forced to live in the broom cubord and fight for scraps :P


The Opterons only go up to 2.6Ghz. I had one. The AM2 chips currently go all the way to 3.2Ghz. I have one of those now. (my socket 939 board failed. I have 2GB of DDR400 and an Opteron 185 sitting around doing nothing right now.) The performance difference is very noticeable. Socket 939 is now totally obsolete.


RE: Will wait for a stepping or two.
By teldar on 10/15/2007 4:40:22 PM , Rating: 3
Are you looking to unload those obsolete parts at any time in the near future?

T


By Griswold on 10/15/2007 2:00:07 PM , Rating: 2
Think for a second.

If the 125W figure was following this new Intel compatible power scheme, it would not be called TDP but ACP (Average CPU Power).


RE: Will wait for a stepping or two.
By Einy0 on 10/15/2007 6:38:40 PM , Rating: 4
Actually AMD rates their chips TDP on absolute Maximum value at peak load. Not on an average like Intel does. So it's TDP ratings are far more strict.


RE: Will wait for a stepping or two.
By JumpingJack on 10/15/2007 10:02:06 PM , Rating: 1
Not true... read the specs....


By raven3x7 on 10/21/2007 4:43:42 PM , Rating: 2
Im not going to link it now but if you look up AMDs thermal guidance PDF it is stated that its max theoretical TDP.


By MikeyJ79 on 10/15/2007 7:34:29 PM , Rating: 4
quote:
on another note, doesn't it seem a bit strange that they are already at 89W and 125W with these chips, even though they are now using a more forgiving method of rating power consumption and even though these are new models?


Don't forget that these are quad-core chips which are fitting into the same thermal envelope as many of the 90nm dual-core chips.


RE: Will wait for a stepping or two.
By deeznuts on 10/15/2007 3:00:35 PM , Rating: 3
Even though I know it was planned abandonment, Socket 754 wasn't all that long either, was it?


RE: Will wait for a stepping or two.
By darkpaw on 10/15/2007 3:26:43 PM , Rating: 4
754 was around quite a bit longer as far as I remember. From the start of the Athlon 64s until even after the 939 was killed.


RE: Will wait for a stepping or two.
By Gul Westfale on 10/15/2007 4:20:22 PM , Rating: 3
yeah, because they made S754 semp0rns for quite a while.


RE: Will wait for a stepping or two.
By cheetah2k on 10/15/2007 9:03:54 PM , Rating: 2
While i dont want to start another "why did AMD ditch 939" debate, apart from HT2, i dont think there was any other advantage of AM2 over 939. I still use my FX-60 rig with 8800GTX in SLi and 4Gs of Corsair's best, and I just can't see any reason to move on up to the latest and greatest. Even on Vista (apart from memory speed rating of 5.1) I still pull 5.9, and 10.5k+ 3dmk06 scores across the board.

What ever happened to the days of socket/slot adaptors? I still remember the old Pentium 2/Celeron FPGA slot adaptors.. While I am sure its not feasible, i dream of the day to see some sort of socket adaptor to allow us die hard 939 platformers to continue to use what we have..

939 - live long and prosper! ^.^


By Yawgm0th on 10/16/2007 1:36:17 AM , Rating: 4
It's the problem of any integrated memory controller. It's great for performance, but the processor is basically married to a given memory architecture. If AMD had stuck with 939, AMD users would be forced to go with DDR, and DDR2 is substantially cheaper (and would be regardless of AMD).

I'm sure there were some other technical reasons, but they couldn't easily switch to DDR2 without switching sockets.


RE: Will wait for a stepping or two.
By Locutus465 on 10/16/2007 1:04:28 PM , Rating: 2
Odd that you're getting a system rating of 5.9 when your memory scores 5.1... Last I knew the over all system score was determained by the score of the slowest component in the system... I.E. if I used integrated graphics in my Athlon 64 X2 3800+ (OCed to 2.4GHz) with 2GB Corsair Twinx and SATA hard drives I would still be getting a system score of say, 2.0 (I can't imagine any integrated graphics gets better than 2.0).


