Pentium 4 has zero market appeal
There is an interest poll at hexus.net. It asks what will your next CPU upgrade be. Not surprisingly, 24% of people want Conroe, which should make Intel marketing smile. Their guerilla benchmarketing paid off.Equally expected is that less than 1% of people want Netburst (P4 and Pentium D). It's reported that Intel will however introduce a new Pentium D 9xx stepping (D0) by the end of 2006. It seems Intel fans will be stuck with Netbust for quite a while as Conroe's 4MB cache will ruin yields. Presler's double die approach does have a yield advantage.
AMD's socket AM2 enjoys a healthy 34% demand, and socket 939 has 16% popularity. This indicates that most people know that Intel is hiding the truth about Conroe, and they would rather choose a proven high performance future proof platform.
posted by Sharikou, Ph. D. @ 6/16/2006 08:11:00 PM
41 Comments:
Netburst? What's that? It's a kind of oven?
But more seriously, what do you think of these reviews?
http://www.tbreak.com/reviews/article.php?cat=cpu&id=456&pagenumber=1
http://www.digit-life.com/articles2/video/conroe.html
(Damn, no Quake4 or Oblivion benchs...)
Yes and we all know how reliable Internet polling is...
Let me predict... Next quarter (Q3) Intel will have roughly 3.5 times the sales as AMD... just like last quarter. I also predict that Pentium 4 will have significantly more than the zero percent market share predicted by this poll. I guess we can say hexus.net readers don't represent the general market ... but you already knew that. Didn't you?
That's before the upcoming price cuts though, which may entice quite a few people away even if Conroe and AM2 is faster. It also shouldn't be used to represent world demand though since Hexus is an enthousiast site. I'm sure the average person would be more interested in the price tag and other big flashy marketing tools like odles of RAM and a "high-performance" graphics card.
Also, Intel isn't standing still with the Pentium Ds either. Even though you may downplay it, their 65nm is constantly improving.
http://intel.pcnalert.com/content/eolpcn/PCN106404-00.pdf
They're now dropping the TDP of the 3.6GHz 960D to 95W, which is quite impressive considering the first single core 3.4GHz 540 had a 115W TDP. The entire Pentium D product line will now come in at 95W. I'm sure all the other models down to the 915D will also see power consumption reductions even though it isn't advertised because it's still under 95W. The single core Cedar Mills also all already have a 65W TDP.
You can't really complain if you are paying less for a better product than before, even if it is a Netburst.
It's funny that you criticize Hexus for being biased toward Intel, but you trust their polls. I guess because the results are in your favour.
It's funny that you criticize Hexus for being biased toward Intel, but you trust their polls.
Dude, Hexus.net is a paid Intel pumper doesn't mean all visitors to their site are Intel pumpers.
With energy costs increasing every day, there really is no reason to choose Intel's current desktop/server chips, all of which use way too much juice and run way too hot.
While "Core Du Over" is more energy efficient than Pentium D, it still runs quite hot. Witness how most Centrino notebooks, single and dual core, run hot compared to AMD versions of the same notebook.
Even if Intel were to give away their Pentium processors, over the lifespan of a computer, they still cost more than AMD due to energy usage.
Intel's massive price drops on Pentium-based chips are basically just a Walmart-like blowout of chips that have no place in a world where energy costs money. It is good to know Intel is replacing Pentium 4 with better chips sometime in 2007 or 2008.
On the consumer front, Pentium 4 and Pentium D is gonna frag all AMD sales for Semprons, A64s and even X2s. The new Pentiums are now much cooler and lower power thanks to applying 65nm process. Heck, push it to 4GHz and it will give A64 and even FXs a run for their money. Coupled with cheaper prices its a basement bargain. Big OEMs like Dell and HP/Compaq still takes them... Now AMD chips beginning to look like zero market appeal?
