Please register or login. There are 1 registered and 3063 anonymous users currently online. Current bandwidth usage: 326.30 kbit/s November 25 - 05:33am EST 
Hardware Analysis
      
Forums Product Prices
  Contents 
 
 

  Latest Topics 
 

More >>
 

    
 
 

  You Are Here: 
 
/ Forums / AMD's price reductions, a fruitless exercise?
 

 
 Author 
 Date Written 
 Tools 
Continue Reading on Page: 1, 2, 3, 4, 5, 6, 7, Next >>
Janick Bernet Jul 19, 2006, 07:46am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 09:28am EDT

Replies: 125 - Views: 17268
What a cheap rant. Intel is finally back in the lead since years, and of course with "the most impressive new processor architecture to date", although its not a new or even inovative architecture at all but just very good optimization of the P3 line. But the important thing is, that all the Intel fanboys can now get rid of all their accumulated anger by writing articles as this one.

Intel may be ONE of the driving forces behind many new technologies; mainly because they have the money and the status to do so. AMD is a much smaller company, they cannot build on three different architectures like Intel (Itanium, NetBurst, Core) hoping one of them will succeed. But they still have the better bus system than Intel, which they introduced years ago already.

AMD was in the lead for the last two years at least, Intel may be for a year, as the new K8L architecture will not launch this year. But if they manage to be on top for the next two years, then, and only then, you can start thinking about writing your flaimbait.


Want to enjoy fewer advertisements and more features? Click here to become a Hardware Analysis registered user.
Sander Sassen Jul 19, 2006, 08:06am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
I'll ignore the flamebait comments for now, as a column is intended to be opinionated by nature. You mentioned a better bus used by AMD, but even that doesn't hold true on all counts. Yes, the bandwidth is higher, but the latency as well. Anybody will tell you that a lower latency is always preferable over raw throughput, in that aspect the Core 2 processors have lower latency than AMD's for memory access. Anand has a few benchmarks to show this in his article, you might want to give that a read.

Best regards,

Sander Sassen
Editor in Chief - Hardware Analysis
ssassen@hardwareanalysis.com
Dublin_Gunner Jul 19, 2006, 08:50am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Well I think you summed it up quite well Sander, I dont see how that article could be called 'flamebait' or whatever the above poster said.

The comments above are more 'flamebait' than anything Sander wrote, and if you think for one second that K8l is gonna take the performance crown back, you're gonna be sorely disappointed.

Its merely a re-hash of the current K8 architecture, with some small optimisations here and there, and a different way of implementing the HTT, which will in no way help CPU or RAM performance.

Yes HTT is a great interconnect, but you dont see the advatntages of it until you use multi processor systems, as all it does is basically give more bandwiadth the I/O peripheral hardware on desktop systems.

Core architcture, although is a re-working of the P3 and Pentium M architecture (albeit with MANY enhancements), is a very good design indeed, and will remain at the top of the performance tree for some time yet.

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Stephen Rayburn Jul 19, 2006, 08:55am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
AMD is holding an impressive marketshare for the desktop and enthusiasts market. I'm not partial to either myself (whatever get's the job done). Nice that Intel finally came out with a fast processor, however it's still an uphill battle to regain their lost marketshare which they let slip over the years; further compounding this, is price point: if you already hold a large share in a market, lowering overall costs to the end-user simply re-enforces this, and further expands brand recognition, AMD is also clearing out the old to make way for the new, it's a tried and true strategy, and Intel follows it also, nothing but smart.

IMHO speeds for the processor's these days are hardly something worth mentioning and a lame-duck arguement, anything right off the shelf today runs all the games and apps out there . It's the motherboard manufacturers we need to go after, to force them to come up with radically new designs that will allow theoretical throughput to become factual throughput for memory, CPU, and VPU; putting the speed issue back into the laptops of Intel, AMD, ATI, and Nvidia. PCIe is hardly an answer...

By all appearances this article did seem skewed towards Intel, and could've been
written from a more balanced viewpoint (however, still informative, and thus a good read).

Incidentally, don't be surprised when AMD comes to market much earlier than anyone expects with their answer to Intel, it should prove to be quite a competition, and maybe it'll give Anand something fresher to write about. =)

Dublin_Gunner Jul 19, 2006, 09:03am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
By all appearances this article did seem skewed towards Intel, and could've been
written from a more balanced viewpoint (however, still informative, and thus a good read).



Thats exactly how it was meant to seem, Sander is no fanboy, the article was driven by the fact that the curresnt situation is unbalanced.

