Case Proven: People that think X2 > Core2 clock for clock

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

imported_Crusader

Senior member
Feb 12, 2006
899
0
0
Originally posted by: dmens
haha, you still didn't explain your quality metric. dilettantes like you wouldn't know quality engineering even if it got jizzed all over your face lol.

Wow. Jizzing in peoples faces ect.. how old are you? You're an idiot. I'm not going to even bother. Good luck in life. :disgust:

Originally posted by: zsdersw
Hey, if Williamette or Prescot is your idea of "quality engineering", then more power to you. Unfortunately for yourself and the other Intel apologist, you're dead wrong.. but thats ok, you are entitled look as dumb as you want to look.

You'd better get on the horn and warn Intel to not follow in AMDs footsteps and integrate an IMC.
What does that say about your proclamations of the IMC being "inferior"? It says they're crap.

Oh, I could not.. on my worst day.. come close to looking as dumb as you already look. Show me where, exactly, I said the IMC is inferior. You can't.. because I didn't say it was inferior. I said that it was not enough to make AMD's chips faster than Conroe/Merom/Woodcrest. I also said that in most cases (as 4P+ servers/workstations are few and other servers, desktops, and laptops are many) the IMC and HyperTransport are irrelevant because their advantages over a FSB are not realized in systems with 1 CPU. That's not saying the IMC and HyperTransport are inferior.. it just says their capabilities and benefits over a FSB model are not needed most of the time.

You implied it was inferior here (at least as much as "I" implied what you thought you heard earlier in the thread.. fair is fair)-

Originally posted by: zsdersw
The IMC (integrated memory controller) and HyperTransport, as nice as they are, don't win the race for current AMD chips in 1P and 2P servers.. and are totally irrelevant in the desktop and mobile market, as their advantages over the FSB model don't show in systems with 1 CPU.

Their advantages MUST show or Intel wouldnt be implementing one onto C2D.
Thats my whole point, it is the better/superior way to go. So what, C2D beats A64 on other merits? Doesnt mean your arguments about "FSB being good enough" or being better hold any water.
Because they dont. IMC is superior, AMD knew it.. blew Intel away with that being part of their architechure.. and now Intel is working on it.

As an aside- Theres no denying Intel has fantastic engineering in their Israeli division (makes sense why they wouldnt go to American "experts" like dmens LOL). They did just enough borrowing of ideas to not look like they are "following the leader" but everyone knows thats what happened when Intel finally gave up on (and practically split internally) over Netburst.

I'm impartial between the two, but Intel elitists and their arrogance is disturbing. I guess I'd be defending Intels decisions if I dumped (or in most peoples cases that dont use their PCs to make money... wasted) all that cash on C2D as well.

But I'm always up for defending AMDs great engineering and engineering choices. Which Intel followed more in the lines of with Conroe. I like both, Intel employs more Americans so I support buying from them actually a tad more. But AMD is still an American based company. Thats how I choose what I buy beyond price/performance, all things otherwise being equal.. who employs the most Americans.
People can argue what in reality are slight differences in performance between chips.. but jobs are real. At least for those of us who are adults here.

I personally cant justify the move to C2D till all games are multithreaded and things are bit mature. A $139 FX55@3ghz fits the bill till that day comes, which could be a few years.
 

imported_Crusader

Senior member
Feb 12, 2006
899
0
0
Originally posted by: dennilfloss
Why do you care so much about what others buy? :confused: Other people buying A64 X2s does not make your Conroe CPU more/less desirable. If one's CPU does what one expects at a price one finds reasonable, all is well for that user. Most other stuff amounts to just e-penis comparison and may be important to members with little self-esteem but not to the vast majority of computer buyers who use and see their machine as a tool to do a job (job can be playing games or productivity work, crunching distributed-computing work units, etc...) Stop comparing yourself to others and judge your CPU performance+ price with regards to your own needs.:beer:

This is some Truth right here.

And what I was getting to in my last post here-
I personally cant justify the move to C2D till all games are multithreaded and things are bit mature. A $139 FX55@3ghz fits the bill till that day comes, which could be a few years.

For a gamer, who is already on 939.. and the FX57 being the fastest single core ever produced by anyone.. its tough to bail on my rig when I got a FX55 for $139.