By cheetah2k on 10/17/2007 12:53:43 AM , Rating: 2
quote:
Odd that you're getting a system rating of 5.9 when your memory scores 5.1... Last I knew the over all system score was determained by the score of the slowest component in the system...


Duh? You obviously miss-read what i wrote...


By 1078feba on 10/17/2007 1:39:10 PM , Rating: 3
I'm right there with ya Cheetah!

I have almost the exact same rig (FX-60 @ 3.2GHz), 'cept I'm sporting dual Ultras and Crucial Tracers at PC4000 DDR500 4x1GB. Pulled an 11.6 on 3DMark06.

The thing of it is, reading proc reviews on sites like this one, well, those bar graphs displaying avg FPS and encoding time are really, really starting to bring me down. Our venerable FX-60 is beginning to show up in the lower third-fifth. Sucks. I was on another site earlier today and say that GSkill has just hit the channel with a 4GB DDR2-800 kit (2x2GB), with a CAS latency of 4 (!!!). Hell, that's what I run. 4!?!? On a dual channel 2GB stick!? 4!?

Unless Phenom FX really knocks one out of the park, I will be going with 45nm Intel when it hits.

939, we hardly knew ye...


TDP
By ajfink on 10/15/2007 1:12:50 PM , Rating: 2
It's interesting that 2.2Ghz and 2.4Ghz parts will have a TDP of just 89w, but bumping it to 2.6Ghz takes it to 125w.

I wonder how well these things overclock, but that being said, I'm sure hitting a 3.0Ghz overclock will involve high-end air coolers and water cooling setups if they generate so much heat at the relatively low frequency of 2.6Ghz.

Dual 3.0Ghz Phenoms, though...I have a feeling they would encode some video right quick.




RE: TDP
By drebo on 10/15/2007 1:44:10 PM , Rating: 2
Remember that AMD is rather conservative with their TDP ratings. Typically you will almost never see the peak rating on current chips. I've got no reason to think otherwise for the new chips either.


RE: TDP
By JumpingJack on 10/15/07, Rating: -1
RE: TDP
By Griswold on 10/15/2007 2:09:22 PM , Rating: 2
Actually, its merely a matter of being honest with your hardware partners. It is clear that a computer builder prefers the absolute maximum number instead of some wague average/maybe/if figures. If it wasnt so, they wouldnt stick to TDP for the hard facts and ACP for the marketing.


RE: TDP
By JumpingJack on 10/15/07, Rating: 0
RE: TDP
By Viditor on 10/15/2007 11:11:11 PM , Rating: 2
quote:
But Intel's TDP is worst case... by spec, meaning the thermal solution must dissipate this for the worst case applications


The difference is that Intel's TDP is "based on measurements of processor power consumption while running various high power applications", while AMD's is based on " the maximum power a processor can draw for a thermally significant period while running commercially useful software"

In other words, Intel uses the measured power of these worst cases, and AMD gives the absolute maximum possible.
Obviously, the measured amount must be lower than absolute maximum...by how much is impossible to tell without knowing which applications we're talking about.


RE: TDP
By JumpingJack on 10/15/2007 11:49:26 PM , Rating: 2
Maximum power used for commerically useful software. You need to specify the load, which both do.... AMD uses commercially useful software.... Intel uses worst case applications where power sensitivity is concerned.

Again, the application Intel uses and is supplied to the OEMs for the core architecture is leaked... you can download it... you can use it... you can test it.. and you can conclude... what I stated above is correct.

You can even check the actual data, you can see it -- unconvolute with system level variability by simply looking at the power at the socket from the sites that measure it at the VRM.

http://www.digit-life.com/articles2/cpu/intel-pent...

How can this be -- Intel 65 Watt TDP processors actually consuming less than AMD's 45 Watt TDP processors... hmmmm weird (if we think like you), however those who know what TDP really means, this is not surprising.


