What I think about [Hard]ocp!

[ArchAngel]

I've read through half of this thread and all of it seems a bit pointless.

If you don't like the site, that's fine. Don't come here. Don't complain that someone is biased and that the entire site is now stupid because of that. You seem to put down the entire site because of a couple of bad articles. That doesn't make any sense. I have come here before and seen things I don't particularly agree with but I don't just close the browser and never come back. Every person, no matter who they are, has some sort of bias. Deal with it.

And on the topic of benchmarks and hardware and such...[H]ard|OCP is dedicated to showing us the best of the best. As soon as a new product hits the streets, we know about it. And they benchmark it on a system that is best suited to doing that sort of thing. In regards to video cards, they use a highend system to remove any chance of the CPU factoring into the benchmark. They want to obtain the purest results they possibly can.

This whole thread destroys the first poster's argument anyways...he is biased against the mighty [H].

Case closed.

And go away if you don't like it. Kyle has more important things to do than deal with whiners.

 

[GabooN]

I fully trust Hardocp more then any site out there. Not only do they review a product but THEY always list several others who have reviewed the same product. I know when I read a review from Hardocp and buy that hardware I can duplicate there scores because I have in the past. If I want to tweak my system I will use sandra soft and several games that I play NOT some synthetic benchmark that I cant play. Keep up the good reviews [H]!!!

 

[burningrave101]

Percentage wise, yes, but with the greater IPC on the AMD procs they actually gain more per clock than the P4 does. The 2.6 3000+ will blow a 3.4 Intel out of the water, just as the FX-55 does

And no, no pif

The performance of the Athlon 64's is attributed to their on-die memory controller and the fact it scales in speed with the processor. The latency of the memory controller in the northbridge really hinders performance. The Pentium 4 had alot less issue with this then the Athlon XP did but you can't beat having it integrated into the CPU itself.

The architecture of the K8 core is not alot different from the K7 and an Athlon XP gets blown out of the water by a high-end Pentium 4. An Athlon XP @ 2.4Ghz can barely keep pace with a stock Pentium 3.0c in the areas its strong in and has no hope in the areas the Pentium 4 is strong in like encodeing and multitasking.

The K8 core has an extra 2 stage pipeline making it 12 stages now, it has a few added registers, it has the x86-64 extensions, and it has the on-die memory controller. Those are the major differences between the two.

If you took the integrated controller away from the Athlon 64 then i'm afraid AMD would be back to trailing Intel again. I'm wondering when Intel will or might start working on a move of the controller from the northbridge to the CPU themselves. This would make the Pentium 4's MUCH faster in games considering games and multimedia already favor deeper pipelines.

 

[SuperSparky]

Let's see, "The New York Times" has got to be the most left-wing biased newspaper in the country and it's obvious. "The Wall Street Journal" is one of the most right-wing based newspapers in the country and it's obvious. Hmm, TV news.... CBS, definitely left-winged ... far left-winged bias. In fact it's obvious. I could go on. Boy, naivity, it's scarey.

I would think the great differences between DOOM3 and Half-Life 2, both DX9 games would prove a lot is involved in the engine's programming and not just the graphical API are what makes a game behave. Using the "it uses all of the DX9 API" logic fails when looking at the differences in games that use the DX9 API as wel,l and are vastly different in speeds on different manufacturer's video hardware.

I think what a synthetic benchmark fails to consider is how differently each engine uses those API paths to achieve similar results. Part of game coding is optimization

If I'm to purchase a card or even a game, I'd like to know how they behave together. If a card, then I'd like to know how it runs on the game(s) I'm to play on it. If a game, then I'd like to know how that card plays that game. Synthetic benchmarks are no more useful than a racecar with my own car's branding on it. It's apples and oranges. Seeing how the racecar performs isn't going to tell me how my car will perform.

Anyone who has worked for NASA can tell you that simulations never can simulate real life, as real life always finds a way to throw in something you never considered. They only use simulations for safety, not accuracy. Nevertheless, they will tell you if there was a 100% safe way to practice in "real life" they would.

Video games and other graphical applications are "real life." Each have their programming methods of generating a scene and thus will behave just as differently as the synthetic benchmark's "engine." Each scene has its own amount of polygons, and surfaces. Each will put different loads on a GPU for a given point in time. None of the scenes in a synthetic benchmark are in a real work game and visa versa. It's comparing apples and oranges.

Now if you had a benchmark fetish and all you did was watch them, then I suppose they are real world to you. People happen to play games, not benchmarks.

 

[oldsk00l]

If you took the integrated controller away from the Athlon 64 then i'm afraid AMD would be back to trailing Intel again. I'm wondering when Intel will or might start working on a move of the controller from the northbridge to the CPU themselves. This would make the Pentium 4's MUCH faster in games considering games and multimedia already favor deeper pipelines.

You're close, the K8 arch does have a few more pipelines, and different logic gates. They streamlined alot of operation into a more logical array, they have also made some arrangement changes for heat issues.