For me, and 90% of the people here whose most intensive thing beyond being in the Windows desktop is gaming.. it would make the most sense.
Price + Performance in regards to my own needs indeed! :thumbsup: :beer:
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,284
16,123
136
Guys... I just got a 6300 C2D and a miserable OC to 2 ghz, but just for kicks, I ran superpi (no comments, I know its crap).. @ 2m the 6300@2 did 1.56 and my X2 3800 @stock 2 ghz did 1:48. Then I ran 1M, the 6300 did 44 rec, and the X2 did 47 sec. So in my case they are very close, and I think the cache made the C2d win@1m. Anyway, they are very close.sometime one wins, sometimes the other does.
 

dmens

Platinum Member
Mar 18, 2005
2,275
965
136
Originally posted by: Crusader
Wow. Jizzing in peoples faces ect.. how old are you? You're an idiot. I'm not going to even bother. Good luck in life. :disgust:

old enough to appreciate fine sexual humor. and you still haven't explained your quality metric yet.

IMC is superior, AMD knew it.. blew Intel away with that being part of their architechure

and conroe blew k8 away with the fsb being part of its uarch. crikey!

They did just enough borrowing of ideas to not look like they are "following the leader" but everyone knows thats what happened when Intel finally gave up on (and practically split internally) over Netburst.

damn, "everyone" knows a lot about intel internal politics and design inspiration. i should go talk to this guy...
 

coldpower27

Golden Member
Jul 18, 2004
1,676
0
76
Originally posted by: Markfw900
Guys... I just got a 6300 C2D and a miserable OC to 2 ghz, but just for kicks, I ran superpi (no comments, I know its crap).. @ 2m the 6300@2 did 1.56 and my X2 3800 @stock 2 ghz did 1:48. Then I ran 1M, the 6300 did 44 rec, and the X2 did 47 sec. So in my case they are very close, and I think the cache made the C2d win@1m. Anyway, they are very close.sometime one wins, sometimes the other does.

This is very misleading. In your case you only ran 2 items, and without giving the rest of the hardware used on each platform this is not a worthwhile comparison. 2 situations is hardly is representative of the picture in general.

It's more like most of the time Core 2 Duo will win given the same available clock frequency, there are only handful of situation where you maybe able to find greater IPC from a K8 and they are rare when you use a large testing suite.

As well no biasedness toward either platform, both platform should have the best RAM available to it.
 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: dexvx
Originally posted by: Crusader
My point was, if you did buy the AMD (or Intel) either way you wont get a crap chip.

For the past years, with the exception of Northwood, Intel has been selling crap chips.
Theres nothing pathetic about using a X2 5000+ unless you are a rabid intel fanboy elitist. Still a great, fast chip..

You're saying its ok to overpay for a chip that:

1) Does less in the performance/watt category (upwards of 40-60% less).
2) Does less general performance (upwards of 20-30% less than a E6600 at stock).
3) Costs upwards of 50% more.

Yet, how is this different when you were buying "crap" Intel chips? I mean P4 Prescotts/Cedar Mills made *ok* machines (they work well and are reasonably fast compared to previous generation AthlonXP's and S754 A64's), but they fell into the same category as #1, #2 but NEVER #3 (except for EE's).

Originally posted by: atom
So, since when is AMD getting a cut from retailer markups? And how is this different from Intel? Intel makes a killing on the EE chips, AMD makes a killing on their FX chips, no surprises here.

I think Viditor can answer this, but its also based on distributor demand. Obviously since retailers like Newegg are selling out at inflated prices, the distributors aren't being retarded and not taking notice. Hence the distributors can mark it up. Since AMD sells directly to the distributors, they are idiots for not noticing the series of markups as well. Also I'm talking about the AM2-5000+, not even an FX chip.

A long answer coming...beware! :)

Firstly, dexvx is quite correct that for a DIY system, using a high-end X2 makes no sense whatsoever! So who the heck is buying the 5000+ chips (obviously they are selling somewhat, or AMD would drop the price)?

The short answer is small businesses...

1. When you are a small business (say 6-12 systems for various jobs in the company), you have a choice between a Dell/HP system or a locally made white box. The Dell/HP will almost always be less expensive on these deals, but they come with a BIG caveat...quality of service.
For example, if you have a HDD failure on a Dell, they will fix it quickly (depending on your service contract) but that means you get a new HDD and any data you had on the old HDD will be lost (Dell does NOT do data recovery at this level). The local white box guy will do everything in his power to recover your data and customize your system for your needs. For many companies this can mean a savings of thousands of dollars (depending on the data that wasn't backed up). Even if you back up your work every day, you've lost a day's work from whoever was on that system.