RE: TDP
By JumpingJack on 10/15/2007 11:56:29 PM , Rating: 2
Ooops... forgot something, you forgot to bold the follow on in your bolding... I underlined the part you forgot to bold.

quote:
The difference is that Intel's TDP is "based on measurements of processor power consumption while running various high power applications ", while AMD's is based on " the maximum power a processor can draw for a thermally significant period while running commercially useful software "


RE: TDP
By JumpingJack on 10/16/2007 12:00:04 AM , Rating: 2
Sorry, had to this again... this is almost silly..

quote:
In other words, Intel uses the measured power of these worst cases, and AMD gives the absolute maximum possible.


Intel measures the power for the worst case scenario, which when talking dissipation would be the max it would produce... what part of "worst-case" do you not understand?


RE: TDP
By JumpingJack on 10/15/2007 2:03:28 PM , Rating: 5
On a side note -- a rapid increase in power for small delta's in clocks means two things....

a) that the are not conservative on the voltage to drive the transistor to saturation at the lower bins -- running the supply voltage at the very hairy edge.

b) 2.4 to 2.6 Ghz requires a voltage bump... compounding the power curve (it is a cubic function if you account both F and Vcore).... if it increases more than this (which is likely), then they have sever leakage problems.

Considering that they can't get a dual core above 2.6 GHz (Brisbane) with reliability high enough to slap a 3 yr warantee on the product, I lean toward the fact they have a fundamental problem with their drive currents and various oxide degradations.


RE: TDP
By coldpower27 on 10/15/2007 9:34:44 PM , Rating: 2
I don't think so...

http://www.xbitlabs.com/articles/cpu/display/core2...

Remember the QX6850 has a TDP of 130W...


Reviews reviews reviews
By dflynchimp on 10/15/2007 12:01:37 PM , Rating: 1
If AMD is confident that their chip can pose as competition to Intel's offerings (both Conroe and the upcoming Penryn), the best move would be to get their pre-production chips out to reviewers to prove they're worth looking at.

Something also has to be done about ATI's position in the graphics market...they've been too quiet lately (in stark contrast from their latest graphics cards, which are positively gas-hogging air-jets)




RE: Reviews reviews reviews
By wordsworm on 10/15/2007 1:39:26 PM , Rating: 4
quote:
If AMD is confident that their chip can pose as competition to Intel's offering

Anandtech recently did an article showing how AMD's 6400+ stacked up against Intel's Q6600. The difference in performance was slightly in favor of AMD while the 6400+ is about $50 cheaper last look-see. http://www.anandtech.com/video/showdoc.aspx?i=3122...
To beat that 6400 you'll have to spend close to 3x that amount of money. I call that competitive. If Phenom's going to make any significant gains, I can't help but feel that AMD is going to be an even greater adversary. The gap between AMD and Intel's high end stuff seems to be closing while the value AMD and performance AMD is bringing to the table is truly impressive and seems to be a touch stronger than what seems to be openly acknowledged.


RE: Reviews reviews reviews
By fleshconsumed on 10/15/2007 2:08:10 PM , Rating: 2
You are comparing quad core processor clocked at 2.4GHz to a dual core processor clocked at 3.2GHz which is not a valid comparison.

One, of course dual core AMD is going to be cheaper than quad core intel. Two, HL2:EP2 still does not utilize multiple CPUs and despite having 33% better frequency AMD CPU is only 5% faster than intel chip. I wouldn't call that impressive or competitive.

I am glad to see that amd will release Phenom CPUs sooner than Penryn, however it seems to me that AMD is falling back into place where it has been for most of it's life - competing at value segment where price/performance ratio matters most.


RE: Reviews reviews reviews
By JM360 on 10/16/2007 1:41:40 AM , Rating: 1
quote:
I am glad to see that amd will release Phenom CPUs sooner than Penryn


Umm... This report says that Phenom will likely be out in "late November." Penryn is set for a November 12th debut, long before what would be considered the "end of November."