What I want to know is whether $530 for a 2.67 GHz Conroe is a retail price, or the per-unit price when you buy in lots of 1000 units. AMD traditionally prices in 1000 unit lots as seen here:
http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,51_104_609,00.html
Also, in the digit-life review, am I reading it right that they're reviewing a single core 2.4 GHz Athlon 64 against a 2.13Ghz a dual core Conroe on the basis that the second Core 2 core doesn't do anything in games and the AMD's extra MHz will make up for it? Now, I'm not really anal about minor discrepencies in benchmarks, but isn't this precisely the thing we have benchmarks to determine?
I'm not sure what conclusion to draw from that benchmark. It looks like Conroe wins overall, but where the Athlon 64 wins, the author blames it on the Conroe system not being fine-tuned, or just states that the Athlon 64's win is "anomolous" and "most likely incorrect."
Oh boy! Another person harping on energy consumption. Now that the entire Pentium D line will be @ 95W it would take about a decade to even make a $100 difference in the electric bill.
"Oh boy! Another person harping on energy consumption. Now that the entire Pentium D line will be @ 95W it would take about a decade to even make a $100 difference in the electric bill."
If you live in California and have PG&E, you will sing a different song. I have a big AMD server with two Opteron processors, 8 drives, and 4GB RAM. It uses the same power as a single processor Dell desktop... when the Dell is at idle. If you do anything on the Dell, the Dell uses more power. I've verified this using the Seasonic Power Angel.
Buying a Pentium 4 is like buying a big hulking SUV with today's gas prices: just plain dumb with a bad attitude towards life.
Dude, Hexus.net is a paid Intel pumper doesn't mean all visitors to their site are Intel pumpers.
if Hexus.net was a paid Intel pumper, they would have surely faked this poll. so Hexus.net is probably a paid AMD pumper...
If you have 800 Netburst CPU's in your Datacenter it really makes a difference. Belive me because i have.
Well looking at those benchmarks, seems the 2.13 E6400 with 2MB cache is performiing better than you guys expected, seems to be on par with the FX60.
So long AMD performance crown, just kidding guys, but it is sure getting close.
"Even if Intel were to give away their Pentium processors, over the lifespan of a computer, they still cost more than AMD due to energy usage."
This is completely untrue for the average desktop user. If you assume 4 hours per day usage, 9cents/kilowatt hour and assume now that Pentium D is running at 95W TDP it would take over a decade to equal $100 of electricity difference when compared with an equivalent AMD product. By that time the average PC user would have upgraded twice.
"This is completely untrue for the average desktop user. If you assume 4 hours per day usage, 9cents/kilowatt hour and assume now that Pentium D is running at 95W TDP it would take over a decade to equal $100 of electricity difference when compared with an equivalent AMD product. By that time the average PC user would have upgraded twice."
The majority of PCs are left on all day, not turned on/off for just 4 hours. That is a factor of six right there. So 10 years just became 1.67.
In California with PG&E, there is no such thing as 9 cents per kilowatt hour. High energy costs are real and even states/places that have lower energy costs will see massive changes in the short term. So 1.67 years becomes 0.85 or less years in California.
And finally, Intel's machines usually use more electricity not just due to the CPU, but ancillary chips, fans, less efficient power supply, etc. Intel machines usually have a bad design which optimizes assembly cost at the expense of all other factors, including cost-to-own.
When you add it all up, Intel machines are bad for your pocketbook and bad for the environment.
It is misleading for anyone to quote the TDP of a few new Intel Pentium 4 chips, leaving out all other factors, and say "look, Intel is great!". The customer cannot specify when they buy a machine what Intel processor they are getting. Most likely it is the cheapest oldest crappiest chip from their warehouse... which as we know has been pumped full of Intel stock. Hence, the vast majority of customers will get Intel energy waster chips.
No amount of mendacity can change the fact that Intel is, at its core, criminal organization and a disservice to this world.
"High energy costs are real and even states/places that have lower energy costs will see massive changes in the short term."
Damn, it's back to my solar calculator. I HATE that.
Well guys, if a processor burns 100W, then the power supplies for the processor (which are not 100% efficient) burn a lot of power as well.