2. When a small business orders a number of systems, one of the most common practices is to buy a site license from Microsoft based on the number of seats. They also tend to create a single image including their common software needs...
This means that they require a single type of motherboard to allow for easy substitution and install (i.e. they install everything on one machine then clone the HDD across to all of the rest). The motherboard choice is actually the most important one in this case as it must match the needs of the company (HDTV, PATA/SATA drives, USB/Firewire, audio, on-board video, type of raid, etc...). Of course they can use different CPU speeds/memory amounts and still swap out a HDD from and to any other system in the place (which gets the guy with the most critical job a guarantee that he will be back up in minutes should his system fail).

3. So, it then comes down to which platform is best suited for the client at the best price (CPU is very much secondary). The typical scenario here would be a single mobo with a wide range of CPUs (say 5 Semprons, 4-X2 3800s, and 1-X2 5000 for the boss).

4. Enter the "Certified Business Platform" (Nvidia's for AMD chips) and "vPro" (Intel's) programs...
This certifies motherboards, chips, memory, software, etc... for those small businesses from a variety of manufacturers. It allows a more equal footing between OEMs and white box manufacturers in that it gives a broad range of testing and qualification standards for companies to help them decide on a solid platform which they can build.
AMD (and more specifically Nvidia) has certainly come a VERY long way on this in the last year (of course they had a much longer way to go vs Intel)! I would say that the 2 programs are now fairly neck-and-neck in capability and reputation...for vPro, the impression is that they have a lead in remote manageability, and the CBP has a better rep for multimedia (in general).
There are other minor differences like CBP boards are a little better in backwards compatability (they have more slots for PATA drives for example), and the vPro boards tend to incorporate ICH8 so they will have better raid functionality (including Raid 5).
These minor differences are what makes choices for businesses (do they need to use some of their older PATA drives? Do they want raid 5 incorporated? Do they do multimedia? Do they have an IT person who will manage all of the systems remotely?)

At the end of the day, Crusader is quite correct that it doesn't really matter for the vast majority of users if they have an X2 or a C2D, either will be just fine...what matters most are the platforms (including reliability and functionality) and the package price for all the systems.

Of course for almost everyone here, it matters a great deal!

BTW, to answer your question on the rationale of the 5000+ pricing, that's the "carrot" for these packages...if a white box maker can throw in a "high-end" chip that has a very high retail value at a huge discount for a package (normally for the boss's machine), then it can quite often close a sale. And remember that we're talking about a 6-12 machine deal...
AMD marketing knows this quite well (as does Intel), and that is the real value of these ultra high-end priced chips...
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,284
16,123
136
Originally posted by: coldpower27
Originally posted by: Markfw900
Guys... I just got a 6300 C2D and a miserable OC to 2 ghz, but just for kicks, I ran superpi (no comments, I know its crap).. @ 2m the 6300@2 did 1.56 and my X2 3800 @stock 2 ghz did 1:48. Then I ran 1M, the 6300 did 44 rec, and the X2 did 47 sec. So in my case they are very close, and I think the cache made the C2d win@1m. Anyway, they are very close.sometime one wins, sometimes the other does.

This is very misleading. In your case you only ran 2 items, and without giving the rest of the hardware used on each platform this is not a worthwhile comparison. 2 situations is hardly is representative of the picture in general.

It's more like most of the time Core 2 Duo will win given the same available clock frequency, there are only handful of situation where you maybe able to find greater IPC from a K8 and they are rare when you use a large testing suite.

As well no biasedness toward either platform, both platform should have the best RAM available to it.

The A64 system is an SFF using DDR400, and all setting@stock.

The C2D is using DDR2 667@284 (the best this motherboard would go) @1:1, so the memory is better than the motherboard can even use. without every piece being equal, its impossible to make a 100% comparison, but I thought this was insightful, as (at least in my case) at the same clock, they are almost equal IN THIS CRAP BENCHMARK.

I am not saying this will allways be the case, but I think that this is at least valuable information...

Also, the X2 3800 is on an Nvidia chipset, and the C2D 6300 is on a via (P4M800PRO-M motherboard)

If I have a problem with my system, all input is welcome, as I would love to speed this puppy up ! I have only had it for 3 hours, and it blows the hell out of the 805@3.5 that I had in the same case with the same memory, way more responsive, just like my X2's.

