Does anyone else find it frustrating

[kogase]

I'd like to see benchmarks done with CPUs that normal people can buy (read: not Athlon 64 FX-54). The benches aren't helpful if I don't know how it will run on my system, a 2500, which I think many people still use.

 

[flyboy84]

I agree, it would be nice if they showed performance at the "poor man's level." However, keep in mind that for these benchmarks, they are trying to test one particular component (read: video card), so they want to minimize any other bottlenecks. This means they have to use the best processors available for testing. Probably the best solution would be for them to add a little section to reviews about CPU scaling with that particular videocard to give us an idea of how our systems would perform. Note that this would be a huge amount of work for even a small variety of platforms to test, so it's not likely to see this, except in perhaps a large bi-yearly roundup of CPU/VC performance.

 

[kogase]

Perhaps, but they could just bench with the most common mid-range CPUs, like one 2500 and a 2.4 P4. They don't test with a plethora of high-end CPUs, so they needn't do that with mid-range either. Just give us all an idea of how our systems will really run.

 

[JBT]

all you have to do is look at the reviews from when those CPU's were on top.... Also notice it is really only a few people who are interested in looking at older hardware. Most of us want to see the latest and greatest.

 

[Akira1224]

Flyboy hit the nail on the head. AT has stated numerous times they bench using FX chips to remove bottlnecks that would affect performance. When doing a GPU review we are looking to see what the peak performance is. I can understand your desire to see how the hardware will run on a similar config to your own but they need as objective an analysis as possible.

Again I think Fly touched on the fact that when they do roundups they include a much larger spectrum of hardware.

 

[kogase]

Akira1224, ok, good point, but... how about when they do a benchmark for a game review? When Half-Life 2 is realeased, I would like to see it benched witha 9800 Pro (what I'm planning on getting), with an Athlon 2500 (what I have). Surely people would like to know how the game, at least, will run on their machines, if not how the latest video card will.

 

[Algere]

You have a point!

It would give a better basis for personal comparison if the CPU being benched is comparable to what the average user uses and/or can attain. That way it would benefit more ppl who may contemplate about upgrading certain components of their system based on CPU scaling/video card.

OTOH benching with a very high-end CPU allows the video card to stretch it's wings and show off it's performance potential which would normally be hidden/restricted by the speed of the "average" CPU plus it makes the graphs look better & helps convince ppl to buy $400 cards . Another reason to bench with a high-end CPU is that @ the time of a AT review the video card being reviewed is usually a high-end card and ppl (w/ fat wallets) who usually buy high-end cards have CPU's to match.

Then there's the majority of ppl who'll buy a video card when the price meets their budget and by the time a high-end card of today approaches the mid-range pricepoint, the high-end CPU of today will most likely do the same. Knowing that, ppl who ask today why AT reviews video cards with high-end CPU's will ask tomorrow why AT didn't review that same card with a faster CPU.

 

[Marsumane]

Originally posted by: kogase
Akira1224, ok, good point, but... how about when they do a benchmark for a game review? When Half-Life 2 is realeased, I would like to see it benched witha 9800 Pro (what I'm planning on getting), with an Athlon 2500 (what I have). Surely people would like to know how the game, at least, will run on their machines, if not how the latest video card will.

Try tomshardware. He tends to use a more variety of cpus. But u have to think of it this way. The review is geared towards a perticular cpu or graphics card. If it was for a game, they tend to use different configurations. Look at the doom 3 review. They have a review for each type of component that could effect perfromance. They only have so much time. Other sites don't do what anandtech does, and are geared more towards doing other things. U may want to check out other sites also. The majority of the people are satisfied with what anand does. He doesnt cater to 3 yr old archetecture in his "high end fx55 review". He saves that for game reviews. I'm sure ull have enough benches of HL2 to make up your mind on if you should upgrade or not.

 

[kogase]

Algere, they could do it with high-end and mid-range. Like I said, how about doing it with: Athlon64 FX-55, Pentium 4 3.6, Athlon XP 2500, Pentium 2.4. That way, they cater to both ends of the spectrum and everyone in between can fill in the lines. At least until a majority of people are running ultra high-end CPUs (basically, when those are no longer ultra high-end). But most people tend to upgrade their video cards (I think) before CPUs. That leaves a bunch of people with a pretty much useless game review.