So yet again, AMD is late to the party, and not in any way fashionable.

AMD is a day late and a dollar short with this launch. Barcelona is failing to live up to the many claims AMD had be trying to shove down our throats for nearly a year before its launch, and even after six months of delays, it's not what they claimed it would be.

Phenom, while faster than Barcelona, is not enough to keep AMD competitive. Finally they've slipped back to the bargain bin sector, where they belong.

- JM.


RE: Reviews reviews reviews
By fleshconsumed on 10/16/2007 3:26:35 PM , Rating: 2
I believe intel is going to release only extreme edition quad core models priced at $1K on November 12th. While good news for those who can drop 1K for a CPU, majority of people me including are much more interested in mainstream CPU's priced at $150-350 and that are going to be released much later. From this point it's nice to see AMD finally enter quad core market with a true quad core CPU and at hopefully reasonable prices.

Regarding your "Finally they've slipped back to the bargain bin sector, where they belong" pray that AMD stays competitive as otherwise it means much higher prices for both you and me. Or did you forget what happened in P2 days when intel dominated market? Or what happened just a year ago when AMD dominated dual core market with X2 CPUs? We were lucky to find entry level X2-3800 for less than $300.


RE: Reviews reviews reviews
By maroon1 on 10/15/2007 2:45:08 PM , Rating: 3
quote:
Anandtech recently did an article showing how AMD's 6400+ stacked up against Intel's Q6600. The difference in performance was slightly in favor of AMD while the 6400+ is about $50 cheaper last look-see


Why you don't you compare it to E6750 ? It seems that E6750 is better performer and cheaper


RE: Reviews reviews reviews
By DigitalFreak on 10/15/2007 2:58:27 PM , Rating: 1
Then it wouldn't fit his fanboy fantasy.


RE: Reviews reviews reviews
By Adonlude on 10/15/2007 2:59:26 PM , Rating: 2
Seriously dude, what the other 2 replies said. You are basically just way off in your comparison. Come back when you have learned about the CPU's and their market.


RE: Reviews reviews reviews
By Spuke on 10/15/2007 4:06:21 PM , Rating: 3
IMO, any CPU that does 100fps in a game is good enough for me. Anything else is just e-penising and a waste of money.

Exceptions are if you are using a computer in a capacity where every little bit of performance counts and your cash flow depends on that every bit of performance.


Where are you Barcelona Barcelonaaaaa
By crystal clear on 10/16/2007 2:07:39 AM , Rating: 2
Did AMD Barcelona and Intel Tigerton really ‘launch’?

AMD officially launched the “Barcelona” quad-core processor last month in San Francisco, but people are telling me that the systems are nowhere to be seen from any of the major server manufacturers nor is there any time line given for availability. To see if there is any merit to these complaints, I posed as a prospective IT shopper and called up some of the major Server makers such as HP, Dell, and IBM to see if I can actually order a server with AMD Barcelona inside of it.

http://blogs.zdnet.com/Ou/?p=816

Yet, Barcelona action is missing from HP, Dell, Sun, and IBM. You can order Intel's Tigerton from Dell and HP - but servers based on that processor probably won't ship until November.

AMD's Quad-Cores: A 1-Month Report Card

"We're extremely disappointed with AMD on a product delivery level. What's nice about the product is it runs great, it's rock solid, it really performs. The real problem seems to be is that AMD doesn't have any of these things," he said.

And to add insult to injury, the chip maker isn't explaining the hold-up on deliveries to its system builder channel, Corn said.

"Their management has no answers for the channel as to 'Where is this product?' Their response, instead of the truth, is silence. It's difficult getting no feedback from AMD, when a year ago Hector Ruiz said we will never turn our back on the channel. And now it looks like AMD has fallen into the same pattern, taking care of their Tier 1's and some favored partners, and turning their back on the channel," he said.

http://www.crn.com/white-box/202402138

To summarize all the above after reading the links-

Another Disaster in the making !