Edit: Interesting note: Both systems were running 2x F@H at the time, so that was equal. When I killed the F@H on the C2D, then 1m time went down to 40 seconds from 46 (new drivers) The X2 went from 47 down to 44 sec, so not as much difference. Even more interesting, not as much affect....

Edit2: BTW, I realize that the C2D is on a $50 motherboard, but many of my X2's are also. The comparison here, the SFF box is using a 240 wat power supply, and the HSF that comes with the SFF ! I still think the motherboards for the C2D's are very immature, and exspensive compared to X2 motherboards, but then again, they did just come out. Just more data I am sharing...
 

imported_Crusader

Senior member
Feb 12, 2006
899
0
0
Viditor excellent post! :thumbsup: :beer:
I've been wondering how the NV CBP has been doing, havent heard to much myself since it was announced.
It will be very interesting to see how CBP and the AMD/ATI "business platform" works out? I'm guessing just more options open for AMD setups.. ultimately making AMD business platforms more attractive?

Originally posted by: Viditor
At the end of the day, Crusader is quite correct that it doesn't really matter for the vast majority of users if they have an X2 or a C2D, either will be just fine...what matters most are the platforms (including reliability and functionality) and the package price for all the systems.

Yeah, for example my FX57 argument presented in this thread actually applies to even 90% of enthusiasts that spend most of their time gaming. I like new, fun hardware to.. but I cant spend the cash right now to upgrade for the sake of upgrading (my G80 is far more important to get ;)). As long as I get the max out of my GPUs, thats all that matters.
IMO its a shame there isnt a high clocked single core Conroe out yet.. because that would be really attractive.. and talk about a wipeout to AMD if the prices are right..
low priced, high clocked single cores.. maybe unlocked? A dream to good to come true.

For people making money with their PC doing some serious daily video editing ect, C2D makes sense of course. There arent many of those around statistically, and of course the minority of C2D owners.
The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.
Huge disadvantage is that dual core can be incredibly detrimental to the gaming experience.. some of those scenarios detailed by Brent Justice@HardOCP. So I'm wary till things progress.
But the old "you have the right to spend your money on whatever the hell you want" is good enough reason to buy a Conroe, and thats most peoples reason.

Its pretty tough to justify in reality, esp for a gamer already on 939 (with $139 FX55s around).
But I've spent/lost more money and gotten less out of it than a C2D for sure! :p Gambling, beer and women can be hazardous to the bank accounts!! :D ;) :beer:

By the time K8L is out, maybe 80% of games will be multithreaded (thats a longshot, but I'm hopeful) and I can have faster options, better multithread support, and :thumbsup: better prices :thumbsup: for my dual core move.
 

zsdersw

Lifer
Oct 29, 2003
10,505
2
0
Their advantages MUST show or Intel wouldnt be implementing one onto C2D.
Thats my whole point, it is the better/superior way to go. So what, C2D beats A64 on other merits? Doesnt mean your arguments about "FSB being good enough" or being better hold any water.
Because they dont. IMC is superior, AMD knew it.. blew Intel away with that being part of their architechure.. and now Intel is working on it.

Your reasoning is flawed. As the results clearly indicate, an IMC and Hypertransport are not required to achieve better performance in desktops, laptops, and 1P and 2P servers. It will be some time before Intel's CSI will trickle down to those markets, probably around the time when the multi-core march and blurring between CPU and GPU take shape. This is likely to be a time when CSI makes the most sense for those markets.
 

zsdersw

Lifer
Oct 29, 2003
10,505
2
0
Originally posted by: Crusader

The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.

There you go being an idiot again...

Gaming performance is not a good measure of CPU performance, as it is mostly a GPU-limited thing at the resolutions people typically play the games. At low resolutions, though, C2D is faster than AMD chips. Those low resolutions are the only way to gauge the CPU's impact on the performance equation in games.

 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: zsdersw

Your reasoning is flawed. As the results clearly indicate, an IMC and Hypertransport are not required to achieve better performance in desktops, laptops, and 1P and 2P servers. It will be some time before Intel's CSI will trickle down to those markets, probably around the time when the multi-core march and blurring between CPU and GPU take shape. This is likely to be a time when CSI makes the most sense for those markets.