 

[DAPUNISHER]

You may be fustrated by this but I am not. Why? Because these reviews are targeted primarily at those who are the most interested in the subject matter i.e. the enthusiasts. Since most of us 'round here were running single channel 3ghz+ NW and 2.2ghz+ XPs approximately 2 yrs ago, the XP3200+ is the low-end.

Owning a 2500+ and not overclocking is your problem. You could have made your system a 3200+ with the proper motivation and research when buying by planning to overclock. Don't overclock? Then don't expect to have a very fast system for a shoestring budget. Besides looking at the dismal performance of the XP3200+ in most those tests, showing a stock 2500+ 333fsb would just leave you depressed instead of fustrated

I am stoked to see Anand back and turning out extensively tested, and exhaustively written articles again, and you come along and start bitching about it. Sorry, but that gets me a bit bent, and I say hit up one of the 2nd rate sites run by hacks if you want to see old hardware included, because I consider it as interesting as watching paint dry :disgust: Besides, if you still have an old 2500+ and haven't already extensively used it with whatever you run to determine how it performs then stop whining and go do it already.

PS, Don't anyone bother telling me how I'm being harsh, or my response is uncalled for, because I don't give a shat

 

[DAPUNISHER]

The other point I'd like to make is that these articles are the most helpful in assisting people planning to purchase in the near future. I doubt most are eyeing a 2500+ since even wal-mart is selling the XP3000+ as a budget system

Everyone can take issue with whatever they like, it is our perogative, and while some will take issue with the article I will take issue with those who take issue with it. I guess some of you just haven't realized how much Anand's work has been missed by some here, well...now you can see it was greatly missed, at least by me.

 

[Killrose]

They are all getting out of touch with their main reader base IMHO. Not too many of us have FX processors and few have A64 compared to XP's and lower end P4's. That one reason I like Firing squad, because they will bench with low end as well as high end stuff.

 

[Zebo]

Originally posted by: kogase
I'd like to see benchmarks done with CPUs that normal people can buy (read: not Athlon 64 FX-54). The benches aren't helpful if I don't know how it will run on my system, a 2500, which I think many people still use.

:roll:

Re read the article. $160 A64 is included. Or is that still too much denero?

Ok then refer back to one of the many articles anand has done it's not that hard. What do you expect him to review from the $44 duron to the $1033 EE??? That's over 70 chips...quite unrealistic IMO. But he's done it anyway for you... just takes some interpolation and research.

Heres your sorry ass two year old 2500 reviewed

 

[kogase]

DAPUNISHER, not everyone has the cash to plunk down for a high-end system. A majority of the system specs I see posted in all manner of forums include an Athlon XP 2500+. That's the way it is. It should not be expected of anyone to overclock. But now that you are getting into it... I can't get my 2500 to run with a 400FSB. I'm running it with a slightly raised multiplier, so it's a 2700. But I don't have the cash to buy a bunch of stuff and HOPE that it runs nice on my system. I have to go to sites like anandtech.com to see if they, who are sustained in part by my interest in their site, will do this hit-and-miss component testing for me. And even most people who are planning to buy a system in the future don't have the money to get really nice gaming components. But they would still like to know how the stuff they can afford will run with newer games, BEFORE they buy the game. It seems perfectly reasonable to me.

 

[kogase]

Zebo I mentioned 4 chips. That's four, the number. Four chips that represent the two extremes that are still commonly used by todays gaming demographic.

 

[Zebo]

Originally posted by: kogase
Zebo I mentioned 4 chips. That's four, the number. Four chips that represent the two extremes that are still commonly used by todays gaming demographic.

Go see hardocps they have the 2.4C and 2500 right there..at the dismal bottom of each chart.

 

[DAPUNISHER]

Originally posted by: Killrose
They are all getting out of touch with their main reader base IMHO. Not too many of us have FX processors and few have A64 compared to XP's and lower end P4's. That one reason I like Firing squad, because they will bench with low end as well as high end stuff.

Without the statistical data to back it up, it's just pure specullation on your part as to who the main reader base is. The majority of those who post in the forums certainly don't leave those older parts@stock speed. What the main site traffic demographics reflect, neither you or I can say, but I'd wager AT gets more traffic than FS, and that would reflect that the testing differences doesn't reflect a lack of audience. Of course that's just my speculation and I qualify it as such.