BTW, on you point about market share & AMD reducing prices to get rid of old stock. Why would anyone now chose an AMD based system, that not only performs far less, but costs a lot more than a new Core 2 Duo based one?

The only people AMD are going to sell those chips to are people looking for a quick and cheap upgrade to their s939 systems. I dont see many users on forums around the web with AM2 based systems, do you?


And why's that? Because everyone is holding out for a Conroe, its that simple. Unfortunately, I fear AMD are going to lose a LOT of the market share that they gained with K8, and unless they pull something out of the hat, this is going to really hurt their stock value.

BTW, I am a fan of neither company, as I've statd many times, I'm a fan of whoever offers the best performace for the price. Which right now is Intel.

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Anthony Lau Jul 19, 2006, 09:21am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 

Janick Bernet Jul 19, 2006, 09:21am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 09:24am EDT

 
>> 
I wasn't aware of the fakt, that this was a column, mainly expressing the authors own point of view. Anyway, I still think the whole article was very subjective. It was written, as if Intel made a brakethrough never achieved in the history of personal computers. They are not much more ahead of the Game as AMD was with their AMD 64 architecture vs the Pentium 4 (not so much Performance wise, but in a general rating of the architecture).

Regarding the price issues: At least here in Switzerland I get a E6600 for about 450.- swiss francs and a Athlon 64 X2 for 430.-, which should be in the same performance range.

Regarding the bus latency, IMO latency is mainly affected by clock speed and Intels is clocked at 1066 Mhz, while AMD's is clocked at 1000 Mhz; so i don't see much advantage there. Also HTT ist a much more flexible protocol, as it can be used to innterconnect multiple different devices in your computer. Also, the on-die memory-controll, pared with HTT still beats the core's memory bandwith (which is, I admit, e very synthetic number): http://techreport.com/reviews/2006q3/core2/index.x?pg=4.

Also I still think that your points raised about innovation (PCI-Express, etc.) and AMD milking theyr K8 architecture is a very mute one. As I said, AMD just isn't in the same position as Intel which was the marked leader for years and has a huge influence still. Also they have not the resource to experiment with three architectures in parallel. IMO it is not at all a time to celebrate Intel that one of them suceeded, but it would have been very bad if none if them did.

R James Jul 19, 2006, 09:32am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 09:44am EDT

 
>> 
This article didn't sound one sided to me it seemed to be based on current trends in the market. Intel is however a bigger company than AMD with a bigger budget.

I'm happy that Intel bought out this CPU and I'm an AMD fan. I'm mostly happy because it means a much needed pricing shake up on the market which is badly needed is about to happen. I'm tired of seeing $1000 CPUs and $1000 dual video cards pushing a PC up to $2000 before you even purchase any of the other components. Especially with Direct X 10 around the corner and all current GPUs being Direct X 9 based.

I'm kind of straying off point here... AMD will still have sales and with the pricing cuts people looking to upgrade or new system vendors will always offer both (apart from the Intel only ones anyway). Currently I'm looking to upgrade my computer and I may go the Intel route then again I've had great experiences with AMD ever since the Thunderbird and that counts a lot when I go to purchase my hardware. Besides as somebody else said in this post anything you pick up off a shelf or online these days will run whatever you throw at it well. My current system does my video conversion and other tasks requested with no problem at all.

The article was a good one, no matter what a lot of forums with Intel and AMD heads are going to be filled with argumentative nonsense now.


__________________________________________________________________________
Current System:
Asus K8V-SE Motherboard
AMD Athlon 64 3200+
Sapphire ATI Radeon 9800pro 128mb
1024MB PC3200 Memory
200GB 7200RPM Western Digital SATA HD
Ben Q 8X DVD/RW
LG 16X DVD ROM



Sander Sassen Jul 19, 2006, 09:33am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Skewed or not, the point of my columns is to provide fuel for further discussion so we're all better informed about future purchases we might be making. I'm in no way stating that my arguments are the be all end all, I'm merely providing a few of them that are provoking enough to discuss further. But clearly some of you raise a few good points as well, and only AMD knows what's going on exactly, I'm sure we'll all agree anything we discuss is (educated) guesswork at this point.

Best regards,

Sander Sassen
Editor in Chief - Hardware Analysis
ssassen@hardwareanalysis.com
Dublin_Gunner Jul 19, 2006, 09:48am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Regarding the bus latency, IMO latency is mainly affected by clock speed and Intels is clocked at 1066 Mhz, while AMD's is clocked at 1000 Mhz


The comment was regarding memory latency, of which the HTT has nothing to do with.