AMD should be now flooding the market with "Barcelona" instead we see shortages/unhappy customers/complaints/delivery problems etc etc.

Major Server makers such as HP, Dell, and IBM with nothing to offer .(Barcelona inside)

Intel stands to gain by this - dont expect buyers to wait for AMD or the Barcelona.

Very bad for AMDs future !




RE: Where are you Barcelona Barcelonaaaaa
By Amiga500 on 10/16/2007 5:42:43 AM , Rating: 2
Or conversely, from your 2nd link:

"There are no hardware conflicts and the power draw is as promised. They delivered on their technicals. On these high-performance compute and memory-intensive applications, they're kicking Intel's butt," said Brian Corn, VP of marketing and business development at Waltham, Mass.-based Source Code.

or Corn, the hope is that AMD gets its act together on getting Source Code the product it wants. Because despite the hassles, Barcelona is definitely a CPU that his company does want.

"The silver lining is that they have a rock solid product. The samples that went out, the benchmarks are great. The demand is huge.

Better to have a load of demand and not enough CPUs to fill it than loads of CPUs and no demand!


By Amiga500 on 10/16/2007 5:43:33 AM , Rating: 2
edit:

"or Corn" should read "For Corn"


By crystal clear on 10/17/2007 3:30:59 AM , Rating: 2
quote:
Better to have a load of demand and not enough CPUs to fill it than loads of CPUs and no demand!


Sorry could not respond earlier-busy-anyway.

Sounds great on paper (your quote),but in reality its not so.

Client/buyers have their commitments/schedules/deadlines/requirements/plans etc when upgrading to a new system.

You got to take these factors into consideration when deciding which system to use-
the Intel one or the AMD(in this case barcelona).

However good/compatible/fitting your requirements etc a CPU may be-
once you are not " guarranted " its supply/delivery,

then you have a problem in hand & a serious one.

You cannot wait endlessly/indefinitely for the delivery of your batch of CPUs.

You need definite/guarranteed delivery dates from AMD to build upon your project plans upon.

Once you realize that your supplier(AMD in this case) is unreliable ,
you SWITCH to an alternative source(Intel)...just remember that.

Every sale or NO sale reflects on AMD's

revenues/profits/market share/ etc etc...just remember that.

NO REPLY from AMD is iressponsible & unacceptable.

Rule 1 for AMD-Never keep a client/buyer waiting for their CPUs.

Rule 2-Keep enough in your inventories/warehouse for every buyer that approach you.

Rule 3-Gear up production schedules/quantities to meet the requirement of Rule 2.

(use J.I.T.-just in time techniques)

Rule 4- Match your supply with your estimated demand or projected demand.
For this you got to survey market to measure demand.

Rule 5-Unhappy/dissatisfied buyers dont return to buy again.

For all the above, Production managers & Marketing managers work in close cooperation together/hand in hand.

AMD has NO (full time)MARKETING MANAGER to coordinate its marketing activities/statergies/etc.

This is not the first time AMD has messed up its delivery schedules-resulting in shortages & LOST CLIENTILE.

AMD should be aware that they are in a fierce/cut throat/ruthless competition with Intel.

SURVIVAL OF THE FITTEST (laws of the jungle) apply.
let be Quality/Prices/Production/delivery/etc.

For AMD its a DO OR DIE SITUATION-FIGHTING FOR SURVIVAL

To summarize it all-

AMD has a very bad management team & management techniques/practices.

A good product behind it a bad management team

Not long ago AMD was playing with the idea of the FABLESS MODEL plan-plain suicide indeed.

Its not only important to get your product into the market but have enough of them to close a sale.

AMD should say to the buyer-YOU NAME IT..... WE GOT IT

Signed sealed & delivered & on to the next sale/client.

(yes dont forget to collect the cheque for the Accts dept- or you will have a serious cash flow problem or liquidity issues.)