That's quite misleading Z...
While I absolutely agree that IMC and HT aren't required to achieve better performance, those same results show that a uA is much better off with them than without them. The question becomes what you are willing to do to incorporate them into your design...
I personally feel that Intel has made a mistake in their NIH (Not Invented Here) policy as far as that goes, and I have a feeling it may cost them later next year. (JMHO)
 

coldpower27

Golden Member
Jul 18, 2004
1,676
0
76
Originally posted by: Markfw900
Originally posted by: coldpower27
Originally posted by: Markfw900
Guys... I just got a 6300 C2D and a miserable OC to 2 ghz, but just for kicks, I ran superpi (no comments, I know its crap).. @ 2m the 6300@2 did 1.56 and my X2 3800 @stock 2 ghz did 1:48. Then I ran 1M, the 6300 did 44 rec, and the X2 did 47 sec. So in my case they are very close, and I think the cache made the C2d win@1m. Anyway, they are very close.sometime one wins, sometimes the other does.

This is very misleading. In your case you only ran 2 items, and without giving the rest of the hardware used on each platform this is not a worthwhile comparison. 2 situations is hardly is representative of the picture in general.

It's more like most of the time Core 2 Duo will win given the same available clock frequency, there are only handful of situation where you maybe able to find greater IPC from a K8 and they are rare when you use a large testing suite.

As well no biasedness toward either platform, both platform should have the best RAM available to it.

The A64 system is an SFF using DDR400, and all setting@stock.

The C2D is using DDR2 667@284 (the best this motherboard would go) @1:1, so the memory is better than the motherboard can even use. without every piece being equal, its impossible to make a 100% comparison, but I thought this was insightful, as (at least in my case) at the same clock, they are almost equal IN THIS CRAP BENCHMARK.

I am not saying this will allways be the case, but I think that this is at least valuable information...

Also, the X2 3800 is on an Nvidia chipset, and the C2D 6300 is on a via (P4M800PRO-M motherboard)

If I have a problem with my system, all input is welcome, as I would love to speed this puppy up ! I have only had it for 3 hours, and it blows the hell out of the 805@3.5 that I had in the same case with the same memory, way more responsive, just like my X2's.

Edit: Interesting note: Both systems were running 2x F@H at the time, so that was equal. When I killed the F@H on the C2D, then 1m time went down to 40 seconds from 46 (new drivers) The X2 went from 47 down to 44 sec, so not as much difference. Even more interesting, not as much affect....

Edit2: BTW, I realize that the C2D is on a $50 motherboard, but many of my X2's are also. The comparison here, the SFF box is using a 240 wat power supply, and the HSF that comes with the SFF ! I still think the motherboards for the C2D's are very immature, and exspensive compared to X2 motherboards, but then again, they did just come out. Just more data I am sharing...

I would also like to know the memory timings, I suggest you get a better motherboard for the Core 2 Duo, that motherboard is just to cheap. Which Nvidia chipset is the X2 3800+ that you tested against running on?

Try the ECS P965 for 94US on Newegg, which can support up to DDR2-800.

Well considering it's a super low end VIA motherboard from ECS I am not too surprised on it's lackluster overclock.

I still think the information is misleading, as like I said in most cases, given the same clock and an Intel 945P or higher motherboard, preferably a 965 Core 2 Duo > X2 per clock.

Obviously you can't make everything equal, but if your going to use a VIA chipset for Intel use a VIA chipset for AMD then. If your going to use the perfered brand for AMD aka a Nvidia or ATI based motherboard. Then use the flagship brand for Intel which is Intel themselves.

You can generally keep most factors constant. I am not saying all, but most.

Like for example the RAM, if your using DDR400 3-4-4-8 then DDR2-533 4-4-4-12 is acceptable, or CAS 5 DDR2-667 or DDR2-800. JEDEC timings for both RAM types is fair.

If it's DDR-400 @ 2-2-2-5 then I would accept, DDR2-800 4-4-4-12 minimum, or CAS3 DDR2-533/DDR2-667.



 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: Crusader
Viditor excellent post! :thumbsup: :beer:

My humble thanks!

I've been wondering how the NV CBP has been doing, havent heard to much myself since it was announced.
It will be very interesting to see how CBP and the AMD/ATI "business platform" works out? I'm guessing just more options open for AMD setups.. ultimately making AMD business platforms more attractive?

It's just one example of why those who are predicting the split of Nvidia and AMD need to do more research...
The CBP program is a HUGE effort from AMD and Nvidia, which could easily double the number of Nvidia chipset-based systems over the next year. And remember that in this case, it's Intel vs everyone else...the only chipsets for vPro systems are Intel's.


Originally posted by: Viditor
At the end of the day, Crusader is quite correct that it doesn't really matter for the vast majority of users if they have an X2 or a C2D, either will be just fine...what matters most are the platforms (including reliability and functionality) and the package price for all the systems.