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Anthony Lau Jul 19, 2006, 09:52am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 09:54am EDT

 
>> 
I agree with you completely there, only AMD knows for sure whats the next step. As for the discussion being aimed towards only the article, well let me address that. First off, I didn't think of HTT as a protocol but as bus technology... what I fould interesting was the increased latency with AM2 vs. Core 2, but even so, the bandwidth available to the FX 62 is FAR above Core Extreme, although AMD seems to benefit little from it thus... maybe in the Opteron arena it is more of a performance factor and of course in 4x4 or dual sockets, the direct interconnect of multiple (though physically seperated) core via the additional HT lanes vs. Via the FSB or otherwise (as on intel) is a better route. I've read that HT3 is due soon and Intel could really benefit from using it as CSI is delayed and that FSB really isn't cutting it anymore. HTT is developed by a community as opposed to a single body and as such Intel could IFF (if and only if) they wanted tom but I guess from a PR perspective it wouldn't fly well. One more thing, my understanding of the HT bus was 1GHz unidirectional or 2GHz bidirectional, not a simple 1 to 1 comparison with a 1066MHz FSB.

Shadow_Ops_Airman1 Jul 19, 2006, 09:58am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 10:01am EDT

 
>> 
Intel made the mistake of going Netburst even under comparison a 1.4 Ghz P3 would whoop up on a 1.4 Ghz P4, All Intel Can do now is continue making major Optimizations to a Already Great Architecture. Lets see the Core 2 Duo Continue for a few generations. K8L is just 2 Windsor Cores Smooshed together in a Package, Like what Smithfield is. What AMD really needs to rethink, is about the memory controller being built directly into the CPU, thats what makes the CPU Price jump way high and its causing AMD Headaches. Also letting the Chipset manufacturers put the memory controller on the mobo will allow for more flexibility (SKT AM2 to 939 adapter and vice versa)

AMD Athlon XP-M 2500+ (133x14= 1867MHz) (209x11= 2299MHz)
DFI LP NF2 Ultra-B (Hellfire 3EG Rev2)
Antec SX800, Neo HE 500, 4 Antec 8CM Fans
Thermalright SI-97 1 Antec Tricool 12CM Fan
CL SB XFi Xtreme Music
2x Barracuda HDs (250/400)
2x Samsung Write
Dublin_Gunner Jul 19, 2006, 09:58am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
re you talking about memory bandwidth? If so, the HTT has nothing to do with it. The memory bus connects directly to the memory controller on the CPU die. More badwidth doesnt translate into better performance on K8 based systems, due to the memory controller, the CPU already has enough bandwidth for any given clock speed.

On Intel systems this is different, as data from memory has to compete for bandwidth over the FSB, but in the case of Conroe, this doesnt seem to make much of a difference, as can be seen with the latency/bandwidth benchmarks.
The FSB still seems to offer enough throughput for single CPU systems.

As was stated, where the HTT shines is in the multi - cpu server environment.

BTW, I thought Intel cancelled the CSI interconnect and actually was considering licensing the HTT (or a variation) for their own systems.

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Shadow_Ops_Airman1 Jul 19, 2006, 10:08am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
they May License it, but for their other Interconnect, not the FSB. I was just thinking on how to get HTT working more efficiently but it would make Motherboard prices Skyrocket.

AMD Athlon XP-M 2500+ (133x14= 1867MHz) (209x11= 2299MHz)
DFI LP NF2 Ultra-B (Hellfire 3EG Rev2)
Antec SX800, Neo HE 500, 4 Antec 8CM Fans
Thermalright SI-97 1 Antec Tricool 12CM Fan
CL SB XFi Xtreme Music
2x Barracuda HDs (250/400)
2x Samsung Write
guru Shane Jul 19, 2006, 10:11am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List

Edited: Jul 19, 2006, 10:11am EDT

 
>> 
By reading this thread, it's easy to see who are AMD fanboys... I can certainly applaud what AMD has done, but they are not the ones currently shining. You can talk about how the HTT is better all you want, but it doesn't currently hold any foundation.

I would also like to take this time to laugh at the people whining about how AMD is a small company and can't afford to work on three chips at once. AMD made 5.8 billion dollars in 2005... -_- It has much more to do with ingenuity than it has to do with the amount of money you spend or the amount of people that you have working on it. Honestly, you guys should already know this considering how much Intel has put into development over the past few years.

