can an AMD Athlon 2 X2 be snappier than an E8500?

[toyota]

Yeah that's hugely noticable. On this crap Dell 530, on the second double-click the frame of Word appears, and over the next 1/10th of a second or so everything is ready to go. Opening Excel and Powerpoint are equally fast, even if I've left the others open.

Something is horribly horribly wrong with that E8500 setup.

the E8500 setup matched or exceeded comparable systems in every single cpu, gpu or full system benchmark I tested it in. when it comes to little things like opening text documents it is slower than this Athlon 2 pc and even my older 5000 X2 pc.

 

[Arkaign]

the E8500 setup matched or exceeded every single cpu, gpu or full system benchmark I tested it in.

Exactly, and I believe you 100% because I've seen tons of systems over the years with every conceivable configuration (latest example, one of my clients at a law firm has an i5 notebook with Win7 Pro x64 and a FAST ssd that is oddly sluggish in windows, yet completes benches/diagnostics with no oddities) have unacceptably or abnormally slow performance.

Something is just wrong behind the scenes, and often I've found it impossible to nail down. The closest thing I've found to a real fix is moving the stuff to a different mobo, something I figured out long ago. I had an early S754 A64 3000+ that ran like total crap, and I tested moving the chip into a system that was running a Sempron S754 that was much snappier, and the A64 cpu was not the cause, system ran beautifully with the A64 in it. Reloading windows, reloading drivers, trying faster memory, tightening memory timing, updating bios, nothing fixed it. Changed board, reloaded, bam, no problems.

 

[Tsavo]

Something is horribly horribly wrong with that E8500 setup.

+1. I've used similar machines side by side and they were not distinguishable in terms of responsiveness.

 

[nenforcer]

This argument has been made going all the way back to the Athlon XP days when nVidia introduced the DASP (Dynamic Adaptive Speculative Preprocessor) on their nForce northbridge to cache memory accesses and the first dual channel DDR memory controller.

I've always felt there was a noticeable difference in simple things like opening file folders from Windows explorer. Since the Athlon 64 introduced it's own IMC this has been maintained.

This is one reason I've always preferred AMD multi-core architectures and their native design when compared to Intel's brute force fuse two separate dies together.

However, if you read the benchmarks since the Core 2 Duo was released while gaming performance may be somewhat close in certain games there are many benchmarks where Intel just dominates (video encoding). Which is one reason why I bought a C2D E6600.

 

[Gigantopithecus]

First, the IPC of the old X2s was really nothing compared to the IPC of the PhII-derived products like the Athlon II.

The fact that you even mention IPC illustrates that you and I have different ideas of what 'snappiness' means. The question the OP posed isn't subjective, it's objective, once you define snappiness and system responsiveness.

"Snappiness" to me is simple: how long does the computer take to launch an everyday application like MS Word or FireFox? How quickly does it respond to a save file command? How quickly does that program close? The problem is things like that are difficult to measure objectively, but the perception is developed because it is really there.

"Snappiness" is not primarily a function of IPC. It's more a function of memory timings, memory subsystems, and I/O subsystems. This is why SSD-based systems kill HDD-based systems, almost regardless of CPU, in system responsiveness.

This is an old topic of discussion in this forum.

 

[Arkaign]

I understand the snappiness/responsiveness idea, I mentioned IPC because you brought an Athlon II into a discussion that had specified a much older A64 X2, which only makes the situation stranger.

All of the things you describe are no different in correctly configured systems of similar era, and to have a much newer cpu (E8500) running like total crap is just a sign of serious problems somewhere. The Athlon X2-5000 he's using is running correctly, and is snappy and performing up to par. The E8500 is suffering abnormal failure at some point in the configuration or hardware in the system. Even much slower C2Ds can produce near-instant responsiveness throughout the windows experience. That is the bottom line.

To re-state, a system built with a CPU/mobo from the past 5 years, with enough ram and a decent HDD, will produce basically instant results in windows common tasks (open word, open browser, save average file, open winamp, whatever). If it doesn't, it's broken somehow, or has a piece of hardware that runs poorly.

 

[toyota]

Exactly, and I believe you 100% because I've seen tons of systems over the years with every conceivable configuration (latest example, one of my clients at a law firm has an i5 notebook with Win7 Pro x64 and a FAST ssd that is oddly sluggish in windows, yet completes benches/diagnostics with no oddities) have unacceptably or abnormally slow performance.