Yeah, for example my FX57 argument presented in this thread actually applies to even 90% of enthusiasts that spend most of their time gaming. I like new, fun hardware to.. but I cant spend the cash right now to upgrade for the sake of upgrading (my G80 is far more important to get ;)). As long as I get the max out of my GPUs, thats all that matters.
IMO its a shame there isnt a high clocked single core Conroe out yet.. because that would be really attractive.. and talk about a wipeout to AMD if the prices are right..
low priced, high clocked single cores.. maybe unlocked? A dream to good to come true.

C2D's only real weakness right now is volume...they just don't have the production volume for a 4MB single core version. In fact, one of the largest C2D Fabs (D1D I believe) is already pulling back on production in order to begin conversion to 45nm for next year.

For people making money with their PC doing some serious daily video editing ect, C2D makes sense of course. There arent many of those around statistically, and of course the minority of C2D owners.
The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.

Well, let's be clear here...you're talking about a very specific segment of the market (and not a very big one at that). In general (and overall in fact), the C2D is a much superior chip to the K8. The point of my post was that this really doesn't matter as much for the CPU market in general. So, for the enthusiast and gamer, the C2D is the only choice right now (unless your budget is constrained or you are upgrading an existing AMD system)...but as far as AMD vs Intel goes, the choice lies more in which platform is best. You should also keep in mind (as it's easy to forget with all of the benchmarks flying around) that 80% of what Intel ships is still the old Netburst design.

Huge disadvantage is that dual core can be incredibly detrimental to the gaming experience.. some of those scenarios detailed by Brent Justice@HardOCP. So I'm wary till things progress.
But the old "you have the right to spend your money on whatever the hell you want" is good enough reason to buy a Conroe, and thats most peoples reason.

Its pretty tough to justify in reality, esp for a gamer already on 939 (with $139 FX55s around).
But I've spent/lost more money and gotten less out of it than a C2D for sure! :p Gambling, beer and women can be hazardous to the bank accounts!! :D ;) :beer:

By the time K8L is out, maybe 80% of games will be multithreaded (thats a longshot, but I'm hopeful) and I can have faster options, better multithread support, and :thumbsup: better prices :thumbsup: for my dual core move.

I agree...be it Intel or AMD, we will continue to see better and faster improvements as time goes on.
 

deadseasquirrel

Golden Member
Nov 20, 2001
1,736
0
0
Originally posted by: Crusader
For a gamer, who is already on 939.. and the FX57 being the fastest single core ever produced by anyone.. its tough to bail on my rig when I got a FX55 for $139.

For me, and 90% of the people here whose most intensive thing beyond being in the Windows desktop is gaming.. it would make the most sense.
Price + Performance in regards to my own needs indeed! :thumbsup: :beer:

But if you truly want to dig into price/performance for gaming, then even the $139 FX55 isn't such a great deal. Take into account the popular Oblivion, for example. My guess is that 1280x1024 is a widely-used resolution, and, as you can see, gaming performance is exactly the same from 2.0ghz to 2.8ghz at that resolution. Which would make the $55 3000+ quite a bargain at almost 1/3 the cost.

I agree with zsdersw in that regard. Gaming performance isn't a good measure of CPU power. Calling a particular processor a "gaming processor" is about the same as labeling some low latency ram "gaming ram".
 

zsdersw

Lifer
Oct 29, 2003
10,505
2
0
Originally posted by: Viditor
those same results show that a uA is much better off with them than without them.

How so? The excess memory bandwidth that K8 has (but doesn't come close to fully utilizing)?

Show me where in those results it says that the desktop, laptop, and 1P-2P server markets are better served in the performance department by chips that feature an IMC and HT. Show me how HT yields better performance than a FSB in a single-CPU server, desktop PC, or laptop featuring a Woodcrest, Conroe, or Merom chip, respectively.

Having an IMC is not without its cons, too.. you know.

 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: zsdersw
Originally posted by: Viditor
those same results show that a uA is much better off with them than without them.

How so? The excess memory bandwidth that K8 has (but doesn't come close to fully utilizing)?

Show me where in those results it says that the desktop, laptop, and 1P-2P server markets are better served in the performance department by chips that feature an IMC and HT. Show me how HT yields better performance than a FSB in a single-CPU server, desktop PC, or laptop featuring a Woodcrest, Conroe, or Merom chip, respectively.