About games favoring deep pipelines, I have a bone to pick there. It's about your style of programming, and the compiler you use. I've seen 3d rasterizers in Linux go from the gcc compiler to an Intel compiler and gain %30 in speed. Also as you utilize more 8 and 16-bit data structures, the matrix you decide to use if you use Octrees instead of BSP.

Pipeline depth, imho, is a not nearly the factor the onchip mem controller is. Also, Intel wants to move back to being conservative with their pipe depths, hence the re-arrangement of their roadmap.

The G5 and the K8 prove that simply throwing more stages for a clock increase is not the most efficient way of accomplishing this. Intel's current situation wrt CPU performance beind behind, is the reason [H] might seem biased. Intel half-admits it themselves that they should have gone a different route though.

 

[_Alx]

I think it's funny that there are eleven pages full of well stated and well supported counter arguments to the original post, yet maleficarus lurks in the dark and only emerges to dismantle the few weak/irrelevant ones; then claims that this whole thread is nothing more than a bunch of testimonials. Another day, another forum battle won!

On another note, thanks Kyle, Brent, Sean - keep up the good work. And I believe a few weeks ago you had another 3dmark article in the works around the release of '05, Kyle mentioned it as a side note in a news post about a lawsuit... whatever happened to that? I was really looking forward to it

 

[Bubba The Barbarian]

If you took the integrated controller away from the Athlon 64 then i'm afraid AMD would be back to trailing Intel again. I'm wondering when Intel will or might start working on a move of the controller from the northbridge to the CPU themselves. This would make the Pentium 4's MUCH faster in games considering games and multimedia already favor deeper pipelines.

If you took the defence away from the Pats, they would be a realy sucky football team too.

Look, both procs have their pros and cons. From a clock to clock standpoint, AMD does more work and always has, with lower clock speeds being a byproduct of 3 primary pipes as opposed to 2. AMD 64 chips DO have the mem controller, making tightly timed DDR act like a big ass L3 cache. Taking it away is like taking away hyperthreading from the P4, it is part of the design!

AMD does have the best price/performace ratio right now, and is better gaming chip due to memory performace. Deep pipelines mean that cache misses are punished harshly, which why you see more shallow pipes on recent chips rather then fewer deeper ones (Why did the FX suck so hard? Why did the 9700 rock?).

 

[El Perro]

Hmm, I've been coming to the [H] ever since the 300A and well... wow... These are a group of guys that have strived to provide the most objective reviews on hardware on the web and people bash them for being biased? These are a group of guys that when a specific widely used benchmark is uncovered to have been skewing up results in favor of one card they seek out to correct it. A group of guys that when intel releases a messed up processor they test and test and test to make sure that they are right and they ally with a website they actually hate just to be able to make sure that the end user (us) receives the highest quality product they can. They can admit when they make a mistake but they also stand by their opinions no matter what and they have done this both in the nVidia debacle and the intle debacle. They benchmark things in the most objective way they can and then they provide an opinion on the hardware. Well, that's what I come here for. I see the numbers and read their opinions. most of the time I agree and if I don't it's because I was able to form an opinion on the numbers.

And the bashing is coming from a guy that denies the facts and is probably more biased than the [H]... go figure...

oh well, been coming here since the beginning and I will be coming here for a long time to come. keep up the good work.

 

[maleficarus]

Just so everyone knows...

I'm maleficarus™, doer of all things evil...

 

[ryankenn]

I only have a couple of questions after skimming over these posts, so I may have missed it.

Did he say where he goes to read hardware reviews he trusts?

Did he make a more specific case for why he feels they are AMD biased, or unfair in their videocard reviews.


I for one find this site extremely useful, and I've switched platforms many times, each time I did a quick search to find out things about the MB/Proc/Video Card I was looking at. Everytime the reviews seem spot on. I see the review Intel as well as AMD, and the reviews can be good for the Intel camp, they even had the one up to 4.44Ghz which was amazing, I'm just not sure I'm looking at the part of the reviews he finds biased?

Just now I read the new 850XT Preview, and although I don't have PCI-E, I found it useful for me, because based on what I've read up until now, I was certain I was going to get a 6600GT or Vanilla 6800. I play HL2 like everyone else, but I also play a variety of other games, and seeing the benchmarks for games like Madden and NFSU2 makes me wonder about my decision. I'm not getting into a debate about video cards, but I find the range of games benchmarked give me a good idea what I can expect. Most other sites test a bunch of FPS which doesn't really help me out.

Keep up the good work [H].

 

[Bubba The Barbarian]

Just so everyone knows...

I'm maleficarus™, doer of all things evil...

Accourding to Jim Romes Rules of Smak talking , one cannot gloss himself. The USSR did not gloss itself as the Evel Empire, Bill Gates did Gloss him as Borg Boy, and Ballmer's attempt to gloss himself as the biggest supporter of developers ever got him named "Monkey Boy".

In that vein, let's brake down your name.
maleficarus -
Mal = bad
efi = girl's name
carus = carion.

NIIIIICE....

 

[burningrave101]

If you took the defence away from the Pats, they would be a realy sucky football team too.

