juanrga (nospam.delete@this.juanrga.com) on August 8, 2014 11:00 am wrote:
> But some of those companies were showing benchmarks behind the doors. And Nvidia has already admitted
> in public that those ARM64 CPUs can feed its faster GPGPUs the same than with traditional Xeons.

"Admitted"? You're spinning this as if it's something NVidia only grudgingly accepts, and using that spin to imply credibility. But it's actually in NVidia's interests to promote this idea. They've spent a lot of money developing ARM SoCs and CPUs.

Your behavior in recent posts is interesting and revealing. When you find that (IIRC) one of three authors of a paper you don't like happens to work in the same lab as people funded by Intel, you shout about tenuous Intel connections being proof that the paper is ridiculous while providing little substantive criticism of the paper's contents. But when it's an idea you want to be true, suddenly this hyperskepticism about motives flies out the window and obvious corporate marketing is treated as gospel.

> If you really believe that those companies will be hiding numbers during next three years expecting people
> to pay them hundred of thousand of dollars in the base of paper claims then I have very little to add.

If you really believe the private sector never pours billions of dollars down ratholes based on paper claims, there's nothing anyone can do to help you wake up, but I'll try to explain one aspect of how the real world works to you anyways.

There are tons of investors who make poor investments. Most of them are the very same people who make good ones: the tech VC industry operates on the principle of spreading money across many high risk / high reward projects. IIRC the failure rate is well over 50%, but successful VCs make up for it because the big hits can pay back their original investment many times over.

Now of course these guys usually aren't fools. They don't want to put any money into companies that have no hope, and they perform due diligence investigations so they're not relying just on personal judgement. But in the end they're going to make some mistakes. It's unavoidable.

Now let's talk about the paper claims. If you're a pile-o-ARMs startup and you think the story you've crafted has some VCs on the hook, but you're aware there's some weaknesses in it, the last thing you want is publish the numbers. The rest of the world will pick them apart, the VCs will realize they've been had, and the next round of VC funding will vanish.

Maybe you even believe you'll be competitive in the long run after another product generation or two, so you're hiding today's numbers to stay in the race. Maybe you're even honest about this with your VC partners. But that's still not a good reason to go public. The unfortunate truth is that honest self assessment does not sell in our society. We demand irrational self confidence from corporations and leaders and athletes and everyone else, so there's a strong pressure to hide bad results with spin and innuendo.

When you tell yourself that the people doubting these unverifiable stories of ARM server superiority must be crazy because real money is being spent on ARM servers, you're not doing yourself any favors at all.