Something is just wrong behind the scenes, and often I've found it impossible to nail down. The closest thing I've found to a real fix is moving the stuff to a different mobo, something I figured out long ago. I had an early S754 A64 3000+ that ran like total crap, and I tested moving the chip into a system that was running a Sempron S754 that was much snappier, and the A64 cpu was not the cause, system ran beautifully with the A64 in it. Reloading windows, reloading drivers, trying faster memory, tightening memory timing, updating bios, nothing fixed it. Changed board, reloaded, bam, no problems.

I just think most people would never notice the difference. the E8500 did that from day 1 but if there was nothing else to compare it to then it seems normal. its slightly slower speed when opening a text document was 100% consistent and reproducible. it was the only thing I remember feeling different after using a 5000 X2 for 2 years.

now even on this bloated oem pc, the text documents open quicker. so lets see, two X2 machines that both open the text documents quicker and one E8500 that was a tad slower. its not scientific but I am thinking the E8500 system was just slower at that and it was perfectly normal. I will be getting on a Core 2 machine within 24 hrs so we will see.

 

[AnandThenMan]

Not in my experience when you compare processors of the same era (obviously comparing i7 against Athlon X2, or PhII against ~2006ish C2D E4000-series is uneven).

http://www.anandtech.com/show/2045/5

A good engineer goes beyond specs and numbers. I can show you tests that "prove" almost anything, doesn't change the end user experience. BTW, that article has many flaws in it, as was pointed out by the readers.

 

[Arkaign]

Most people probably wouldn't, but I know you're one of the types like me that gets annoyed at sluggish systems. I have gone through hell replacing components in 100% working units to get acceptably snappy/instant responsiveness, and have run into this with all kinds of combinations in CPU/mobo/OS/HDD/etc.

I can tell you with confidence that any properly working system from the past 5 years (I should exclude special low-power/performance stuff like Atom/etc) should produce no serious lag in simple windows tasks and apps. Literally the longest thing you should see is the Windows animation routine if you have even left that enabled. A full second or close to it to open Word is inexcusably bad.

 

[Arkaign]

A good engineer goes beyond specs and numbers. I can show you tests that "prove" almost anything, doesn't change the end user experience. BTW, that article has many flaws in it, as was pointed out by the readers.

Oh, so you're telling me that AMD CPUs are authoritatively faster than Intel CPUs in 'snappiness' based on your overwhelming credentials as a 'good engineer'? And you cast slander on Anand's testing methods?

The bottom line, yet again, is that a properly working system with either an Intel or AMD processor should NOT be sluggish in any way. That's it, end of story.

 

[nenforcer]

This is why it's important to follow the benchmarking done by sites like AT since they typically do clean installs of Windows every time before testing new hardware.

Also don't discount the effect that L2 and L3 cache play on processor performance.
I've used a Pentium 4 2.8 GHz Celeron CELERY side by side with a Pentium 4 2.8 GHz regular and the performance gap was clearly evident. The CELERY machine was slower to boot and load / close applications. Now once you got the application going it was snappy because of the clock speed but everything outside of that was chugging.

 

[AnandThenMan]

Oh, so you're telling me that AMD CPUs are authoritatively faster than Intel CPUs in 'snappiness' based on your overwhelming credentials as a 'good engineer'? And you cast slander on Anand's testing methods?

Some of the testing methods in that article were flawed, go back and read the article. That doesn't make the entire article bad.

The bottom line, yet again, is that a properly working system with either an Intel or AMD processor should NOT be sluggish in any way. That's it, end of story.

This makes me wonder if you've ever used a variety of systems at all, let alone "10's of thousands" as you claim. The Athlon X2's were more responsive vs. the competition at one point, I saw it time and time again. And feedback showed the same thing, the systems "felt faster". The term sluggish is not the right term anyway, just less responsive. I don't generally like car analogies, but it's like two cars with the same horsepower, one has better throttle response vs. the other, so the car feels faster and more responsive.

Don't get all butthurt over it, it was the reality back then.

 

[Arkaign]

This is why it's important to follow the benchmarking done by sites like AT since they typically do clean installs of Windows every time before testing new hardware.

Also don't discount the effect that L2 and L3 cache play on processor performance.
I've used a Pentium 4 2.8 GHz Celeron CELERY side by side with a Pentium 4 2.8 GHz regular and the performance gap was clearly evident. The CELERY machine was slower to boot and load / close applications. Now once you got the application going it was snappy because of the clock speed but everything outside of that was chugging.