Dublin_Gunner Jul 19, 2006, 10:12am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Airman, the randomness of your posts never fails to amuse me!

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Dublin_Gunner Jul 19, 2006, 10:15am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
BTW, 9to the new members posting here) did you join HWA to comment on the article alone? Just curious.


Also, welcome to HWA.

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Anthony Lau Jul 19, 2006, 10:15am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
This excerpt is addressed to the person that stated K8L is just 2 cores bolted on (ala intel fashion), it's an article based on the Keynotes of AMD's Chuck Moore about the K8L:
Quote:
"First, it has a shared expandable L3 cache, necessary because it is a native quad-core design. The one massive enhancement to the mix is that AMD finally has the ability to independently change core voltages for power savings. It now can also change the north bridge voltage independently of the cores. This is a huge win, we are told voltage differentials and problems with them were one of the main scaling headaches of the K8 core to this point.

Next is memory. The new core will support 48-bit addressing and 1GB pages. Cray and SGI will be very happy with this, until they hit that memory wall again. There is also official co-processor support, strongly hinted to be on a HTX card. The key here will be the platform is aware of them vs having to hack them in.
The other whopper Chuck dropped was that DDR2 is coming and DDR3 is in the wings when the spec 'settles down'. Old news, FB-DIMMs are the future, right? AMD has said they are supporting them, but the big news is that they are not forcing support. Unlike Intel's approach, Blackford supports only FBD, AMD will let you choose. This seems to strongly suggest that the controller on the later gens will be quite flexible indeed.

Next up is RAS, another area where AMD is sorely lacking. It is addressing the major sore points with support for memory mirroring, data poisoning support, and HT retry. It looks like it is following the IBM roadmap more than the Intel one here.

IPC is also going up in a big way. It is doing the obvious doubling of SSE/FP resources, old news now, but it goes a lot deeper than that. There are a bunch of added instructions, starting with the bit manipulation instructions LZCNT and POPCNT. It also added SSE extensions EXTRQ/INSERTQ and MOVNTSD/MOVNTSS. No word on SSE4 though.

The last bit is much more aggressive prefetch to 'feed the beast'. It has gone from 16B to 32B, an obvious step with the added SSE number crunching power. On top of this, it has out of order loads, and other tweaks to use the available bandwidth in a much more efficient manner.

For those who thought K8L was more or less a tweaked K8, you are wrong. It looks like no part of the core has been left unmolested by the elves working the CAD stations. It looks like AMD will have a credible response to the Intel MCW architecture after all. 2007 will be a fight after all.
"
That should shed some light on what K8L brings to the table.



Dublin_Gunner Jul 19, 2006, 10:20am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Where did you get that article? I remember reading it before but I cant remember where!

Lancool PC K62
Phenom II x3 @ 3.5Ghz
4GB DDR2-800
ASUS GTX570 DirectCU II
Shadow_Ops_Airman1 Jul 19, 2006, 10:24am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
Sounds all good on paper, lets see AMD pull this one off. Course by that time Intel will have their Quadcores that dont repeat the mistake of Smithfield, and also tweak Core 2 with additional features.

AMD Athlon XP-M 2500+ (133x14= 1867MHz) (209x11= 2299MHz)
DFI LP NF2 Ultra-B (Hellfire 3EG Rev2)
Antec SX800, Neo HE 500, 4 Antec 8CM Fans
Thermalright SI-97 1 Antec Tricool 12CM Fan
CL SB XFi Xtreme Music
2x Barracuda HDs (250/400)
2x Samsung Write
Anthony Lau Jul 19, 2006, 10:25am EDT Reply - Quote - Report Abuse
Private Message - Add to Buddy List  
>> 
@ Guru Shane, 5.8B is great, but how are they spending it? as stated before, for all intents and purposes its plain to see they are trying to boost manufacturing capabilities in a BIG way (ergo Fab36 and plans to move into NY). I doubt building a FAB cost pennies. I do agree that the overall capabilities of a processor has alot to do with the ingenuinity of the engineers, but who do you work for if you're the einstein of CPU design? the upstart company? or the industry leader who will pay you more? money talks man. 'Honestly you should know that', its how the world turns unfortunately.


Write a Reply >>

Continue Reading on Page: 1, 2, 3, 4, 5, 6, 7, Next >>

 

    
 
 

  Topic Tools 
 
RSS UpdatesRSS Updates
 

  Newsletter 
 
A weekly newsletter featuring an editorial and a roundup of the latest articles, news and other interesting topics.

Please enter your email address below and click Subscribe.