Having an IMC is not without its cons, too.. you know.

I have a feeling that you don't understand the difference between memory latency and bandwidth...

Memory Latency = The time between initiating a request for a character in memory until it is retrieved
Bandwidth = the maximum amount of information (bits/second) that can be transmitted along a channel

While it's true that even the K8 has more than enough Bandwidth, it's performance is very sensitive to Latency (for example DDR400 is about equivalent to DDR2-667 in performance because the latency of DDR2 is much higher).
Memory performance is a combination of bandwidth and latency...

So, are you saying that you don't believe that the latency is much lower with an ODMC?
 

Munky

Diamond Member
Feb 5, 2005
9,372
0
76
The main reason C2D is faster than K8 is because it's a 4-issue processor, and the IMC cant overcome the 3-issue limitation of the K8. However, if another cpu comes out that can decode and execute as many instructions as the C2D, but also has an IMC, which one will most likely have the performance lead? The IMC has helped the K8 a lot against the P4, because the K8 core is not much different from the K7, and those were getting slaughtered by the later model P4's. Intel does not need an IMC now due to a wider core, but when the K8L arrives, that might change. With everything else being equal, IMC > FSB.
 

dexvx

Diamond Member
Feb 2, 2000
3,899
0
0
Originally posted by: zsdersw
Originally posted by: Crusader

The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.

There you go being an idiot again...

Gaming performance is not a good measure of CPU performance, as it is mostly a GPU-limited thing at the resolutions people typically play the games. At low resolutions, though, C2D is faster than AMD chips. Those low resolutions are the only way to gauge the CPU's impact on the performance equation in games.

Thats the most retarded statement by Crusader.

In GPU limited benchmarks, you'd also get the same FPS with a higher end Pentium-D system. So by transition, is Crusader suggesting that Pentium-D = A64 X2 = Core2 Duo? Basically, yes.

If you extend Viditor's arguement, it will also apply easily to Pentium-D (which isn't even that bad after C1 revision) as being a perfectly viable alternative, which Crusader has hand-picked out as "poorly engineered."
 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: dexvx

In GPU limited benchmarks, you'd also get the same FPS with a higher end Pentium-D system. So by transition, is Crusader suggesting that Pentium-D = A64 X2 = Core2 Duo? Basically, yes.

If you extend Viditor's arguement, it will also apply easily to Pentium-D (which isn't even that bad after C1 revision) as being a perfectly viable alternative, which Crusader has hand-picked out as "poorly engineered."

I agree (though I have intentionally left out the name-calling portions...).
And Pentium-D certainly is a perfectly viable alternative, in fact it's by far the market leader!
That said, the biggest market-share gainer is currently AMD. Much of the reason for this is a successful (for a change) marketing campaign by AMD on the downsides of the Netburst Architecture...it also helps that Netburst is associated with older technology.
Those of you who feel that AMD has a terrible marketing division have been looking in the wrong spot. AMD's marketing is extremely successful in the areas that make them the most money...business!
 

Conky

Lifer
May 9, 2001
10,709
0
0
I thought it was already established that Crusader(who many suspect to be a Rollo alias) is not only an Nvidia fanboy but an AMD one as well.

His assertation that you can easily get an FX-55 for $139 that is guaranteed to run at 3GHz is fairly comical too. I'd much rather roll the dice on an $180 E6300 and it's overclocking potential but this is still off topic...

Anyone who is capable of building a new system by themselves would never pick a high-end high-dollar AMD cpu at this point unless they were, 1)wealthy to the point of not caring about value, and 2)stupid enough to think the expensive high-end AMD CPU's were the fastest that could be had.
Either way, they would have to be misinformed in one way or another. (Cue OCHungry, lol)
 

zsdersw

Lifer
Oct 29, 2003
10,505
2
0
Originally posted by: Viditor
I have a feeling that you don't understand the difference between memory latency and bandwidth...

Memory Latency = The time between initiating a request for a character in memory until it is retrieved
Bandwidth = the maximum amount of information (bits/second) that can be transmitted along a channel

While it's true that even the K8 has more than enough Bandwidth, it's performance is very sensitive to Latency (for example DDR400 is about equivalent to DDR2-667 in performance because the latency of DDR2 is much higher).
Memory performance is a combination of bandwidth and latency...

So, are you saying that you don't believe that the latency is much lower with an ODMC?

No, that's not what I'm saying at all.. and yes, I know the difference between latency and bandwidth. I'm annoyed that you presumed I didn't know it.