Yeah the Netburst Celeron, similar to the Atom, was/is just too slow to produce results competitive with mainstream midrange processors of the era.

The cache indeed had a HUGE impact on those Celerons, as the netburst architecture's long pipeline was still there, but with only 128k of cache (which IIRC was also crippled in a couple of other ways as well), there were tons more misses than with the 512k of northwood. This produced incredibly bad results, something like the 2.0 P4 usually handily outperformed Celerons that had 50% higher clocks.

 

[Arkaign]

Some of the testing methods in that article were flawed, go back and read the article. That doesn't make the entire article bad.

This makes me wonder if you've ever used a variety of systems at all, let alone "10's of thousands" as you claim. The Athlon X2's were more responsive vs. the competition at one point, I saw it time and time again. And feedback showed the same thing, the systems "felt faster". The term sluggish is not the right term anyway, just less responsive. I don't generally like car analogies, but it's like two cars with the same horsepower, one has better throttle response vs. the other, so the car feels faster and more responsive.

Don't get all butthurt over it, it was the reality back then.

Okay, we'll take a step back and start over. First, they tested fine in that article, unless you really want to nitpick, but the results can't really be argued with in any meaningful way. Second, if you are talking Athlon X2 (or Athlon 64 Single-Core even) against Pentium 4 / Pentium D, then I 100% agree. If you're talking equal-era Athlon X2 (let's say 4400+ for the heck of it), vs. an equal-era C2D (let's call it E6400), then you should have no responsiveness issues with either unit at all. If you do, it's a problem. And I'm one of the worst offenders at being nitpicky. I used to disable all windows animations in every system I used for any period of time because I couldn't stand it, reducing the menu draw timing to zero, etc. Even back in the 9x days I was using regedit or tweakui to turn off anything that brought additional delay in something happening.

 

[toyota]

like I said earlier even Anandtech mentioned in some article that AMD cpus and the newer i7 cpus were indeed more responsive than Core 2. heck if I can find the article since I never imagined I would need it and google is failing me.

 

[Arkaign]

like I said earlier even Anandtech mentioned in some article that AMD cpus and the newer i7 cpus were indeed more responsive than Core 2. heck if I can find the article since I never imagined I would need it and google is failing me.

That would indeed be interesting, and I'd be the first to admit that I was wrong if there's some authoritative proof. It seems incredibly doubtful though, one must remember the era right around the C2D's launch date, there were a TON of AMD fanboys, including me, simply because what they had was so definitively better. After C2D starting seriously drawing blood and winning basically every performance benchmark, the loyalists to the company rather than just the best performance would have jumped ALL OVER something that proved that the X2s were faster/snappier. Other than some ad-hoc commentary by end users, I never remember seeing anything of the sort. Maybe at AMDZone you could find many people that would agree with that.

Like I said, I personally had an old E5200 that was supernaturally quick in windows, and it was with a cheap mobo/ram/hdd/etc as a 2nd system. I've seen similarly superquick X2s, even in Socket 939 form with DDR1. As in, virtually instant everything in windows until you start loading something that takes considerable HDD access (say launching a large game). But menus, opening simple apps, etc, have no delay at all beyond say 1/10th of a second.

As I've seen over time though, I've seen a ton of systems of every conceivable type which just weren't fast/snappy at all, from Athlon 64s to i5 sandy bridge laptops with SSDs. Simply too many variables.

But reducing it to this :

Have you ever seen a C2D system that produces no noticable delay in common tasks? If yes, then it's impossible to say that a C2D CPU alone should produce poor results unless something is broken/poorly configured, or mated to hardware that is reducing performance.

As I've seen personally systems with C2Ds, Core Duo, Athlon 64, Opteron, Athlon 64 X2, Core 2 Quad, Phenom I, Phenom II, Athlon II, i3, i7, etc, etc all perform with beautiful snappiness (I'm picky as hell), I can't agree with blaming it on a CPU, unless it's broken or throttling.

Perhaps more Intel-based motherboards suffer from performance issues than AMD ones, it wouldn't reflect my real-world experience but I couldn't put it past the realm of possibility.

 

[apoppin]

can an AMD Athlon 2 X2 be snappier than an E8500?

We *always knew* something was wrong with your system from some of the nonsense that you posted in video; always generalizing about your CPU and minimum frame rates.