Shared L2 cache?
By laok on 10/15/2007 12:16:03 PM , Rating: 2
I wonder why AMD quad does not share L2 as the core 2 duo does instead of sharing a L3.




RE: Shared L2 cache?
By KristopherKubicki (blog) on 10/15/2007 12:18:10 PM , Rating: 3
It has to do with the split-power planes I would figure.


RE: Shared L2 cache?
By mino on 10/16/2007 5:10:49 PM , Rating: 2
No, not even intel has a reliable 4-core shared cache production-ready.

Besides, 3-level approach is much more efficient in high-load server scenarios if perf/transistor is important. And this is critical for AMD.
Not to mention they would have build their whole cache architecture from scratch for shared L2 to work.


Prices?
By Proteusza on 10/15/2007 12:03:55 PM , Rating: 2
quote:
The 9600 will come clocked at 2.4GHz and the 9500 will have a slightly lower operating frequency of 2.5GHz.


Either the frequency should be 2.3 GHz, or that doesnt make sense at all.

No prices announced yet?




RE: Prices?
By caqde on 10/15/2007 12:10:00 PM , Rating: 2
Somehow I think the numbers are reversed.... 2.5 for the 9600 and 2.4 for the 9500... going by the rest of the article...


RE: Prices?
By KristopherKubicki (blog) on 10/15/2007 12:11:25 PM , Rating: 2
It was supposed to read 2.4 and 2.2 -- that's been corrected.

Thanks


Shirley there musht be shome mishtake....
By AMDfreak on 10/15/2007 12:06:12 PM , Rating: 2
quote:
The 9600 will come clocked at 2.4GHz and the 9500 will have a slightly lower operating frequency of 2.5GHz.


Huh?? Maybe you mean the 9600 is at 2.6GHz, or maybe the model numbers are switched....




RE: Shirley there musht be shome mishtake....
By Martimus on 10/15/2007 1:48:56 PM , Rating: 2
... and don't call me Shirley!

I haven't seen anyone bring up the old Airplane line in a long time. It could be that you just mispelled surely, but I will give you the benefit of the doubt and say you were using the old joke.


By fic2 on 10/15/2007 6:19:10 PM , Rating: 2
Looks like I picked the wrong week to quit sniffing glue.


This isn't competitive against Penryns
By crystal clear on 10/16/2007 2:23:50 AM , Rating: 2
I don't see how this lineup is going to turn things around financially for AMD the next two quarters.

The Ed of this site speaks his mind-

First, they announced a price cut, allegedly to clear out inventory for Phenoms.

At first glance, that sounds fine, except that AMD announced a new round of old models just a couple days later.

We'll see a couple of low-speed items and a 65nm 2.9GHz chip shortly, then something a little bit faster in Q1 2008.

Why would AMD do that, especially the 2008 product, if they have faster and better Phenoms coming?

The answer is they're not going to have faster Phenoms coming anytime soon. If you thought a 3GHz Phenom was going to show up with Santa, surprise!

No, the most you're going to see from Phenom by Ho Ho Ho time is what looks to be maybe a bit overvolted 2.6GHz, if the power requirements are any indicator.

Will there be a 3GHz Phenom X4 in Q1? No. Will there be one in Q2? No. If you want a 3GHz K10, you're going to have to buy an FX in Q2, probably in June.

I don't doubt you'll be able to overclock a December or later X4 to around 3GHz (and maybe those reports we got about 3GHz processors were reports of overclocked processors), but by the time you can do that, you'll be able to overclock a Penryn to around 4GHz.

Any speedy cripplecores coming soon? Uhhh, no. They show up in March, at a whopping 2.3 and 2.5GHz speeds.

Some may ask, "Surely the duallies will do better?" No, they're going to do a bit worse; they aren't showing up at all until Q2, again, read that as meaning June.

The sad, sad truth is for about the next six months on the desktop, the fastest AMD CPU for just about anything that doesn't take serious advantage of multithreading (i.e. most apps) is going to remain an X2. A 90nm X2.