Why did the switch from DDR to DDR2 not benefit the K8 very much? Because of its higher latency. What matters most in memory latency? The timings. What do the timings depend on? How the memory was designed and what the memory standard is capable of.

In spite of the lower memory bandwidth and presumed higher latency of the C2D with its FSB, it soundly beats K8 in nearly every test. What does that say about an IMC? It says that CPU architectures can be designed to raise the performance bar without an IMC. Would Core 2 Duo be better off with an IMC? I doubt it. It's performance doesn't depend on memory latency or bandwidth as much as competing chips and the FSB is far from being an Achilles heel for it in most market segments.
 

Griswold

Senior member
Dec 24, 2004
630
0
0
Gee, you really have to compensate for something if you feel the need to make a thread like this. Poor guy. :disgust:
 

imported_Crusader

Senior member
Feb 12, 2006
899
0
0
Originally posted by: dexvx
Originally posted by: zsdersw
Originally posted by: Crusader

The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.

There you go being an idiot again...

Gaming performance is not a good measure of CPU performance, as it is mostly a GPU-limited thing at the resolutions people typically play the games. At low resolutions, though, C2D is faster than AMD chips. Those low resolutions are the only way to gauge the CPU's impact on the performance equation in games.

Thats the most retarded statement by Crusader.

In GPU limited benchmarks, you'd also get the same FPS with a higher end Pentium-D system. So by transition, is Crusader suggesting that Pentium-D = A64 X2 = Core2 Duo? Basically, yes.

If you extend Viditor's arguement, it will also apply easily to Pentium-D (which isn't even that bad after C1 revision) as being a perfectly viable alternative, which Crusader has hand-picked out as "poorly engineered."

I left the insults intact to shine poorly on you guys, not on myself.
When you grow up, you'll understand. I'm not offended.

In response to that cherry picked statement out of my post (while everythign else was left alone, which tells me you just want to argue and are annoyed someone sticks up for the merits of AMD tech)-
All I said was:

For people making money with their PC doing some serious daily video editing ect, C2D makes sense of course. There arent many of those around statistically, and of course the minority of C2D owners.
The proof of A64=C2D really does lie in the fact that an Athlon gets the same FPS as a C2D.

Yes, its a small part of the market. But its the MAJORITY of the market looking at early adoption of C2D.
If we took a poll of who earns their living entirely off their desktop PC where shaving off a few seconds on some arcane benchmark could possibly matter, the results would be rather slim.
When the majority of people here use their PCs for games, forums and basic Windows apps..maybe use Wildfire + some programming apps.. the upgrade doesnt make sense.
In gaming C2D=A64 when you are GPU limited.

Are you seriously disputing this? Because you argued with me, calling me names.. you know its the fact of the matter.. but you still want to argue? :roll:
 

zsdersw

Lifer
Oct 29, 2003
10,505
2
0
Originally posted by: Crusader
I left the insults intact to shine poorly on you guys, not on myself.
When you grow up, you'll understand. I'm not offended.

In response to that cherry picked statement out of my post (while everythign else was left alone, which tells me you just want to argue and are annoyed someone sticks up for the merits of AMD tech)-

I'm sorry you're not offended. Better luck next time, I suppose.

I picked out that statement of yours because it's the only thing that I disagreed with.

I'm not annoyed that you stick up for "the merits of AMD tech". I'm annoyed that you simultaneously criticize, wrongly, Intel tech by making idiotic statements about "quality engineering", etc.

Originally posted by: CrusaderYes, its a small part of the market. But its the MAJORITY of the market looking at early adoption of C2D.
If we took a poll of who earns their living entirely off their desktop PC where shaving off a few seconds on some arcane benchmark could possibly matter, the results would be rather slim.
When the majority of people here use their PCs for games, forums and basic Windows apps..maybe use Wildfire + some programming apps.. the upgrade doesnt make sense.
In gaming C2D=A64 when you are GPU limited.

Are you seriously disputing this? Because you argued with me, calling me names.. you know its the fact of the matter.. but you still want to argue? :roll:

Since when does the "early adopters of C2D" market have anything to do with the broader discussion about how one CPU stacks up to another? It is but one element in the discussion.. and not a terribly conclusive one at that.

As far as the name-calling is concerned, when it looks like a duck and sounds like a duck, chances are it's a duck. I call 'em as I see 'em. I'm not going to apologize for that.. and if you or anyone else has a problem with it, T.S.