:\

 

[Tsavo]

The Athlon X2's were more responsive vs. the competition at one point, I saw it time and time again. And feedback showed the same thing, the systems "felt faster". The term sluggish is not the right term anyway, just less responsive. I don't generally like car analogies, but it's like two cars with the same horsepower, one has better throttle response vs. the other, so the car feels faster and more responsive.

I have an old P4 2.4 GHz Northwood that felt snappier than any X2 system I built or used. And it wasn't just me either; others used it said the same thing. It defies explanation and logic, but there it was.

I also had a Phenom II very briefly and it felt like a complete pig next to all my other boxes. I attributed it to a crappy chipset and returned it instead of swapping parts until I was blue in the face.

On the other hand, I've got an AMD M300 which doesn't feel any different than any of my other boxes.

I'd wager that any unresponsiveness is due to bad/flaky BIOS/Motherboard/Chipset than CPU.

 

[toyota]

We *always knew* something was wrong with your system from some of the nonsense that you posted in video; always generalizing about your CPU and minimum frame rates.

:\

thanks for popping in and acting like a smartass and trying to bait. I guess you missed the other part where I said my system matched or exceeded benchmarks for comparable systems. and I guess my 5000 X2 system had problems too sense that was the system I first started checking for cpu bottlenecks. seriously if all you can do is try and start crap then stay the heck out of this thread.

and I doubt this will shut you up but I just left Best Buy where they had an E6800 system and it did the SAME EXACT THING as my E8500 did. just a slight hesitation when opening up a text document compared to this X2 system. they even had an X2 a couple of comps down from the E6800 system and the difference was clear. its hardly a big deal in the whole scheme of things but it certainly shows that I was not crazy and the difference is there.

 

[yottabit]

I have to kind of agree with the OP. I built a similar system with an Athlon II x2 and an old WD 250gb drive from ages ago. It is very snappy- similar in response to my i5 750. I think it is not so much an outright performance thing but something about the AMD cpus that make them hold up better to scheduling lots of crap. I honestly *feel* like the AMD system has retained its snappiness as the system has been loaded up with software, etc better even than my i5 rig. Of course I have no way of proving that though

I also have an occasional *hanging* for a split second issue in Starcraft II even on the lowest settings with my C2D laptop, but that could be due to a lack of CPU power in general.

 

[Arkaign]

Again, individual examples of systems is not indicative of the situation as a whole. If it were true that there was a 'hesitation' on all C2D systems that was not there on all 'X2' systems, then there would be tons of videos on it, articles on it, the folks at fanboy places like AMDZone would have been in a frenzy for years over it.

Yes, there will be some systems, regardless of CPU type, that will have sluggishness or even slight hesitations where none should exist. The FACT that I've seen ridiculously snappy systems using almost every type of processor imaginable makes this ludicrous, as it removes the CPU from the equation (so long as it's not some super dumpy POS like an Atom, etc). What does that leave? Individual configurations, of which there are literally thousands of variables. You can get as granular as you want, even down to things as minute as individual chip revisions on the mobo, as well as bios settings that are impossible to access as they are untouchable by the user.

 

[apoppin]

thanks for popping in and acting like a smartass and trying to bait. I guess you missed the other part where I said my system matched or exceeded benchmarks for comparable systems. and I guess my 5000 X2 system had problems too sense that was the system I first started checking for cpu bottlenecks. seriously if all you can do is try and start crap then stay the heck out of this thread.

and I doubt this will shut you up but I just left Best Buy where they had an E6800 system and it did the SAME EXACT THING as my E8500 did. just a slight hesitation when opening up a text document compared to this X2 system. they even had an X2 a couple of comps down from the E6800 system and the difference was clear. its hardly a big deal in the whole scheme of things but it certainly shows that I was not crazy and the difference is there.

Doing what you do on an open public forum? Commenting, Criticizing and Questioning your methodology? Don't be so sensitive just because it is happening to you; or the "way" it is being done. i am not being a smartass, i am simply making an *observation* that i believe to be true.

Just because benches "match" does not mean that everything is set up perfectly - especially by your past posts in Video that showed your CPU to be consistently giving some pretty sub-par performance in some game minimums with little change when you upgraded the video card.

And you also made a funny .. "popping in"

OK, i'll make another couple of observations:

You cannot use your even older 5000 system as any good example for anything - "feelings" are difficult to measure; especially looking back year later.

