Audio/Video, Hardware

The Haswell Leak and IT Websites? Dilemma: An Insider?s View


This summer it?ll be exactly 28 years and nearly 2,000 articles since I published my first IT feature story, about, what else, high end PC CPUs. It was a paper publication owned by a state company in a slick 15-storey tower in the capital of Slovenia, Eastern Europe, and at that time the respect for IT press was quite high. We, in fact, even had our trips paid by the magazine company, as asking vendors for sponsorship was considered only a matter of last resort. Advertisements rarely, if ever, affected the editorial policy and the trust between the press and the readers was reasonably high, although varying from publication to publication.

Since then, I wrote technical features in many places, including the famous BYTE magazine, the pinnacle and gold reference of print IT journalism, likely never to be surpassed in that aspect at least, although this website may have some surprises for you in that respect.

Of course, at that time there was far more to actually write about, as the end of the last century had half a dozen of major CPU players alone, not to mention different instruction set architectures ? many of them far more advanced than the complicated old X86: remember the Motorola 68K, which enabled the innovative Xerox PARC and Apple Mac GUI that was far later followed by Windows? Or the superfast, elegant MIPS and Alpha RISC processors killed off for the bigger corporate interests in the US of A, but happily developed further in China now? Eh, I had honor and luck to write about, review and, yes, integrate such stuff. At the platform level, this was accompanied by a great deal of variety in the component, memory and overall system design. Remember how many GPU, memory and disk drive vendors we had, say, 15 or 20 years ago?

Fast forward 28 years: both the market and the press, over time, have changed ? in some ways for the better, and in many for the worse. Yes, the once space age-like awe for the IT industry has now deteriorated to a level of respect one would have for a neighborhood car repair shop, in many ways ? the same has happened to the media. Now we have a de-facto duopoly of two architectures which both were considered kind of lackluster technically ? yes I?m talking about X86 and ARM. The GPUs are same, with just AMD and Nvidia are around, with the remnants of PowerVR and likes hidden in a variety of mobile processors. The memory, storage, etc are pretty much one or two ?standards? played around by a small group of vendors. And, except for in the mobile market, there isn?t much competition ? or worthy news ? to talk about most of the time.

At the same time, the paper IT press is now not much more than a distant memory, with a few still hanging on ? and matching lower level consumer attitude to survive on sponsors and ads, while the once-myriad of online IT sites is being pared down to fewer and fewer every month, competing for the vendors? and readers? attention against thousands of on-and-off bloggers who, sometimes, purely for fun or to extort marketing dollars, affect the vendors and their products.

Do you think that a near-monopoly vendor like Intel would, for instance, use that pitiful situation to ram down all the press and control them? Wrong ? the ?elite chosen ones? can shake even the almighty Santa Clara company, without any ways for them to retaliate.

Case in point: last month?s Tom?s Hardware ? whose founder is happily retired right here in Singapore ? with a pretty detailed Haswell desktop CPU benchmark story. While the Core i7-4770K 3.5 GHz Haswell CPU they reviewed is only expected to be announced on the 2nd of June by Intel at the start of this year?s Taipei Computex, a detailed review exactly a quarter beforehand surely helps the readers know whether or not it is worth avoiding the current Ivy Bridge desktop processors and waiting for the new ones. So, we could guess, a bit of sales damage to Intel, isn?t it?

Now, if it was a ?normal? IT website, any company affected this way, not just Intel, would have surely pulled their support ? no ads, samples, conference visits and so on. Not in this case ? just like Anand?s early leaks over a year ago with the Ivy Bridge, this was the ?elite? site not to be touched, no matter what rules they break.

You may remember that we gave regular, mostly accurate, early updates on how Haswell will look like, and you can guess our editors also have access to early Haswell samples just like Tom?s does ? but we didn?t go for early benchmark leaks, knowing the level of upset that would cause even to our confidential sources who trust us, despite the worst dry spell in two decades across the PC hardware news field. For instance, right from China, I myself could have exposed detailed 10-core and 12-core Ivy Bridge EP Xeon 5600v2 benchmarks in larger supercomputer clusters months ago, which of course would spoil a number of Intel HPC tenders worldwide trying to sell off the existing Sandy Bridge Xeon?s, isn?t it? Intel would not be able to punish these customers, for sure, as they are the world?s biggest for this year ? some 100,000 of these CPUs are already on the location here, and some ARMed competition is wrestling in there too, surprisingly.

However, in this particular desktop Haswell case, the party that dared go against Intel knowing that Intel won?t retaliate against them won that round of the game, pissing off all the rest of the websites. And all Intel will likely do is try to trace and punish the guy within the OEM company that let Tom?s guys play with the i7-4770K.
Are they more technically competent than this site, or many other well respected hardware reviewer sites around the world? Likely not, but they got the right kind of owners, budget and, yes, relationships. Can these be used to skew the outcome of vendor?s product perception or the readers? opinions? Sure they can.

Would AMD or Nvidia behave differently from Intel in this case? Likely not – AMD was, at points, far worse than Intel in this aspect, especially during that short lived period of a few years when their CPU high-end team had an advantage over Intel in performance and features. Of course, they weren?t just bad to most of the press except the ?anointed few?, but also to many of the early adopters who took sacrifice to support the early Opterons, for instance, before AMD shifted their focus on the likes of Dell and HP. That?s one of the reason why, once the leadership was again lost, AMD never again had those heroes to come back with support.

Would Tom?s or Anand website example raise a reaction among other websites possessing similar stuff to publish even more detailed ?early leaks? and dare Intel to punish them, knowing any such thing would easily backfire against Intel itself, even legally? Yes, very likely.

Now, why should this bother any of you, our readers? Because, if this kind of policy continues, coupled with the increasingly boring hardware development focused on tightly controller ?connected people?, also known as ?Androids?, rather than the quest for performance and capability that propelled the computer world for the previous decades. You?ll end up losing many of the open and technically competent IT news and coverage sources, just like many smaller but technically savvy products and vendors that were sidelined in the so called ?open market? environment.

So, would you like to be limited to a world of a few ?dominant? vendors not leading by technology quality but by sheer brute force, covered by a few ?chosen? sites not as a result of quality but by relationships, and likely, in the end, owned or controlled by the same all powerful ?shadow shareholders?? Is this not an easy beginning not just for the slowing down of technology development, but also an important element of a ?Big Brother?-like ultimate control of information? It is the choice of you, the readers, at the end?