So if you have a 100W processor and a 80W processor, the 80W processor saves more than 20W of power. It saves approximately 20+20*(1-efficiency of power converters). Since you have an AC/DC converter and then a DC/DC converter (let's assume each has an efficiency of 80%), a 80W processor would save 20+20*(1-0.8*0.8)=20+20*0.4=28W
Another point you got to keep in mind that energy costs are much more of a problem in data centers with a huge number of servers than in regular homes. The reason why low power for desktops is becoming important (in my opinion) is the fact that people want small form factor desktops with small heat sinks (have you noticed how small Mac boxes are?). The PC is also coming into the living room, and one doesn't want a noisy fan in there, right? Of course, energy costs saved are a nice bonus.
"And finally, Intel's machines usually use more electricity not just due to the CPU, but ancillary chips, fans, less efficient power supply, etc. Intel machines usually have a bad design which optimizes assembly cost at the expense of all other factors, including cost-to-own."
Really. So you can guarantee for me that all AMD PCs use super high efficiency power supplies, and the best fans? Somehow I doubt it. Your Intel experience will obviously vary just as your AMD experience will vary. You can't say it's indicative of the entire product line. As always, the customer must do some research before making there purchase, which is nothing new.
if Hexus.net was a paid Intel pumper, they would have surely faked this poll. so Hexus.net is probably a paid AMD pumper...
Well, can they write PHP code?
here is how to recreate pentium d's appeal:
1. release pentium d 805 for $10 dollars and top of line pentium d 965ee no more than $30.
2. bundle the old (non-conroe compatible ones)for extra $20 dollars.
3. bundle linux distribution that can automatically create HPC cluster with the above combo.
universities are ready to absorb them all for their entry level computer science lab that needs el cheapo linux cluster for such reason, since each node should not cost more than $100 and still provide enough juice.
hingsun
"I have a big AMD server with two Opteron processors, 8 drives, and 4GB RAM."
I believe I was comparing a D 950 to a comparable Athlon not server products. For most home users & gamers the electric bill is negligable.
"I believe I was comparing a D 950 to a comparable Athlon not server products. For most home users & gamers the electric bill is negligable."
Unless you have a T3 at home, downloading 2gb file requires opening the computer for the whole night.
Age group from 25 below will be the one most likely opening their computer 24/7, rather they are too lazy or downloading.
26+ will probally use computer 2-4 hours a day to check out news and some online banking.
Only the 25 below will buy new computer every 2-3 years or so.
Well I am just wondering when everyone here is gonna read the...HKEPC article.
It seems that Conroe temp are quite good, and also its TDP.
How long before someone here finds ways to discredit HKEPC?
On a lighter note it also seems the benchmarks with higher resolutions are quite good, and AMD is still slower... here come the flames...lol
I know K8 is three years old, but if AMD was to come out with a new chip they would benchmark it against what ever there rivals best chip was, so that is a non factor.
Thanks guys.
I say open and fair competition should be everyones mantra and is a win win situation for everyone. However, there are those who maintain that an Intel only x86 market place (can you say monopoly)is the only answer therby allowing Intel to name its price, hence forth driving up prices while allowing it's investors to reap insane profits while stifling all competition and industry inovation. Going forward, a STRONG AMD and Intel is to everyones advantage, INCLUDING YOURS AND MINE.
Just My 2 Cents
That is another benchmark of Woodcrest Vs FX-62. Woodcrest seems ahead of of FX-62 in most benchmarks
I am convinced that Woodcrest has the performance crown in games at this moment. (attn : the memory is Corsair DDR2-800 1GB x 2 (CL5-5-5-15) for FX-62). If AMD wants to retain the performance crown in game, it needs 4X4 now! not a year later. What unknown if Woodcrest could have performance edge in server applications.
http://www.hkepc.com/hwdb/x6800vsfx62-10.htm
The price war is official now. In the next 6 month, what really count is the yield of Woodcrest, and the manufacturing capacity of AMD.
Next six month is going to be a fun for consumers.