Look, both procs have their pros and cons. From a clock to clock standpoint, AMD does more work and always has, with lower clock speeds being a byproduct of 3 primary pipes as opposed to 2. AMD 64 chips DO have the mem controller, making tightly timed DDR act like a big ass L3 cache. Taking it away is like taking away hyperthreading from the P4, it is part of the design!

AMD does have the best price/performace ratio right now, and is better gaming chip due to memory performace. Deep pipelines mean that cache misses are punished harshly, which why you see more shallow pipes on recent chips rather then fewer deeper ones (Why did the FX suck so hard? Why did the 9700 rock?).

The clock for clock performance doesn't explain the better gaming performance on the Athlon 64. The Athlon XP did more work clock for clock then the Pentium 4 and yet the Pentium 4 "c" series was faster in games which was one of the Athlon XP's strongest areas. The Athlon 64 is also doing less work per clock cycle now then they did with the Athlon XP or at least according to AMD's rating scheme it is. The fact they pumped the stages up to 12 does have something to do with it though. They had to increase the stages though because otherwise they wouldn't be able to increase the clock speeds further.

The integrated memory controller decreases the latency time associated with communication between the CPU and the RAM. This is one of the biggest bottlenecks in performance so this helped performance alot where access times are important. Latency is everything in gaming performance.

 

[Sizer]

Tom's Hardware is soooooooo shitty, after all, their admin banned an assload of posters from their BB because they pointed out a flaw in.... oh wait

 

[covertclocker]

I think everyone who thinks [H]ardOCP is biased needs to lay off the crack. This site has always been objective. Is it so wrong to support the winner? If anything, the stronger argument would be that [H[ardOCP flip flops between who is better. But here again it is only because products change. Currently Intel Processors are dogs. They don't do as much work as AMD's do, so [H] is in AMD's corner it seems from time to time, but if Intel breaks out with a better product, i am sure [H] will go to their corner. I, like many people am looking for the best bang for the buck, and [H] gives very intelligent, and good reviews. Kyle and the crew do their reviews on the best bang for the buck, hence the apples and oranges tests to see which card can give you the best resolution and IQ. I don't have money to be pissing away on return fees etc etc. Kyle, Steve, and the rest of the crew here at the [H] please keep up the good work!

For the haters, FUCK OFF!

 

[el rolio]

I'm not sure if you understand the true nature of benchmarking. Let's take your example. You consider that Elvis is the King and performs better than anyone else. However, a recent British poll found that John Lennon was the greatest rock icon. Who is to say who is right? How do you determine who is better? On the surface it seems subjective, but it's possible to benchmark the two.

If you use the 3DMark method, you would sit each singer down and have them sing the same four songs. The end result would give you some number saying which singer is truly better. Now that isn't quite fair is it? While they are quite good at what they do, it doesn't tell you much about those singers. This is especially true if you have songs like Twinkle, Twinkle Little Star; Viva Las Vegas, Imagine, and Mozart's Requiem. Each singer would rock the first song, do great on their respective songs and bite the big one on Mozart. Again, this isn't a good test.

Using the HardOCP method, you test each singer singing their own albums. Here you see a "real world" examples of how they perform. You are able to see a greater range of the performer, see how they are "optimized" and also see where improvement is needed. I don't have time to listen to every album to see how they might do on the next Nelly album that's coming out.

That's where metric benchmarking comes in. Take each singer and break them down to see how they would perform. You start measuring every nuance of Elvis and Lennon. How many albums were sold, length of career, ec. Elvis is a better singer, but Lennon is the better writer. Elvis had more number one hits, but Lennon was more influential. Elvis has a better and larger vocal range and dances better, Lennon is better at instrumentation and branched out away from music. They both have downsides like doing drugs and Yoko Ono. It's hard to cheat as well. Either the singer gives me a C# when I ask or he fails.

Once you have that out of the way, you can start comparing. Lennon would do well at writing the rap lyrics, but would probably suck at performing it with his Liverpool accent. Elvis would probably do better overall. We don't have the Nelly album, but because we have good numbers to compare the singers we can extrapolate who would do better.

Hopefully, this gives you an idea of why metric benchmarking it the best way to go. Comparing physical hardware is even easier with metrics than it is with singers. I can measure specific areas and know how the card performs exactly.

APPLAUDED. as one who loves and uses analogies frequently... i give this 2 thumbs up! ha

 

[NEVERLIFT]

The performance of the Athlon 64's is attributed to their on-die memory controller and the fact it scales in speed with the processor. The latency of the memory controller in the northbridge really hinders performance. The Pentium 4 had alot less issue with this then the Athlon XP did but you can't beat having it integrated into the CPU itself.

Up till the Northwoods core the P4 was slow, so can we take away its HT that helps its bloated pipeline, weak FPU, 33% less IPC and barely any lvl1 cache.

The architecture of the K8 core is not alot different from the K7 and an Athlon XP gets blown out of the water by a high-end Pentium 4. An Athlon XP @ 2.4Ghz can barely keep pace with a stock Pentium 3.0c in the areas its strong in and has no hope in the areas the Pentium 4 is strong in like encodeing and multitasking.

