In case you haven’t read Chew*’s latest forum post regarding his decision to say goodbye to the overclocking community, do it now (source). His story is one that characterizes how the industry deals with community resources: low-cost exposure and low-cost R&D assets.
Chew*’s situation is a bit different than usual, although it’s possible that you have the same experience on a smaller level. As many of you know, Chew*’s pretty much regarded as one of the most knowledgeable AMD tweakers in the community. His knowledge and, more importantly, drive to figure out every detail of the system eventually led him to hardware vendors supporting him with plenty of hardware. In his case, most of it was provided by ASUS, although he also covered products of other brands like GIGABYTE and MSI. Important to note is that, although some may see it as such, this hardware cannot be qualified as free or even payment for done work.
The process works as follows. First you demonstrate your insight and knowledge on forums or overclocking rankings, then a company asks you if you’d be up for testing on of their products and provide exposure for it. Of course you agree (free gear!) and do what you’re asked. If the company thinks you did a good job, they might put you on their white-list and you might end up receiving more products in the future. In most cases, the first batch of products becomes yours when you receive them. However, over time it’s more likely that you’ll end up having to send them back for rotation (as you’re not the only one allowed to test the hardware).
In above scenario, however, the enthusiast just focuses on showing what a specific product can do. Or to put it differently: the enthusiast is providing exposure for a certain product of the company. Free exposure. As long as you can get something in return for the exposure, all is fine. But when you reach the stage where hardware has to be returned, you’re nothing more than a very cost-effective marketer. The industry justifies this by stating that in return for the free labor (providing exposure) the enthusiast has the chance to test the latest hardware. It’s true, I suppose. Or is it? When looking back at my own experience, summing up the costs and benefits, I come to the conclusion that there’s a lot more hidden costs than meets the eye. Yes, the enthusiast will be happy to test the latest products, but to do proper testing the enthusiast has to spend time getting to know the product (how to fine-tune), has to spend money and resources to test the product (e.g.: liquid nitrogen), and is, in most cases, expected to provide some feedback to the company to improve their products. Each on their own, the costs are (or seem) small, but in the long run they all add up to a cost higher than you expected.
Returning to Chew*’s situation. It’s safe to say that his dedication was at another level than the one I described above. Not only did he provide a lot of exposure to the end-users, but also did a lot of work behind the scenes, providing various companies with much needed feedback on their products. In addition, he’s been actively supporting some of the bigger live events. Of course, he’s most known for his input on community forums. His dedication eventually almost landed him a (well-deserved) job at a company. In that case, the work he’d put into testing all the hardware would have been worth it. Sadly enough, it didn’t go through as planned.
In the long run, reality caught up with him. Just like how I described the process a few lines ago, in the long run the costs do not outweigh the pleasure of testing new hardware. And this is nothing out of the ordinary: proper testing requires a lot of time, money and (important!) various hardware samples to compare your findings against. Especially the latter turns what outsiders perceive as free hardware into a cost much rather than a benefit. A (personal) example:
In the early days of X58, at Madshrimps we discovered that some mainboards had flawed memory bandwidth when pushing the memory frequency due to an incorrectly fine-tuned B2B-timing. It was only because we had several mainboard samples on our test bed that we could come to this conclusion as the mainboards that had this option in the BIOS served as reference for those that didn’t. Through comparison, through hours and hours of testing, we figured out the memory bandwidth issues were caused by this one timing. Eventually, we provided this feedback to (in this case) MSI and they implemented the feature in a new BIOS. If we would only have had the MSI mainboard, it would have been impossible to consider the B2B timing as cause of the flawed memory bandwidth. (source)
Chew* found himself to be in this exact situation. In order to provide founded feedback to both the company as the community, he needed to keep the hardware samples as future reference. Or: he couldn’t sell the products to cover expenses.
That brings me back to the very first paragraph of this editorial: the industry regards the overclocking community as a low-cost advertisement/r&d tool. Marketing/strategy-wise, this is a solid methodology as most of the public demonstrations of the hardware appear to be written by normal end-users and not (biased) company representatives. Because of this, they seem more genuine and closer to the truth. Sadly enough, reality is much different. More than once I’ve picked up on a story about a company pushing a so-called normal end-user into a direction to overly praise the ‘free’ hardware. Perhaps that’s the best proof that exposure through enthusiasts is just a cheap way to advertise; after all, there’s very little room for public criticism when you received a product for free. Just have a look around the forums and you will find the people who can only talk good about a brand/product.
I could end the editorial with the conclusion that the community should stand up against this practice, but I’m confident that this will never happen. The companies are in a very powerful position as they have the decision-making power to ship hardware to anyone they want. As history has proven over and over again, if one end-user says ‘no’ to the testing procedure as dictated by the company, the next in line will respond ‘yes, yes’. Yes, it’s that dirty. To provide an example: last year, when several overclockers were banned after the LOC incident (source), some other overclocker(s) turned to the company representatives that used to support one of the overclockers with the message: “since he’s a cheater, can you give all your support to me now”. As the Dutch saying goes: one person’s death is another person’s bread. It’s a (sad) reality, but complaining about it doesn’t change anything.
Instead, I’m going to end the editorial with the note that things could’ve gone differently for Chew*. But, in light of recent disputes and discussions, it’s definitely not a shock that he took the decision to put overclocking to a halt. Being told your records are not important or significant by a representative of the same company that is providing you the hardware to set those records is demotivational to say the least. Perhaps I’m also happy that he chooses family over this ‘hobby’(*). But, we all know that the right ending of his story would be having a community-supporting paid job by any of the vendors that he gave so much of his time for to improve and demo products.
Hope to see you back soon, Brian!
(*): it’s not really known in the community, but many of the present and past top-level overclockers have relations broken off due to their passion for overclocking. Of course it would be not-done to put a list of names here, but the list is long and, quite frankly, shocking. Of course it’s up to the individual to make decisions, but perhaps the unhealthy structure of community support contributes to this.
PS: the way company representatives deal with the community varies from region to region and from company to company. In some regions, the pressure to only be positive about a free product is much, much higher than in other regions. There is a general trend of certain companies being much more aggressive community-wise, though.