 

[DAPUNISHER]

Originally posted by: kogase
DAPUNISHER, not everyone has the cash to plunk down for a high-end system. A majority of the system specs I see posted in all manner of forums include an Athlon XP 2500+. That's the way it is. It should not be expected of anyone to overclock. But now that you are getting into it... I can't get my 2500 to run with a 400FSB. I'm running it with a slightly raised multiplier, so it's a 2700. But I don't have the cash to buy a bunch of stuff and HOPE that it runs nice on my system. I have to go to sites like anandtech.com to see if they, who are sustained in part by my interest in their site, will do this hit-and-miss component testing for me. And even most people who are planning to buy a system in the future don't have the money to get really nice gaming components. But they would still like to know how the stuff they can afford will run with newer games, BEFORE they buy the game. It seems perfectly reasonable to me.

I see them to but rarely at stock speeds, even yours' turns out to be mildly overclocked so a 2500+ forces you to still do the math on how your's stacks up, just like with the 3200+. Also, for your stated needs, user results in the forums are far more illuminating for the data you desire. Reading a thread about barton overclocking results with various steppings is far more useful than anything Anand can offer so it seems illogical to rely on or expect him to do it since we all buy from newegg, mwave, ZZF, ect. have far more diverse selection of boards and other hardware, and thus provide far more pertinent info on the topic

 

[Killrose]

Originally posted by: Zebo
Go see hardocps they have the 2.4C and 2500 right there..at the dismal bottom of each chart.

The only thing that keeps me out of the bottom feeder class is the fact that I overclock my 2500+ mobile to 2.4gig

People at least need to overclock if they can't or don't care to buy the high end stuff.

 

[Zebo]

kogase... you relize you can get a A64 setup for about the same price as the ancient XP stuff? A64 combo for $200 with top fight mobo and a64 2800???

This is why it's not included. It's non-sense to be buying a AXP for this type of review, as shown by the hummiliation even the top drawer 3200XP gets. This is a performance review...top of the line. No pintos allowed.. only sports cars. Anand just did a budget CPU shootout even including the lowest end A64 (which dominaited BTW) for your view pleasure I think you should look at.

 

[DAPUNISHER]

Originally posted by: Killrose

Originally posted by: Zebo
Go see hardocps they have the 2.4C and 2500 right there..at the dismal bottom of each chart.

The only thing that keeps me out of the bottom feeder class is the fact that I overclock my 2500+ mobile to 2.4gig

People at least need to overclock if they can't or don't care to buy the high end stuff.

But that's just it, you did the research, found the mobiles were overclocking sweeties, and got some serious bang-for-buck. I believe many who read these sites and on a tight budget do the same thing.

 

[Zebo]

Originally posted by: Killrose

Originally posted by: Zebo
Go see hardocps they have the 2.4C and 2500 right there..at the dismal bottom of each chart.

The only thing that keeps me out of the bottom feeder class is the fact that I overclock my 2500+ mobile to 2.4gig

People at least need to overclock if they can't or don't care to buy the high end stuff.

Oh I enjoy being a bottom feeder and laugh my ass off at people buying $1000 chips. But I overclcok bottom feeder chips to $1000 chip levels...

That's not what OP is talking. He's stocking which leaves all XP, celeron, Duron, and Intel chips below 3.0 out in the performance cold.

 

[Killrose]

Originally posted by: DAPUNISHER

Originally posted by: Killrose

Originally posted by: Zebo
Go see hardocps they have the 2.4C and 2500 right there..at the dismal bottom of each chart.

The only thing that keeps me out of the bottom feeder class is the fact that I overclock my 2500+ mobile to 2.4gig

People at least need to overclock if they can't or don't care to buy the high end stuff.

But that's just it, you did the research, found the mobiles were overclocking sweeties, and got some serious bang-for-buck. I believe many who read these sites and on a tight budget do the same thing.

As far as i'm concerned, if you don't build your own system, do some overclocking/tweaking, the enthusiast sites do not represent your profile ,i.e., integrated graphics/sound/non-overclocking types need not apply.

I think my first overclock was a Cyrix PR200@PR233 speeds, big up-grade from that P133 I had :beer:

 

[Zebo]

PCmag.com

 

[DAPUNISHER]

Originally posted by: Zebo
PCmag.com

I haven't been to their site or read their mag since I got on-line way back and discovered I was still a n00b despite building experience with several platforms. Don't they test a lot of OEM systems for their readers? I would think that is the place for those folks to go.