I have saw AthlonXP 1800+ 1.53ghz beat 3ghz P4C in FPU benchmarks and in all fairness it usually takes a 2.6ghz or there about P4 to really beat a 1.5ghz AthlonXP.
Guess that quad pumping just causes to much friction. (non pun intended)

The K8 core has an extra 2 stage pipeline making it 12 stages now, it has a few added registers, it has the x86-64 extensions, and it has the on-die memory controller. Those are the major differences between the two.

Ya your pointing out the fact AMD did this first and got it right. Thats why intel is copying their 64bit instructions and intel made the Pentium M. Looks like intel wont be using the P4's either for their dual core and instead will be using the Pentium M. Also why did intel move to the PR rating scheme.... guess they was embarrassed to see there CPU's getting beat by AMD CPU's running over 1ghz slower.

If you took the integrated controller away from the Athlon 64 then i'm afraid AMD would be back to trailing Intel again. I'm wondering when Intel will or might start working on a move of the controller from the northbridge to the CPU themselves. This would make the Pentium 4's MUCH faster in games considering games and multimedia already favor deeper pipelines.

And on the same front if you took away intel's HT (which is a stop gap for the P4's bloated pipe-line) and their special instructions like SSE1/2/3 which helped a great deal to make up for the P4's weak IPC and weak FPU and very low lvl1 cache. (I love the SSE1/2/3 cause when AMD uses them it makes them even better)



Back on topic.

Kyle and Steve and the guys working for them are a great bunch of people that call it as they see it, they report the facts in your face and if you dont like it then no one is forcing you to come to this site. But anyone with common sense and that visits other good sites like Anandtech and Techreport for example will see that its not bias, its just the facts being reported. And I'm very proud of the stand HardOCP has took with the videocard industry! This shows they are standing up for us.


Thank you HardOCP and keep up the great work.

 

[rogue_jedi]

Basically my view is that [Hard]ocp tries to push its views on us readers without being fair and unbiased. That would not go over if this was a newspaper as the readers would not tolerate that kind of journalism in a publication. So why does it get over looked on the www?

QF(not)T

have you EVER read the Washington Post? pretty liberal right there... ALL newspapers/news agencies have an agenda. they ALL push it as much as they can... no reason the internet is any different.

personally, when i read reviews on hardocp and i don't have much time, i just skip straight to the conclusion. why? because for almost every review i have read on hardocp i have taken benchmarks into consideration, as well as problems they reported, and agreed with their conclusions. if they make a conclusion that seems ridiculous i'll try to figure out why they said it.

i like the idea of using unrealeased timedemos for benchmarking, because it prevents much of the driver "optimization" that is possible.

no offense and don't take this the wrong way or anything, but nobody's forcing you to read hardocp's reviews. as you said, there are other review sites that you like, so you can just read reviews there.

really, though... the only reason (IMO) that you find "bias" in a review is because you disagree with the opinion, or are completely impartial. you don't find said bias in other reviews because they happen to agree with you. i'm not saying i don't do this too, i'm just mentioning it.

just thinking out loud a bit, that's all.

 

[Glimfeather]

1. Once upon a time Tom's Hardware was a legit source of information on building myself a computer for at least $1500 less than retail so my wife would'nt get all mad and stop cooking for me and stuff. Then Tom went down the road to stupidville.
Now, today, in my opinion HardOCP is perhaps the most legitimate source of information on not just system building, but on many other areas as well. As for bias, we all are a bit biased; much like each poster's thumbprint is unique, so will each opinion differ. What matters is that the people at HardOCP strive to be honest and straightforward about their advice, and that is all we in the community can ask for. Incidentally, I also appreciate alll the knowledgeable people who respond with suggestions, assistance, and advice to question/problem threads. You guys are my help desk.

2. To add to the CBS/bias/journalism portion of this thread, IMO the most incredibly funny thing I have ever read about Dan Rather: "What's the frequency, Kenneth?" http://www.ratherbiased.com/bizarre.htm#kenneth

3. Recently, just before upgrading to my excellent A64 / Abit AV8 / GeF 6800 "platform of solidness", I wrote a quick email to Kyle to say thanks for no-BS advice on quality affordable components, AND HE ACTUALLY TOOK THE TIME TO WRITE BACK. In my mind, this fact alone sets HardOCP apart from the rest.

4. At first I was in angry disagreement with the piece about "CS:S is beta at best..." but then after thinking about it for a few weeks I realize how true the [H] assertion is. Even Cliffe's .5xx series betas of the original One True Counterstrike had certain features, details, and fidelity that surpass the current CS:S offering. I'm still a fan of Source, but I look forward to major improvements in my "anti-drug". Also, the [H] gaming CS:S server is [H]ard.

5. Infinium Labs Expose'. Now that is a great testiment to embarrassing stupid people. May they never sell a single console.

6. It seems that instead of making some of us open-minded, the net has somehow made some of us more closeminded. Example: if you don't agree with something you read on the intarweb, for the sake of all that is good in the world don't take the thing personally! Keep perspective, people. That's all I ask...try to keep things, EVERYTHING YOU READ ON THE UNREGULATED INTERNET, in mature perspective. You will be the better for it.

 

[Skolar]

Stupid ignorant people!

Nobody forced their views on you becuase it was your own choice to read the reviews in the first place. By definition voluntary means: "Done or undertaken of one's own free will." Thus it is impossible for [H] to force their views on you because you voluntarily read them. Idiot

Somebody please ban maleficarus's IP...Thank you.

Skolar

 

[MeanieMan]

12 pages of posting, the following is what I find:

Good
Comments on stuff that could be changed
Hints on why Hard may appear bias, not bias.

Bad
Comments about whiners
Whiners whinning about whiners
Pentium 4 vs. AMD (wtf?)
Off topic deleted posts
Down right insults

Why is so hard for you people to stay on topic? Since when did an opinion warranty a flame in return? What does one person's opinion have to do with the amount of level 1 cache a Pentium4 has? What is the use of posting a response that calls everyone else an idiot?
Yes, I know this off topic, but after 11 total pages of insults and BS, it needed said.

 

[EQTakeOffense]

Basically this thread has been broken down into the following:

Is it fair to say AMD is better then Intel when Intel has it's own strong points?

Is it fair to bash 3Dmark because nVidia and ATi cheated via drivers to get higher scores?

First thing first, fair is all a matter of opinion. Secondly, there is no anwser.

I think [H] has many time stated that although AMD and Intel both have their strong points, AMD bests Intel once the final pros and cons are drawn. The debate over which one is 'better' right now is actually fairly close, depending on which features you view as important and what exactly you're looking at. Since this site is mostly viewed by gammers and power users, [H] needs to say which one is better to use, which in this case is the Athlon64.

I agree that nVidia and ATi are the ones to blame, but I don't think 3Dmark should be left out either. All are guitly one way or another. If you're in a car, and your buddy has drugs that you don't know about, you still both go to jail. Same kind of concept.

 

[SquireSCA]

You gave me front page!!!! You just made all my years of internet forum flaming worth it. Thank you kyle. I now have a new found respect for you

Dude, making first page like that, is like being put up on Cliff Yoblanksi's site.

He is showing everyone what an idiot you were. Now people can get a taste of the kind of antics that hove gotten you banned from 10+ forums since 1998 when many ofus met you at Delphi Forums...

 

[SquireSCA]

You're right. This thread is basically a testimorial. Just post your thoughts on what you think about synthetics, this site and the games we play. Let's be honest here. This site is one of the most popular when it comes to the games we play. I mean its link city galor. If I want to find out about a game the first place I look is here. He has all the latest news cause he links it up. Another thing like is [H] posts links to other topics. Some are kinda dumb but some actualy make me sit down in front of my PC and read.

But.. and there is always a but. The dam site pushes its views on certian topics that really piss me off as a gamer. Like to try to take down synthetic benchmarks was like saying elvis was a bad singer you know? He was the king!!

Dude, why do synthetics matter? You have a low end system that does equally poor at benchmarks and game scores, so why would either one matter to you, in all seriousness?

 

[chrisf6969]

I think Kyle is egotistical pigheaded self-centered prick, but then again so am I. If you know everything why not be conceited about it. j/k

I think the [H] way of reviewing videocards kinda blows, damn apples to oranges comparisons confuse me!

And when did 16xAF become the standard? I thought 8xAF was generally the standard!

That being said... keep up the good work, do what you think is right, don't be lemmings and do what every other review site in the world does.

 

[jacuzz1]

Hmm, I've been coming to the [H] ever since the 300A and well... wow... These are a group of guys that have strived to provide the most objective reviews on hardware on the web and people bash them for being biased? These are a group of guys that when a specific widely used benchmark is uncovered to have been skewing up results in favor of one card they seek out to correct it. A group of guys that when intel releases a messed up processor they test and test and test to make sure that they are right and they ally with a website they actually hate just to be able to make sure that the end user (us) receives the highest quality product they can. They can admit when they make a mistake but they also stand by their opinions no matter what and they have done this both in the nVidia debacle and the intle debacle. They benchmark things in the most objective way they can and then they provide an opinion on the hardware. Well, that's what I come here for. I see the numbers and read their opinions. most of the time I agree and if I don't it's because I was able to form an opinion on the numbers.

And the bashing is coming from a guy that denies the facts and is probably more biased than the [H]... go figure...

oh well, been coming here since the beginning and I will be coming here for a long time to come. keep up the good work.

werd

 

[mentok1982]

I believe this photo sums up my feelings for [H]ardocp.com expressed through the medium of Counter Strike Source.

Opinion

 

[GeForceX]

Nottt sure what kind of opinion that is. You priding in that logo or you actually cutting up that logo 'cuz [H] sucks?

My opinion, despite all the flames...

[H] is my start up home page.

Should I even say more?

<3...

-J.

 

[GilmourD]

Most of the people who have issues with the [H] are n00bies, IMO. Well n00bies, here you go.

3DMark is good for nothing but stability testing, and to see if my video card has a problem or not (downclocking due to overheating, artifacting, etc. etc.) due to my overclocks. That's all it is.

The [H]'s methods of testing has saved me in over $1000.00 in hardware purchases, whether it be for family, friends, or myself. Before that, I was dancing to the beat of Tomshardware (bias, anyone? ), which promptly helped me regret about half of my purchases. So I see absolutely no issues with Kyle's methods of reviews. One thing I learned with the [H] was this:

If the hardware was good, the [H] is "biased" toward it. Perfect example: AMD was whipping Intel's ass all over the place with the Athlon XP. Intel came out with the Northwood C, and the same happened for AMD. First, it seemed the [H] was biased AMD, then you could have sworn they were biased Intel. Same goes for ATI/nVidia. Do you see the [H] bashing nVidia like they used to, now that nVidia has released the 6800 series GPUs? We all know the FX series was utter trash. And even I bought an FX, so I'd know.

Bottom line: Kyle and the [H]'s "bias" is nothing but honesty in trying to get us (the enthusiast community) the best bang for the buck. If you can't handle it, then you might as well leave.

I couldn't have said it better, except to add one thing. It's not the n00bst1cks that bash the [H], it's those people who simply can't accept blunt honesty. Everybody wants candycoated reviews. Everybody wants to read all the good reviews for stuff and scoff at bad reviews. Everybody wishes that there could be pretty little flowers and smilies everwhere. Well, maybe not that extreme, but sometimes knowing what sucks is just as important as what's good. Kyle, Steve, and the rest of the [H] crew have the balls to be honest with me, you, and the world. If you don't have the balls to handle the truth, that's your own problem.

I've been a PC tech for roughly 10 years. I've seen a lot of stuff, used a lot of stuff, seen a bunch of things suck, and stuff that should work simply blow up. [H]ard|OCP reviews simply reflect my personal findings of what's good and bad, and saves ME PERSONALLY a lot of money when I'm making purchases on things that I don't have experience with, yet. Every time I've said "Eh, I'll try it even though the [H] said it bit the donkey on the ass" I've been bitten in the ass.

If you want candycoated, paid for reviews, go check out Tom's Hardware. I've seen him give praise to stuff that I personally could not get to work remotely like he had them working in identical configurations. That's why I stopped reading Tom's Hardware cold-turkey a few years ago, cuz I realized that he's biased and his opinions fluctuate depending on how many ads for that manufacturer are on his page.

 

[GeForceX]

I wish Tom could come by and actually say something. Defend himself? I love seeing defense talks.

-J.

 

[Brent_Justice]

And when did 16xAF become the standard? I thought 8xAF was generally the standard!

In finding the highest playable settings that includes AF as well.

I do only use 8XAF for the Apples to Apples section.

 

[NEVERLIFT]

I think Kyle is egotistical pigheaded self-centered prick, but then again so am I. If you know everything why not be conceited about it. j/k


I think the [H] way of reviewing videocards kinda blows, damn apples to oranges comparisons confuse me!

And when did 16xAF become the standard? I thought 8xAF was generally the standard!

That being said... keep up the good work, do what you think is right, don't be lemmings and do what every other review site in the world does.

You almost hit the nail the on the head.

There are just to many settings to set and test with. Yes its all nice to do, but when its all said and done... and the $500 videocard is laid out and slapped around you must realize that if the the card does good at 16x af and 4x aa etc then surely it will do even better with lesser settings.
I saw alot people post about the higher end cpu's used in these tests. This is common sense to me, if a person has the money to get the videocard then they surely have the money to get the best mobo's and cpu's for them. Like Kyle said they are on a tight time frame and try to get out the best game experience review possible.
I like the apples vs oranges reviews but look at the numbers and not the charts with graph. That chart gives me seizures almost. j/k It reminds me of a the monitoring done in hospitals for someones heart beat and vital signs.

The HardOCP put out real gaming experience reviews for the said games used and even thou their reviews are small and to the point. Anyone should come away from them with an general ideal of which card is better in said game.

There are sites like Toms that has every videocard since the TNT2 and they benchmark showing all these cards and sometimes even have multiple cpu's, The Techreport is good about this also. I like seeing the same videocard used in both a intel and AMD system and compared apples to apples. It shows the weakness of and strong points of both systems.
I really like when they do multitasking and benchmark the systems. It shows how the pc responds and which is faster. Seems that even thou the P4C/P4E are smoother due HT and high bandwidth they are indeed slower. And will lose.

The [H] said they see enough review sites doing these kinds of reviews and so dont feel the need to do them also. This why they do their own kind of reviews.
A man really interested in a certain CPU or Videocard should visit several sites like Anandtech, The Techreport and PCStat, Toms(lord forgiveme) and more sites! And take in all the data and draw their own opinion as to what is good for them and their needs. And yes these sites will be bias to the best and tell you what they think is the best. If you disagree and feel otherwise then thats your call. No one makes you read these site.
And if you visit 5 sites for example and then one other site is reporting something different *cough* Toms... then if you cant see who has SOL out thats your own fault.

As a last word.... If sites like HardOCP and others had not stood up and been so vocal about 3DMark, then both ATi and nVidia would both still be spending money and man power to make drivers that run good in these lame ass worthless benchmarks. I have saw to many driver releases in the past that were released for 3DMark and yet broke gameplay in real games! This is not productive and thanks to HardOCP and other sites it has pretty much stopped.


PS: Opinions are like assholes and everyone has them. People are intitled to both so leave it at that and stop the bitching and crying and if you dont like the [H] then stop spending so much time here, cause you will change nothing.


Quote from the book of [H]

[H]ard is the process of enlighten and knowledge of hardware. It’s about taking your hardware to the extreme, just to the point of failure, were it still works, and doesn’t fail. But performs the task needed by it faster or quieter than what it was designed for. It’s sharing that information with your peers, information that will make each others system performs better and stable. It’s a family, a set of values, a set of goals, respect.

[H] Is about living life at another pace, it’s about going farther than the rest, achieving a little bit more. Is a community of people of different backgrounds, ages, raze, sex and believes, that share the same set of rules and commitment, that have to do with everything in life.

 

[Brent_Justice]

It's these "nvidia uses cheating drivers" and "3dmark05 is crap" type of claims that seem to drag on and get overplayed (like an OJ Simpson trial). Most of it is drama.

They aren't called editorials for nothin

 

[KingPariah777]

I don't mind the [H] way of doing reviews as long as they include tradtional framerate benchmarks as well.

 

[GeForceX]

^^^ Agree.

Besides, it's usually my one stop read for the afternoon. Best to have all the apple/orange/peach/banana/grapes all in one article.

-J.

 

[NiViK]

Sheesh...

I dont post too often, but I couldn't pass this one up. Here is a little story.

Way back in the mid 90's (1996 or 97 I believe) I found a great hardware site called Tom's Hardware. Back then Tom himself ran the site...it was a small site and Tom did all the work himself. He was pretty honest and tried to get vendors to give or loan him hardware to test out so he could post his reviews.(Back then it was unheard of for vendors to give hardware to smalltime hardware websites...which was basically all of the review websites)

Well...after a few years (2-4), something happened. Tom sold out to the man. I dont know what happened for sure...maybe he was paid a lot of money to change things...maybe he was tired of all the work...I am not sure if anyone really knows.

I for one noticed it almost over night. His website became overrun with ads, and the reviews changed also. Here is an example... At one time Tom was an avid supporter of AMD. Then all of a sudden, AMD was the Devil. AMD could do nothing right.
This not about what preforms better. At the time, Intel was faster...but AMD was a hell of a lot cheaper. It was a choice...loose a little speed and save a few $$$. For most people that was a nobrainer. Now the AMD K5 was a dog, I'll admit...but the K6 was much better thanks to AMD aquiring NexGen and their Nx686 processor...(go RISC) for the price at the time it was a decient alternative to a PI/PII or a Cyrix(I have always hated Cyrix).

Also...a big factor was that suddenly..Tom wasn't doing the reviews himself anymore. Suddenly there were several new people doing all the reviews and articles that. Did I trust them like I trusted Tom??? No way.

This was a low point for me. I had no reliable source for hardware information anymore.

This was about the time I found [H], Arstechnica and Anandtech.

Over the years things change...Arstechnica used to have lots of great hardware reviews...now they are more news orientated...and that works for them. I like [H] because they seem to cover most of the bases. They are not obsessed with just the highest end newest equipment. They also post guides on how to get the most for your buck.

What really makes me laugh is when some of you forum whores start spouting off that [H] biased. I am sorry...you obviously do not know what biased reviewing is. What I see here is people that are so stuck on a given technology they can not stand it when anyone says that a competing technology is as good/better than what they like.

Example time again...

I used to hate ATI. Back in the days of the Rage Pro/98/2000 days I would rather claw my eyes out than use them. I was 3DFx all the way. Well now I have a ATI card in my gaming box at home...I still use a Nvidia card in my video editing machine...but the ATI card blows away the Nvidia one for the price I paid. I am sure there will be a time that I will stick a Nvidia card in my gaming rig again. But for now, ATI has won my vote.

One more thing that I like about [H]...
They link to other reviewer sites. They are not scared to listen to someone elses opinion. They are also not scared to call someone out if they are not reporting the truth. Also they are not above apologizing when they make a mistake.

What really kills me is that maleficarus states he is here only for the forums...but he has only been on the forums for close to 2 weeks(11/17/2004)? Is that honestly enough time to make a valid opinion of a websites reviewing pratices???

Come back in 6 months and post your opinion when you have more information under your belt.

 

[erexx]

Its not about 2 things opposing and one being the winner.
The US vs Them mentality is crippling

Why not collect information from more than one source and then make your own decisions?

If everyone wrote without any bias or editorializing there would be no challanging popular opinion.
No new ideas...
No exchange of new information...

Only complaint I have about [H] and I have been reading since 98.
Is the use of "crazy" bar graphs in their reviews over the last year or so...
Can you guys visually represent the data in a simpler easier to absorb way?
Everytime I read one I think there has been an earthquake somewhere.

Other than that HardOCP is a top notch source of information.

 

[rcolbert]

When do anti-lemmings simply become lemmings headed in the other direction?

How many high school students run out and get tattoos to express their individuality just like everyone else?

When is Kyle going to admit that he truly thinks the world would be a better place if we all had Billet grills on our cars and trucks?

Why do I even bother buying CPU magazine just to read incomplete and forward-looking articles about things that were published on the web seven weeks ago?

Do you look after you wipe?

Sadly, these are my deepest thoughts.

 

[blackadde]

Well...after a few years (2-4), something happened. Tom sold out to the man. I dont know what happened for sure...maybe he was paid a lot of money to change things...maybe he was tired of all the work...I am not sure if anyone really knows.

All I remember is an 'Approved by NVIDIA' jpeg on a review, and it was all downhill from there.

(I've been reading the [H] since they actually knew what the OCP meant, I just never bothered registering here until very recently )

 

[Torgo]

The only thing you truly have to fear in the [H]ardForums is Kyle giving you a silly forum title. So far I've escaped his notice.

To tell the truth, threads like these everyone once in a while are good for introspection. While it may not be true, you have to ask yourself, why is everyone saying this about me? Is there something here that could give the appearance of bias? You might find some ways of improving the site even more.

 

[burningrave101]

Up till the Northwoods core the P4 was slow, so can we take away its HT that helps its bloated pipeline, weak FPU, 33% less IPC and barely any lvl1 cache.

Thats because the new 20 stage pipelines were not operated at a high enough frequency to keep them full. When the stages aren't kept busy they are pretty much idleing and this causes the performance to be less then what it is capable of. If the stages aren't full then it takes alot longer for them to empty when there is a branch misprediction.

I have saw AthlonXP 1800+ 1.53ghz beat 3ghz P4C in FPU benchmarks and in all fairness it usually takes a 2.6ghz or there about P4 to really beat a 1.5ghz AthlonXP.
Guess that quad pumping just causes to much friction. (non pun intended)

Provide links or dont bother even mentioning what you've saw because there is no way an 1800+ can beat a 3.0c in anything worth mentioning.

Here is a long thorough review of an Athlon XP 2500+ overclocked to 2.4Ghz.

http://www.techreport.com/reviews/2004q1/athlonxp-m-2500/index.x?pg=1

It can barely keep up with a stock 3.0c in the areas the K7 core is the strongest in.

Ya your pointing out the fact AMD did this first and got it right. Thats why intel is copying their 64bit instructions and intel made the Pentium M. Looks like intel wont be using the P4's either for their dual core and instead will be using the Pentium M. Also why did intel move to the PR rating scheme.... guess they was embarrassed to see there CPU's getting beat by AMD CPU's running over 1ghz slower.

Intel could of done their own 64-bit instructions at any point in time they chose. EM64T was based off of x86-64 for compatibility reasons. If the design of it was changed it would require two different operating systems for AMD and Intel processors. Who do you think designed the ENTIRE x86 architecture that AMD has used on all of their processors? Where do you think AMD got MMX, SSE, SSE2, and now SSE3 which will be included in upcoming processors?

The reason Intel is using the Pentium M architecture is because its damn fast and because you dont need high clock frequencies for a dual core processor. Intel is reaching the limits of the speed a processor can handle on their current fabrication processes.

And how would the PR rating scheme change anything at all NEVERLIFT? A rating scheme doesn't make a processor faster or slower.

And on the same front if you took away intel's HT (which is a stop gap for the P4's bloated pipe-line) and their special instructions like SSE1/2/3 which helped a great deal to make up for the P4's weak IPC and weak FPU and very low lvl1 cache. (I love the SSE1/2/3 cause when AMD uses them it makes them even better)

If you disabled HyperThreading on the Pentium 4c your gaming performance would actually go up a few percent. HyperThreading has no affect on anything except multithreaded applications and multitasking.

AMD uses those special instructions like MMX, SSE1, SSE2, and soon to be SSE3 as well .

A strong floating-point unit doesn't mean as much as it used to. Todays applications are optimized for instructions like SSE2 which will allow the Pentium 4 to bypass its floating-point unit. A strong FPU is useful for older applications and scientific applications but it has little affect on most of todays applications and especially games. Most all software is optimized and built on Intel hardware because Intel owns over 80% of the desktop market share. They aren't going to optimize for something the Pentium 4 isn't strong in.

The Pentium 4's L1 cache is smaller because the smaller the cache the better the latency. The larger the cache the longer it takes to search through it. You want the L1 cache to be small because it only stores important data that needs to be accessed very quickly. The Pentium 4's L2 cache is also much faster then the Athlon XP's.

Deeper pipelines will always be faster then fewer pipelines if the branch prediction unit is good enough and the clock frequency of the processor is fast enough to keep the stages full simply because you can move more data. Its just like a 4 lane bridge and an 8 lane bridge. You can make the cars drive faster on the 4 lane bridge (AMD) but your going to still be able to move cars more quickly over the 8 lane bridge (Intel).