And now you are comparing the setting up of your 8600 PC to a 6800 set up by a Best Buy clone with the way it "feels" again. They might just be using out of date drivers. You are digging yourself deeper and really clutching at straws with that one imo.

 

[notty22]

Again, individual examples of systems is not indicative of the situation as a whole. If it were true that there was a 'hesitation' on all C2D systems that was not there on all 'X2' systems, then there would be tons of videos on it, articles on it, the folks at fanboy places like AMDZone would have been in a frenzy for years over it.

Yes, there will be some systems, regardless of CPU type, that will have sluggishness or even slight hesitations where none should exist. The FACT that I've seen ridiculously snappy systems using almost every type of processor imaginable makes this ludicrous, as it removes the CPU from the equation (so long as it's not some super dumpy POS like an Atom, etc). What does that leave? Individual configurations, of which there are literally thousands of variables. You can get as granular as you want, even down to things as minute as individual chip revisions on the mobo, as well as bios settings that are impossible to access as they are untouchable by the user.

Agreed, its in part why we see motherboard reviews of 5 p67 boards.

also I had spent over 4 hours playing Fallout 3 on my E8500 pc and it always would hitch every few seconds when walking around no matter what settings I used. well on this Athlon 2 X2 pc there was none of that when I played a couple days ago. it would only hitch when it had to load that new area but the other time no hitching at all. that was on the same version of the game and video card drivers.

and no that was not in my head because I spent hours just trying to figure out why it was hitching on my old pc and this played perfectly smooth in the same exact spots.

This alone should add enough to the equation that there is possibly something wrong, and its not 'logical' to declare it platform related.

Windows GUI environment can be effected by a plethora of hardware and software combinations.
I recall one cpu review where the author felt the system had excellent snappiness, and in his mind, he attributed this to possibly hyperthreading. Of course that does not mean that a system without hyperthreading would not feel as snappy.

 

[toyota]

Again, individual examples of systems is not indicative of the situation as a whole. If it were true that there was a 'hesitation' on all C2D systems that was not there on all 'X2' systems, then there would be tons of videos on it, articles on it, the folks at fanboy places like AMDZone would have been in a frenzy for years over it.

Yes, there will be some systems, regardless of CPU type, that will have sluggishness or even slight hesitations where none should exist. The FACT that I've seen ridiculously snappy systems using almost every type of processor imaginable makes this ludicrous, as it removes the CPU from the equation (so long as it's not some super dumpy POS like an Atom, etc). What does that leave? Individual configurations, of which there are literally thousands of variables. You can get as granular as you want, even down to things as minute as individual chip revisions on the mobo, as well as bios settings that are impossible to access as they are untouchable by the user.

and again you are trying to be way too specific. from what I have seen the X2 systems open the text document quicker compared to Core 2 architecture cpus. we are talking a minuscule amount of difference here but it is indeed snappier. its not something hardly any person would ever know unless doing it back to back. if I was on various comps of all different types the subtle differences might not be noticed between just two systems. I was 100% accustomed to how my E8500 handled doing that for 2.5 years though so it was something I noticed immediately since I use a lot of text documents throughout the day.

 

[toyota]

Doing what you do on an open public forum? Commenting, Criticizing and Questoning your methodology? Don't be so sensitive just because it is happening to you; or the "way" it is done. i am not being a smartass, i am simply making an *observation* that i believe to be true.

Just because benches "match" does not mean that everything is set up perfectly - especially by your past posts in Video that showed your CPU to be giving some pretty sub-par performance in some game minimums.

And you also made a funny .. "popping in"

OK, i'll make another couple of observation:

You cannot use your even older 5000 system as any good example for anything - "feelings" are difficult to measure; especially looking back year later.

And now you are comparing the setting up of your PC to one set up by a Best Buy clone with the way it "feels" again. They might just be using out of date drivers. You are digging yourself deeper and really clutching at straws with that one imo.

anybody can see your intentions here and I don't play that passive aggressive crap. that's the kind of stuff that girls do just to stir up crap and then claim to be the victim if someone retaliates.

all I was doing was mainly pointing out a subtle difference here in how the X2 responded to normal daily stuff. no matter how small the difference it is there and is noticeable. so going up there and seeing the exact same difference between another X2 and Core 2 system shows it was nothing to do with my pc. if you don't accept that then fine.