What can I say? This isn't competitive against Penryns. This isn't even competitive against C2Ds. I don't see how this lineup is going to turn things around financially for AMD the next two quarters.

Ed



http://www.overclockers.com/tips01229/

Let him read about those delivery problems/shortages/unhappy customers etc etc in my ealier comment,then he would fall off his chair or his keyboard.




RE: This isn't competitive against Penryns
By Proteusza on 10/16/2007 7:36:37 AM , Rating: 2
AMD's biggest problem doesnt seem to be developing new products, it seems to be getting them out in a timely fashion, and doing press releases so that people know what is happening.

They are silent for months at a time, then one day we hear of Bulldozer etc, as well as their other long term plans.

AMD really needs to sort this out - if they have learnt anything from Intel, it should be that you can sell a bad product with good marketing.

Currently, their products are second to Intel's, and their marketing is non existant.


By crystal clear on 10/17/2007 3:48:11 AM , Rating: 2
To summarize it all-

AMD has a very bad management team & management techniques/practices.

A good product behind it a bad management team

Its not only important to get your product into the market but have enough of them to close a sale.

AMD has NO (full time)MARKETING MANAGER to coordinate its marketing activities/statergies/etc.

This is not the first time AMD has messed up its delivery schedules-resulting in shortages & LOST CLIENTILE.

I intentionally put it into bold font since the issue is a serious one.

The full management team has to GO.

Expect to hear on their departure very soon.


By bysmitty on 10/15/2007 2:33:11 PM , Rating: 2
I've been holding off on a computer upgrade until I can see what AMD brings to the table. I don't like that they are starting low and working their way up. I'm getting tired of waiting and seeing low benchmarks on the first gen Phenoms might just push me over to Intel's performance proven C2Qs.

...bysmitty




By darkfalz on 10/21/2007 6:19:26 AM , Rating: 2
So you wait more than a year while everyone are enjoying their overclocked C2Ds and miss out on all that performance for...? To wait and see?


speaking of next gen cpu's...
By maverick85wd on 10/16/2007 10:41:56 AM , Rating: 2
has anyone heard anything about AMD or Intel coming out with High-K technology parts in the somewhat-near future? From what I've heard it will improve performance on the same scale as SOI technology, I just haven't heard of any potential products or when they may come about




RE: speaking of next gen cpu's...
By JM360 on 10/16/2007 12:17:11 PM , Rating: 2
Intel's Penryn processors will make use of High-K technology, so yes, High-K will be released in the very-near future.


Link
By imperator3733 on 10/16/2007 11:04:08 AM , Rating: 2
Did anyone notice the link http://www.dailytech.com/AMD+Demonstrates+30+GHz+Q... (fifth paragraph)? I thought that AMD showed off a 3 GHz quad core, not a 30 GHz one!




RE: Link
By AnnihilatorX on 10/17/2007 5:59:57 PM , Rating: 2
lol well using "." in naming of webpages in http isn't too smart that's why 3.0 -> 30


Conference call
By crystal clear on 10/16/2007 12:57:48 AM , Rating: 2
Here we go again-

Q3 2007 Advanced Micro Devices Earnings Conference Call
Thursday, October 18, 2007 5:00 p.m. ET
Webcast Presentation Help
Click here for webcast
Click here for webcast

Event Details

Title Q3 2007 Advanced Micro Devices Earnings Conference Call
Date and Time Thursday, October 18, 2007 5:00 p.m. ET
Duration 1 Hour


http://phx.corporate-ir.net/phoenix.zhtml?p=irol-e...




#!$# amd
By GlassHouse69 on 10/15/07, Rating: -1
RE: #!$# amd
By xti on 10/16/2007 12:40:49 AM , Rating: 1
interesting


"I want people to see my movies in the best formats possible. For [Paramount] to deny people who have Blu-ray sucks!" -- Movie Director Michael Bay

DailyTech Poll
Which web browser do you use on your primary personal machine? 






44 Comments









botimage
Copyright 2009 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki