Anandtech does it better

A

asmielia

Limp Gawd
Joined
Aug 10, 2003
Messages
337
Hopefully this thread will get read by editors of [H]ardOCP. I love [H]ardOCP but I really must say that today's Anandtech nForce4 SLI motherboard roundup is head and shoulders above the motherboard reviews done by [H]ard recently. It has been a trend where Anandtech has been doing a better and better job and I hate to say it, but [H] has been getting worse. Maybe it's because Morry has been doing the motherboard reviews lately.

The problem is that [H] isn't reaching their target audience as much as they used to. The motherboard reviews aren't digging deep enough into the features and ESPECIALLY, the overclocking. Overclocking needs to get a major overhaul in thoroughness. Look at the overclocking section on the HardOCP review of the DFI SLI-DR. It's pretty pathetic. It's one thing to talk about the voltages and settings available in a BIOS, it's a whole other thing to really use them and test them all for performance. The review states that the DFI board hit 350 mhz FSB. That's interesting, but no one is going to run 350 mhz FSB. What's more important is how far you can take a CPU and memory on the board and compare it to other boards in its class. A motherboard's overclocking potential is a lot more than voltages and settings, it's about stability and compatibility with fast RAM and CPUs at high speeds. Look at the review on Anandtech, they cherry pick an ultimate CPU and unbelievable RAM and then overclock the crap out of it to really push the motherboard to its limits, giving you exact FSB, multipliers and RAM timings that worked. Look at the discrepancy in overclocking potential between the Gigabyte/Asus to the MSI/DFI. This is a huge huge deal to the majority of [H]'s core audience and yet it was barely touched in any of [H]'s nForce4 reviews.

Anyway, I'm not here to bitch at [H]ardOCP. Just giving some tips on appealing to their audience a bit more. Because [H] has always been supported by enthusiasts and these days, overclocking is becoming more and more mainstream.

I've talked about this before. I don't really care about motherboard performance comparisons. These stupid benchmarks done comparing Doom 3 performance at stock speeds between a thousand different motherboards is totally useless. Unless something is seriously wrong with a motherboard, it will perform virtually identically to other boards using the same chipset. These benchmarks waste reviewers' times that could be spent on much more important motherboard issues such as stability, compatibility with many different types of RAM, hard drives, CPU coolers and need I say it again: Overclocking, overclocking, overclocking!!!

My 2 cents,
Adrian
 
I have always liked Anandtech's reviews. The more indepth, the more I like it, but that isn't saying [H] reviews are bad by any means.
 
dave_graham said:
you're dating a man?!?!?!?!

:D

dave
Click to expand...

Nope, just a girl with a man's name :cool:.

Its hilarious when a bunch of us are sitting around watching one of the Rocky movies and we all yell AAADDDRRRIAAANNN over and over again. She hates it :).

BTW, sorry for thread crapping lol.
 
I like to read both. They usually both touch on things that the other doesn't.
 
For me....the OCP (which I had always thought stood for Over Clockers Paradise) is one of three sites that I visit regularly to help me make upgrade and purchasing decisions. Sadly, as much I like the OCP I have to sort of agree with asmielia on this. I still think that OCP does AWESOME reviews, and let's be honest...THEY WERE THE ONES WHO IDENTIFIED PROBLEMS with reviews and benchmarking and then this legal BS came up and I really think it had an impact on the direction that they WANTED to take...that direction being changing the scope and nature of the reviews. I still think there is plenty of room for innovation for me (personally), I would like to see a move further away from the the OCP side of things...and more toward the [H] side. What does this mean exactly?

Building a system for me now is the same as it was 6 years ago.
Install Mobo, Put in RAM and CPU, Put in Vid and PCI cards, Make sure jumpers are set right, put in psu, power up, tweak bios, memtest86, install and tweak os and drivers...bam you are done.

What has dramatically changed for me is how this technology is impacting the world outside of the gaming comp. I am one of the people who has flung himself into the digital chasm head first and I AM LOVING IT! (like mickey d's). I would like to see more articles on setting up the [H]ome network, the [H]TPC...[H]ome theatre audio, how to build-up and manage media, how to make accessable, what devices add functionality. Maybe it's just me but 1200 extra MHZ is useless unless it is actually accomplishing something aside lengthening your e-penis.

I guess when it's all said and done, we should really be thankful for places like the [H] and AT for edumucating us prior to buying anything. And for those ASUS SLi owners.....tisk tisk...you should have waited....we all knew that DFI would layeth the smaketh down.

M
 
I am very happy with my A8N-SLi, your results will allways very. [H] hit 300 with their A8N, AnadTech hit 255, I hit 260 (1002 bios, near stock voltages, with 350watt HEC 15amp on 12v powersupply that isn't really cutting it).
 
The [H] doesn't target the exact same crowd as Anand, at least IMO. They cover a lot of stuff and have more of a magazine feel to them. For a huge, comprehensive, and well laid out roundup, Anandtech has it down. They're a bit more mainstream than the [H], and that's not a bad thing. They do quality work.

I feel the [H] is geared much more toward the enthusiast, and I think Kyle has said as much.
 
Yeah, Master [H], I agree that used to be the case. But I'm afraid in the case of the latest nForce 4 board reviews, Anand has far and away the better reviews from both an enthusaist and mainstream point of view.

Again, it all comes back to the overclocking, [H] was one of the pioneers of introducing an overclocking section in their reviews years ago, but yet look at how little they've progressed. If anything, they might have regressed because it's just pathetic now. Overclocking has become more mainstream, hence why Anand is covering it more in depth, but what does that mean for [H]? They should have an extreme overclocking section IMO.

Here's my wishlist for an overclocking section:
-Best 1:1 1T speeds using many different types of memory CHIPS (Samsung TCCD, Winbond BH-5, Hynix, etc). Use the most popular RAM used by enthusiasts these days, like OCZ PC3200 EL Rev2 or GE Skill LE PC4400. Don't forget timings as those make a big difference.
-Best OC on CPU possible, ignoring RAM
-Test locking of all SATA ports, PCIe bus, PCI bus
-Check that voltage, multiplier and FSB settings work properly
-Check compatibility with most popular large CPU coolers (Zalman 7000/7700, Thermalright XP90/XP120)

Most of all, relevant comparisons are necessary. For example, for an nForce 4 SLI review, I would say a comparison against the other nForce4 SLI boards, one top ranking Intel board, one top ranking nForce3 board and possibly one ATI or VIA AMD based board would be the most important. Actually, only a comparison against other nForce 4 SLI boards is necessary unless it's the first review of a new motherboard chipset.

Adrian
 
I thought the Anand review was not very good. They did not mention any of the problems with the boards. They all have their own quirks, and none are mentioned. Also, its all SLI boards, and they state that the non-sli boards should be the same. Well, no they are not the same, and there are other nforce4 boards like the Chaintech that aren't SLI at all. Not everyone wants SLI. Not everyone can afford 2 6800GTs, or they may prefer to get an ATI card.

Also, they tested overclocking with a 4000+. WHY?!? Most people buying these boards are NOT buying a 4000+. They are buying a 3000+ or 3200+ to overclock.

Not very good at all.
 
[H]'s overclocking sections are written to the average overclocker. That's why they use average components such as the Koolance Exos and stuff. However, in real life, most overclockers progress past that initial learning curve relatively quickly. For example, most people know that while it is of good quality and easy setup, you can do alot better than a Koolance exos when it comes to watercooling. It isn't hard to dive into a custom watercooling setup with carefully picked system compenents once you've made the initial leap into overclocking and advanced cooling, and have done a little reading (like from here in these forums).

In other words, I think that the average overclocker is smarter than [H] thinks he is, and the reviews need to target that audience. I don't think you need to restrict your reviews to what you perceive to be the mean intelligence of your readership; if you write a smarter review, your audience will be smarter. And the case is true with this audience that no one is ever going to complain that a review was too indepth or too intelligent.

I second asmielia's wishlist.
 
Mysogonist said:
What has dramatically changed for me is how this technology is impacting the world outside of the gaming comp. I am one of the people who has flung himself into the digital chasm head first and I AM LOVING IT! (like mickey d's). I would like to see more articles on setting up the [H]ome network, the [H]TPC...[H]ome theatre audio, how to build-up and manage media, how to make accessable, what devices add functionality.
Click to expand...
Anandtech's hardware reviews are pretty good and their Tech articles are just as good as the ones [H]ardOCP does but alot of Anandtech's articles concerning multimedia and HTPCs have been worth dick.

With the direction that the hardware industry is taking it would be great to see H sort of de-emphasise performance alittle and do some articles on things that you touched on but H has always been geared towards enthusiast while Anandtech has always been more accessible to the average person and I don't think that will/should ever change
 
if i may speak my piece just briefly:

Anandtech is currently hiring a few more editors to cover areas of weakness (so to speak). among those is the HTPC market. However, don't look for anand to focus intently on the overclocking market as this is at the fringe of what they normally cover.

dave


 
I agree with everything Asmielia is saying and even thought I was writing the post.
I've said the same things in various threads. I've signed up on a few forums just to make comments on the MB reviews of today.

I'm not an extreme OCer, but that section of the review is what interests me the most. OCing is a hobby to me, and I want more coverage in it.

Currently, review sites do one of three things.
1. No OCing at all just running benchmarks. I don't bother staying in those articles.
2. OC the CPU on the stock heat sink and cooler, usually getting stuck around 235mz.
3. OC the MB to the highest FSB it can reach with a reduced multiplier on the CPU.

All of the above don't prove much to the OCing fan who wants to go higher.

MBs have been hitting roadblocks when people try their own OCs, the best example of this is the Asus A8N-SLI which most got stuck at 245mz.

The answer? Do a real world OC with the Koolance Exos setup. I'm not looking for vapo, of phase change, just water. The H is already doing this, you've got everything in place.

Koolance, good CPU at default multiplier, good memory, raise Vcore as needed, raise memory voltage as needed, try for a goal of 260mz. Take the same CPU, put in next MB, repeat. Print every detail of the test, all voltages, all temperatures, everything.

That's what I want to see if MB reviews of today. The OCing section needs more detail.
I mean, c'mon, you guys are the H. You're supposed to try for records and break stuff. Not reduce the multiplier, raise the FSB to some rediculous level, and claim the board a must have. That's nuts and doesn't prove anything. Do the OCing with the CPU at the default multiplier and go for it. The goal is 260mz, but hey, try for 275.

That will let fans know which boards have roadblocks.
 
Actually our focus has really moved towards stability, feature set, and the stability of that feature set when pushed.

I understand some of what you guys are talking about on not enough OC coverage, but then again, what you are wanting is more of a CPU OC article, not a motherboard OC article.

We cover in our motherboard reviews whether or not the features are there that are needed to OC. We USE them. We report if things are wrong with them. We don't show you CPU bottlenecks.

Truly though, I could give a damn less about benchmarks any more for the most part. We run them to see if anything is wrong, or a company has made advances that others have not. If it were my choice we would leave them out.

Stability and ease of use are out two main focuses. I want to see it work, just like we think it ought to work without having to jump through hoops.

That said, we push both FSB and memory busses which are what are needed for CPU overclocks. Voltages and tweaks are pretty heavily outlined in our BIOS section.

As for doing big roundups, I don't like that format, as it tends to not focus on what is important to us and again that is stability and features.

That said, I think we do some of the finest motherboard reviews in the world that both the mainstream user and hardcore enthusiast will find very helpful. Our standards for good motherboards remain very very high. When we suggest a motherboard, we truly believe the abilities of that product and I have to say I think we have been spot on for a good while now of picking great boards.
 
NoGodForMe said:
I mean, c'mon, you guys are the H. You're supposed to try for records and break stuff. Not reduce the multiplier, raise the FSB to some rediculous level, and claim the board a must have. That's nuts and doesn't prove anything. Do the OCing with the CPU at the default multiplier and go for it. The goal is 260mz, but hey, try for 275.

That will let fans know which boards have roadblocks.
Click to expand...

Actually that I what I think YOU guys are supposed to be doing. We are supposed to be showing the way to good equipment that can get you there.

As for our testing not showing anything at high FSB and memory bus rates, I will have to toally disagree with you there sir. Again, making the CPU the bottleneck in a motherboard review is irresponsible. If the board will run 24 hours at 100% load with a 350MHz FSB speed, that does mean something to me and I think a lot of others.

Again, making the CPU the lmitation needs to be showcased in a CPU review, not a motherboard review.
 
Thanks for the clarification Kyle. I think your poignient point, points out a key issue ;-) Someone raise your hand if I am way off but I think that a good percentage of people here at the [H] are almost looking for someone to tell them what to buy. We in tha [H] community can be classified as MAXIMIZERS (in every sense of the word) as described in this book i am reading:

The Paradox of Choice: Why More Is Less, by Barry Schwartz

Since we all attempt to get every last ounce of functionality out of every dollar we spend making a decision is becoming more and more time consuming and thus more and more stressful. This is becoming more and more apparant as different companies offer diverging strategies for hardware...some have all the features, some are for modders, some are for stability. Because everyone has different needs....there is likely not a BEST board..or chip....or Vid card. I think this is something that has to be acknowledged by [H] readers and by reviewers. For instance, take a look at AT's roundup of DL-DVDRs. When you consider price, performance, noise, DL DVD, CDR, DVD+, DVD- certain drives are better at certain things. There is no drive that is the best at all....and the cheapest to boot. Don't even get me started on the PSU thing! Anyway....the [H] does a great job....and I like how the team is open to new ideas and allergic to stagnation. Keep the reviews coming.

M
 
I think what people are getting to with the lack of overclocking coverage, is that saying that the board will get to an fsb of 350 doesn't really mean much for the mainstream user since nobody is going to be using a multiplier of 4x. That same board that hits 350fsb might not get as high as another board using the same cpu and same default multiplier with the same RAM. That is the type of overclocking that people want to see, the kind that most people will be using.

I realize that you are trying to take the CPU out of the equation, but in the real world the CPU is never out of the equation. You can equally take the CPU out of the equation by using the same one for your reviews. I dont think you guys use a different CPU for each motherboard review...if you do can I have one of the extra CPUs you have lying around? :)

I like the emphasis on stability and features, that is very important, but I too find the recent overclocking coverage a little detached and meaningless. I really like watercooling, and a good analogy would be pumps. Pumps are rated for a max flow rate at zero resistance. This would relate to the max overclock with the cpu multi set to 4x and ram timings loosened all the way. For watercooling however, the max flow rate of a pump does not mean anything, and is not indicative of the kind of performance you can expect when using it in a watercooling loop. I feel the same applies for the recent overclocking sections of the motherboard reviews, I believe they are not really indicative of what one can expect when they actually use the motherboard in a system (and try to overclock).
 
Here is a quick thought...


What about a Kyle's Box0r section. Listing the components that Kyle would choose to put in his comp?

I could see that pissing off some sponsors though....it could get them competing to get a place in the [H]otBox
 
Yes, I fully agree with Erasmus.

Kyle, I understand what you mean about taking the CPU/memory out of the equation, however, things are not always how they should be in the motherboard world. Prime example is the Asus SLI board. While it does hit 300mhz FSB+ when taking the memory and cpu out of the equation, it has serious trouble going over 250 mhz FSB when the RAM is set to 1T timing. Unfortunately, this has meant the difference for a lot of enthusiasts to not buy the board, but yet wasn't covered in your review.

The fact is, you can't just take the RAM/CPU out of the equation because even given the same voltages, same exact settings, no two motherboards will produce the same results, and in fact there can be rather large differences in overclocking performance. Voltages and tweaks are only part of the equation. A user can go look on a motherboard manufacturer's website for the feature sets and the BIOS settings available. What the user can't do is figure out how stable and compatible a motherboard is going to be with different hardware when overclocked. And that's what I think [H] should be focusing on.

I like the fact that you focus on stability, but my personal opinion is that less than a third of all [H]ard's readers will actually be running their boards at stock speeds. Therefore, stability at overclocked speeds is much more important.

I also disagree with you when you say that you're testing the limits of the CPU when doing this sort of testing. If you go into the review knowing the previous maximum speed for a CPU is x, then you can take that same CPU and plug it into a motherboard and if it fails to hit that x speed, you know something is wrong with the motherboard. Same thing with the RAM. Since you're aware of the limits of those components, they are not variables in the equation when testing.

Adrian
 
Erasmus354 said:
I think what people are getting to with the lack of overclocking coverage, is that saying that the board will get to an fsb of 350 doesn't really mean much for the mainstream user since nobody is going to be using a multiplier of 4x. That same board that hits 350fsb might not get as high as another board using the same cpu and same default multiplier with the same RAM. That is the type of overclocking that people want to see, the kind that most people will be using.

I realize that you are trying to take the CPU out of the equation, but in the real world the CPU is never out of the equation. You can equally take the CPU out of the equation by using the same one for your reviews. I dont think you guys use a different CPU for each motherboard review...if you do can I have one of the extra CPUs you have lying around? :)

I like the emphasis on stability and features, that is very important, but I too find the recent overclocking coverage a little detached and meaningless. I really like watercooling, and a good analogy would be pumps. Pumps are rated for a max flow rate at zero resistance. This would relate to the max overclock with the cpu multi set to 4x and ram timings loosened all the way. For watercooling however, the max flow rate of a pump does not mean anything, and is not indicative of the kind of performance you can expect when using it in a watercooling loop. I feel the same applies for the recent overclocking sections of the motherboard reviews, I believe they are not really indicative of what one can expect when they actually use the motherboard in a system (and try to overclock).
Click to expand...


OK, so you don't want to know the motherboard is capable of 350MHz FSB should you ever try to use it to that level, what you want to know if that it will do 275MHz with a CPU that is topped out?
 
Mysogonist said:
Here is a quick thought...


What about a Kyle's Box0r section. Listing the components that Kyle would choose to put in his comp?

I could see that pissing off some sponsors though....it could get them competing to get a place in the [H]otBox
Click to expand...


Product guides. We did three last year. :)
 
asmielia said:
Yes, I fully agree with Erasmus.

Kyle, I understand what you mean about taking the CPU/memory out of the equation, however, things are not always how they should be in the motherboard world. Prime example is the Asus SLI board. While it does hit 300mhz FSB+ when taking the memory and cpu out of the equation, it has serious trouble going over 250 mhz FSB when the RAM is set to 1T timing. Unfortunately, this has meant the difference for a lot of enthusiasts to not buy the board, but yet wasn't covered in your review.

The fact is, you can't just take the RAM/CPU out of the equation because even given the same voltages, same exact settings, no two motherboards will produce the same results, and in fact there can be rather large differences in overclocking performance. Voltages and tweaks are only part of the equation. A user can go look on a motherboard manufacturer's website for the feature sets and the BIOS settings available. What the user can't do is figure out how stable and compatible a motherboard is going to be with different hardware when overclocked. And that's what I think [H] should be focusing on.

I like the fact that you focus on stability, but my personal opinion is that less than a third of all [H]ard's readers will actually be running their boards at stock speeds. Therefore, stability at overclocked speeds is much more important.

I also disagree with you when you say that you're testing the limits of the CPU when doing this sort of testing. If you go into the review knowing the previous maximum speed for a CPU is x, then you can take that same CPU and plug it into a motherboard and if it fails to hit that x speed, you know something is wrong with the motherboard. Same thing with the RAM. Since you're aware of the limits of those components, they are not variables in the equation when testing.

Adrian
Click to expand...

I understand what you are saying and I am sorry that our content does not cover all you personally need to see. Of course that is another reason we keep these purely technical forums here, so that we can all share our experiences.

The issue you point out about the 1T timing is something that we are simply not going to uncover in our testing. It would take weeks for us to prove something like that being the motherboards fault as we would have to test about 20 different sticks of ram to be sure. Pile that on with the fact that a current review takes 40 to 80 man hours and it just is not happening.

That said, I just finished testing a Gigabyte board to the point of CPU failure, as we usually do when testing. It acted as we thought it should overclocking. Had it done differnetly, we would note the issues.....
 
Kyle, I really appreciate you being a part of this thread as I know you guys are always trying to improve. I don't want you to get the impression that I'm not a fan of your site or anything. Just trying to offer a few ideas.

I realize how long it takes to do motherboard reviews. Like I said in my original post in this thread, I wouldn't care if you cut out a lot of the stock speed benchmarks as I rarely do anything but glaze over them. Like you said, it's essentially to see if anything is wrong, therefore putting less time into it wouldn't be a problem by me.

I don't think absolute proof is required for what you're saying about the 1T timings. Just look at the Anandtech roundup. They cherry picked some RAM, then overclocked it as high as it would go while keeping timings as tight as possible. If you have 2 sticks of OCZ Platinum Rev. 2 that are doing DDR630 in other boards but yet can't do DDR500 in another, even if it's not definite proof, that's a definite warning sign that the board isn't up to snuff. 3 different sticks of RAM using 3 different memory chips would be all that's required. If you already know the limits of the RAM, then you can just set the FSB to that speed, and if the board boots up and runs benchmarks, it passes the test, otherwise it fails. I would think it would take a lot less time than running several motherboards through a whole suite of benchmarks.

Adrian
 
asmielia said:
I realize how long it takes to do motherboard reviews. Like I said in my original post in this thread, I wouldn't care if you cut out a lot of the stock speed benchmarks as I rarely do anything but glaze over them.
Click to expand...
Quiet honestly thats one of the most stupid suggestions I've ever seen. Just because your not interested in how a board performs at stock speeds doesn't mean that anyone else isn't, you go on to talk about how the board performs when OCed with different speed memory could show that the board isn't "up to snuff" so what if the board doesn't work or performance is really low when at stock yet when OCed it performs incredibly? Wouldn't the fact that it's a low performer at stock speeds mean it isn't "up to snuff?" :rolleyes:

I know theres alot of people out there at don't OC every god damn thing in their PCs that also read HardOCPs reviews, hell I find many of H's reviews to be an incredible asset when choose parts for something like an HTPC, it's one thing to know how your hardware is suppose to perform out of the box and it's another thing to know just how far your hardware could go.
 
Look at pretty much any motherboard review. How often does a board fall well below other boards using the same chipset in stock performance? I can't remember a single time that's happened. I think performance reviews should be nothing more than a sanity check type of test because for the most part, all motherboards using the same chipset perform identically.

No one would want to get a board that has serious issues at stock speeds, but I'm just saying, I'm willing to cut out time taken to do all those useless benchmarks and focus on more important things.

It's equivalent to doing a video card roundup of 6800GTs all clocked at the same speed, then spending hours and hours benchmarking to reveal that lo and behold: They all perform pretty much the same. Sure there's one FPS difference here and there, but it makes no real world difference. There has to be a way to show the board doesn't have any serious defects at stock without running through an entire benchmark suite.

Adrian
 
I like the [H] reviews on mobos.....it's all about the OCing. I don't care about benchmarks either. I just want to know what the max a certain mobo can hit is.
 
asmielia said:
Look at pretty much any motherboard review. How often does a board fall well below other boards using the same chipset in stock performance? I can't remember a single time that's happened. I think performance reviews should be nothing more than a sanity check type of test because for the most part, all motherboards using the same chipset perform identically.

No one would want to get a board that has serious issues at stock speeds, but I'm just saying, I'm willing to cut out time taken to do all those useless benchmarks and focus on more important things.

It's equivalent to doing a video card roundup of 6800GTs all clocked at the same speed, then spending hours and hours benchmarking to reveal that lo and behold: They all perform pretty much the same. Sure there's one FPS difference here and there, but it makes no real world difference. There has to be a way to show the board doesn't have any serious defects at stock without running through an entire benchmark suite.

Adrian
Click to expand...


The problem here is that each board from a manufacturer will behave different! Same as procs. For some the capacitors will be cleaner, the relays will be faster, the silicon will be neater. And then...what constitues a reprisentrative sample? 3 boards? 5, 7, 15?
 
Once again, I'm not asking for definitive answers. I'm just asking [H] to do the tests with one board. I would never buy anything after reading one review anyway. I read as many as I can find. But if [H] isn't even reviewing that aspect, it makes my life more difficult as that's one less *trustworthy* source I can use to make a decision.

Adrian
 
Hey if you really want to know what boards are the best for overclocking, go hang out at the xtremesystems.org forums.

I feel that the purpose for covering a motherboard is to reveal to potential buyers the usefulness of the features they may be buying it for. If stock settings weren't covered, don't you think the review would be incomplete, especially if there's a chance that the board may NOT perform the same as other similar boards? And what if no one covered stock settings? How would we know?

Kyle, thank you for responding in this thread.
 
Mr. Tinker said:
Hey if you really want to know what boards are the best for overclocking, go hang out at the xtremesystems.org forums.

I feel that the purpose for covering a motherboard is to reveal to potential buyers the usefulness of the features they may be buying it for. If stock settings weren't covered, don't you think the review would be incomplete, especially if there's a chance that the board may NOT perform the same as other similar boards? And what if no one covered stock settings? How would we know?

Kyle, thank you for responding in this thread.
Click to expand...


Ok guys, Here is my 2cents and what I see and do.

I check several reviews to look at the products out there. [H] is the first one I start with. I to also hit up others like Anand and the likes. Its not that I dont like or trust the [H] review but It helps to get many different views that way I can make a good edjucated decision on the stuff I want and besides its only smart to do so. It gets me started in the direction I need to go by using features and settings and the overall feel of the product that was reviewed. But I agree with the way that Kyle and crew are doin things. The whole experience of OCing is the test and trial of parts with an unknown out come. Sometimes they rock and sometimes they suck. Thats the chance that we take. I find that to many times some of us want the easy way about it. If you want more info be apart of the forums here on [H] or like mentioned in the post before mine at Xtreme systems. There is where you can find your OC info as well as be part of the input of that info. Why should Kyle or anyone else DO all the work for you and then tell you what you need and how to do it. That sucks all the fun out of it. As well as the fact its what seperates those of us that are truely [H]ard and wanna OC on our own vs being told how to do it the easy way. Think about this issue when talking about what you want some to do to review a piece. Take for instance Mobos like the K8N Neo2 Platinum have proven to be a great board to oc the 939 platform on, BUT as most know there are even production date and manufacturing plant preferences for getting a good Ocing board, It would not be possible or responsible for Kyle and crew to try to give you the info you are asking for given thus such circumstances as like the one with the Neo2. The one they review might OC great and the one you get wont do shit or the other way around and maybe it kept you from bying that board. But this is just what my 2 cents on the topic is. Keep up the good work Kyle and crew!!!

The Cpt.
 
I also read Xtremesystems.org and it's a great site. But relying on the forums and other people means waiting for them to try the board before buying it.
If all the review sites would do what is suggested here, the reviews would paint a big picture as to how a board OCs in the real world. Instead, all the sites do the same thing, OC the MB to it's highest FSB with a reduced multiplier and that doesn't tell fans if there will be a roadblock.

Want more proof? Look at the story posted by Kyle on 03/02/05 talking about the Asus A8N-SLI.
"we have been told that ASUS wants to do even better in light of folks like MSI and DFI putting out SLI motherboards that are outperforming the ASUS board in some benchmarks."

If OCing in the real world doesn't matter, then why is Asus making changes to their board? Why? Because OC fans have stopped buying it. Why? Because it gets stuck at 250mz.
 
Don't change anything Kyle I like it just how it is. :D

my bottom line:
If I want to see how Anand does it-- I'll go there (and I do.).
 
What is the main focus of [H] anyways? There's enough diversity here that I really can't tell. The system building guides brought me here (via mention on anandtech's forums), it's a good source of news, the reviews seem well rounded (my bias is that I look for stability & compatibility), the forums are well rounded and contain great info and [H] has a large enough fan base to a) spot problems early and b) keep some of the manufacturers in line if they're fudging a bit. It may not all be original content but [H] is a pretty package overall. It's simple enough to attract n00bs yet contains enough information to keep the hardcore. Sites specific to one and one thing only are great; I think if [H] as the hub to them all.

The only bitch I have is the amount of useful information in these forums that are lost or buried in time.
 
I have been kicking this around some and I am going to make an honest attempt to add a "real OC" scenario back to the basics, at least with the AMD systems. Meaning that I will take 3200+ and some speedy ram and see what we can crank the whole shebang up to...

So thanks for the thoughts you guys have pointed out here. We will see how it goes. :)
 
Great, I hope other review sites do the same and make this the new standard.

Please give all details when doing your tests no matter how small. Make the OC section a page, instead of a tiny paragraph. Reviews that give more attention to OCing will be posted on forums, talked about, bookmarked, and get lots of traffic from those trying to OC their systems. You will be rewarded for the extra effort.
 
You must log in or register to reply here.
Back
Top