AMD will let out “anti- -Hyper-Threading” in this year: the secret weapon Of socket AM2
Lexagon / 18.06.2006 13:24 / reference to the material / version for the press
Still in April this year appeared rumors about the fact that the company AMD develops the technology, reverse on its action of the known technology Hyper-Threading from Intel. Let us recall that the latter makes possible for one processor to substitute by itself two virtual nuclei, and to divide between themselves the flows of data. Intel forewent further development “virtual [mnogoyadernosti]” in favor real [mnogoyadernosti], but it does not exclude the possibility of the revival of this technology in the new performance.
According to our personal information, obtained from the familiar with the plans AMD of sources, already in this year the company will open access to technology, reverse on its action Of hyper-Threading. If the latter divides the resources of one physical processor nucleus, then know-how AMD will make it possible to unite the resources of two physical nuclei for accelerating the tasks, which optimally work on the mononuclear processors. For example, this will make it possible to double a quantity of decoders, and “united” processor will carry out six instructions for time. This already by itself can become not bad answer to the output of the processors Of conroe, and the debut of the corresponding technology for sure takes place on the eve on July 24.
The corresponding functionality is already build it the binuclear processors Of athlon 64 X2 in execution Socket AM2. For its activation it will be required to only renew the driver of processor and basic input-output system of maternal pay. The company Of microsoft must let out appropriate [patch] for the operating systems, which will make it possible to receive two nuclei Of athlon 64 X2 as one. By the way, according to the preliminary data, processor will be able to be switched into “united” regime dynamically, depending on the type of the utilized application. It is no secret that many tasks as before win from the use of productive mononuclear processors, but not working at the lower frequencies binuclear processors.
It is amusing, that AMD itself will be simultaneously occupied by the popularization of the platform 4x4, designed for the enthusiasts, thinkers of its existence without [mnogozadachnosti]. Judging by everything, two sufficient different approach will be spread on the different market niches. In any case, the possibility “to unite” two nuclei into one virtual will manage to the possessors of the processors Of athlon 64 X2 in execution Socket AM2 absolutely free of charge, if we do not consider possible expenditures for the running off of new basic input-output system and driver for the operating system.
It is no secret that the start of the technology Of hyper-Threading sometimes caused reduction in the productivity, since two virtual nuclei began to dispute with the attempt to determine the priority of processing the flows of data. Is possible to assume that the antipode of the technology Of hyper-Threading from AMD will make it possible to avoid such annoying conflicts, and even will increase the operating speed of system.
In proportion to the appearance of new information about this know-how AMD we will acquaint you with it on the pages of our site.
http://www.hkepc.com/hwdb/x6800vsfx62-12.htm
Marginaly faster Core2 Xtreeme than FX62.Memory was CAS 5 which hurted FX 5-8% and the clock diference was ~4.6% in favour of Conroe.If these numbers were taken in acount,FX IS FASTER IN MORE THAN ALL THE GAMING TESTS AND IN APROX. 50% of NON GAMING TEST
I just think that the poll was skewed. There is something fishy when amd wins out by a long shot on all the budget processors and older ones (intel 0%). when that occurs, obviously the next gen will go to AMD as well. I just feel people who buy one type of processor are different than others, thats why intel should have won somewhere. Not that I care, i just think the sample was of only a certain type of people and not indicative of anything.
Electricity prices are ramping up very quickly in the US and around the world.
Take a look at the US averages chart below and you will see why it is dumb to buy a Pentium 4:
DOE Feb 2006 electricity report
Pay careful attention to the NOMINAL price, not the constant dollars curve. From the steepness of the nominal price curve, we can see it is foolish to buy computers with Pentium 4 whose cost to own will dramatically increase from year to year.
Hence for Pentium 4, the smart decision is "pass".
For home user, playing games with intermitent frame rates is more irritating no thanks to Pentium4 with its smart throttling - it's a 3.2GHz chip with hyperthreading perform inconsistently slower than 350MHz Pentium2. Really, Diablo2 jerk so bad in my P4 that my P2 seems so smooth.
I leave my PC on for 24hr. So heater grade PC and vacuum-noise PC are of a major concern.
1. PD 805 to 4GHz w/ water cooler
Bump FSB from 533MHz to 800MHz.
Ordinary DDR2-400 needed
2. PD 930 to 4GHz w/ stock cooler
Bump FSB from 800MHz to 1066MHz.
Ordinary DDR2-533 needed
No hidden costs. Expensive high performance RAM NOT NEEDED. No need for tight RAS/CAS timings (those CL thingies, due to higher than usual memory bus speeds).. Much cheaper than X2 3800+ (and even overclocking it).
Hence.. "basement bargains"!
Btw, a PD930 and PD940 has been known to go past 5GHz on water cooling!
"1. PD 805 to 4GHz w/ water cooler
Bump FSB from 533MHz to 800MHz.
Ordinary DDR2-400 needed
2. PD 930 to 4GHz w/ stock cooler
Bump FSB from 800MHz to 1066MHz.
Ordinary DDR2-533 needed
No hidden costs. Expensive high performance RAM NOT NEEDED. No need for tight RAS/CAS timings (those CL thingies, due to higher than usual memory bus speeds).. Much cheaper than X2 3800+ (and even overclocking it).
Hence.. "basement bargains"!
Btw, a PD930 and PD940 has been known to go past 5GHz on water cooling!"
After you add the cost of cancer treatment to your home brew radiation machine, you will see that you get what you pay for.
Crazy high gigahertz is dumb to begin with. And cranking an Intel chip way up past what the case is designed for is just dumber than dumb.
All those water-cooled mini-reactors giving Intel bargain hunters cancer. Just makes me laugh... what a bunch of losers.
Concerning the link http://www.hkepc.com/hwdb/x6800vsfx62-12.htm
A faster memory and better latency will not help FX62 either (its still limited by K8 architecture and clock speed). Notice some of the margins are pretty big, especially in the gaming apps with XFire? That's total pwnage by Conroe.
"A faster memory and better latency will not help FX62 either (its still limited by K8 architecture and clock speed). "
Wrong. Won't help Core, because its architecture hides memory latency very well. K8, just as K7, loves low-latency DRAM.
"Notice some of the margins are pretty big, especially in the gaming apps with XFire? That's total pwnage by Conroe. "
X6800 2.93 GHz normalized to FX-62 2.8:
3DM'03 104%
3DM'05 109%
3DM'06 108%
Doom3 113%
HL2 108%
BF2 104%
SCCT 106%
Don't get me wrong, I'd love to see 'that' ownage for keeping AMD's EE's afraid and busy, but these results simply don't cut it.
Even a 20% advantage in some apps isn't enough because just as clockspeeds increase K8's memory controller improves on. I'm glad Intel is back, K8 is still very competitive though.
"After you add the cost of cancer treatment to your home brew radiation machine, you will see that you get what you pay for.
Crazy high gigahertz is dumb to begin with. And cranking an Intel chip way up past what the case is designed for is just dumber than dumb.
All those water-cooled mini-reactors giving Intel bargain hunters cancer. Just makes me laugh... what a bunch of losers."
Another idiot talking about things he doesn't know about. Cancer? Really? Wow! Truly dumb.
"All those water-cooled mini-reactors giving Intel bargain hunters cancer. Just makes me laugh... what a bunch of losers."
Thats a pretty bold statement. Not everyone is rich, and going by your statement you seem like an ostentatious prick.
I think everyone out there would like the best performance they can get with the money they have, even us poor people...
X6800 2.93 GHz normalized to FX-62 2.8
you normalized a not released processor? does it make sense to normalize processors concepts? not really...
AMD Machines are verry memory sensitive period! if you dont have good speeds and latencys, it will be slow period! The K8 arch has at least 2 hyper-transport links, wich are full duplex, 1 is for the memory, another is for io. intels procseors have 1 slow halve duplex bus.
If you live in cold countries, buying a pentium 4 will save you heating costs
"If you live in cold countries, buying a pentium 4 will save you heating costs "
Long live, Netburst!
Post a Comment
<< Home