Bizarrely, they do have to build a revenue stream. You can't invest billions of dollars in research and billions of dollars in Fabs that quickly go out of date if you don't have a revenue stream. That's why they've always crippled the chips in software (or AMD laser etching contacts and other schemes to configure chips post-production).

Intel's profitability is mostly determined by how successfully they create value. If they can build a better chip than AMD, because they invested more money in research, then yes they can charge a higher price for it. If they do this in the long run, rather than crippling the company chasing short term profits, then they will acquire economies of scale because their leading technology position will (hopefully) hand them a leading market position.

You could argue AMD is evil for under-investing in research, or over-charging for CPUs, since AMDs CPUs are often offering you less performance at the same price. If they were you wouldn't even care that Intel was offering software upgrades because you'd only be buying AMD CPUs.

Similarly, the DLC mentality in games is partly driven by the rising cost of building games. Many games now involve a team of 100+ people working for multiple years full-time. By contrast, many popular games of the past were created by one person in 6 months.

Fortunately, due to economies of scale in the game market, you can now choose to buy new $10 games that two person built in two years through Steam / XBLA / PSN / Wii Shop, or you can choose to buy $90 games with $120 of optional expansion packs created by dozens of people over several years. You can also buy past hits that you missed (sega classics, etc.) for $1-5 a pop, a tiny fraction of what people 10-20 years ago had to pay for the exact same game.

Both gamers and CPU purchasers have never had it so good thanks to past R&D investment by companies selling amazing products at low prices. Contrast that with home buyers if you want to see financial manipulation and greed.

The problem is that no matter how much technology evolves, the relative performance / efficiency of the FPGA will always be inferior to hard-wired logic by a similar performance ratio.

ARM chips with FPGAs are not, and will never, be competing against fully integrated SoCs like you find on a modern mobile phone with hardware video encode / decode functionality and other special circuitry. That's because implementing these functions in the FPGA is orders of magnitude worse (not just cost but power consumption and performance) than hard wired logic, at best usually as good as implementing on a GPU, and for a large set of problems worse even than a general purpose CPU.

The main reason Intel is providing the FPGA is for I/O connectivity. An FPGA can have a small amount of seperate, dedicated logic for multiple tasks, allowing it to elegantly handle difficult timing requirements. Its allows you e.g. to connect an ISA device to your modern CPU without having to buy a customised CPU with an ISA bus, or a seperate ISA - PCIe bridge. I wouldn't call myself an expert but I have succesfully programmed FPGAs to do things more efficiently than the equivalent price / power Intel CPUs. Its certainly not a good general purpose solution to most problems though. You tend to select an algorithm more amenable to FPGA implementation, rather than simply implement the best algorithm (the one that gives you the best result) more efficiently. I've seen people who didn't understand how FPGAs work try to implement inappropriate algorithms and end up needing a $2,000 FPGA chewing 100W to do a task slower than a $200 30W Intel mobile CPU.

In some ways, FPGAs are less exciting for computing tasks than they were historically, because nowadays we can buy chips with hundreds of simple, low performance CPUs (ie. a GPU) that are as good or better than FPGAs for many simple, repetitive calculations.

By contrast, for I/O, FPGAs are more interesting than ever, because you can implement a lot of low-moderate performance I/O on a small FPGA, rather than having to build specific SoCs for specific devices, or connect external I/O chips. The cost of building this FPGA into the CPU has decreased much more than the cost of designing / packaging two chips, or manufacturing a custom chip in small quantities.

A large chunk of Intel's cost is R&D. Intel could do the R&D, give it away for free to companies who then manufacture and sell the chips at close to cost. Unfortunately this model would only work if Intel was doing all the research for free. This means either Intel borrowing enormous amounts of money, then declaring bankrupcty and causing a massive financial loss to those who lent the money, or alternatively engineers with families getting paid $0 salaies and paying for their equipment out of their own pockets on top of that.

Another scenario is Intel skipping on research, continue to manufacture on their previous technology, and create only 1.8GHz chips. What they do instead is invest in research to build 2.0Ghz chips, then using the new technology manufacture 2.0GHz chips (for those who are prepared to help fund the research) and as a side benefit produce cheaper 1.8GHz chips for those who aren't. Everybody wins.

Investment and research is not necessarily a zero sum game. From Intel's perspective, this research is hurting their short term profits. Its competition in the market between Intel, AMD, etc. that drives the continuous improvement in value. If Intel stopped refining its technology then AMD would eventually have a large cost / performance advantage and destroy Intel's business. Sure some of Intel's short term profits derive from brand power, but this would quickly evaporate in the long run if they failed to compete on technology.

When companies giving you better value products every year causes you to go into militant nerd rage then you probably need to re-evaluate your priorities. The reason their are not many competitors is because what AMD and Intel do is very difficult, and they are very good at it. Even then AMD makes a loss most quarters because prices are so low in the CPU market. Competing against Intel is suicidal because they are a very efficient business (aided by vast economy of scale), and they've shown a willingness to stay price competitive with their rivals, rather than always trying to extract an "Intel" premium based on brand power alone. Their immense profitability derives from their phenomenal quality record (due to strong investment in testing at the expense of short term profit) and strong investment in research (again at the expense of short term profit).

I think market cap is an important metric of a company from an investor's perspective, but has only indirect relevance to the size of the company.

Basically, the largest company by market cap (roughly) is the one expected to make the highest total discounted future profits, not the "largest" which would more obviously be defined by total sales, total employees, total capital (cash, land, buildings, equipment). Exxon represents a bigger slice of the US economy, even if its profits are lower than Apples. The difference is Exxon spends more money paying employees (and indirectly income tax) and other things that don't directly improve value from an investor's perspective.

Obviously investors want to own a company that is going to net them strong profits in the future. Therefore a massive company (tens of billions in sales, hundreds of thousands of employees) is not particularly valuable if its cost of good sold and marketing expenses leave it with a profit of 0.00001% of revenue. In fact, this sort of company is high risk because a very small increase in costs could quickly send the company bankrupt. Similarly, a company that has massive revenues but averages a loss on each sale is often near worthless to an investor. It may have valuable assets that can be salvaged, but that may require a write off of a portion of its debts which will usually wipe out the existing investors.

Similarly, a company like Microsoft who has high revenue, negligible expenses and a cash mountain is an excellent proposition from an investor's perspect (unless its sales are expected to decline, but even then its got some intrinsic value simply from the pile of cash it could pay out if it decided to wind up.)

Market cap is therefore roughly a function of the investor's average expected future profits. This can swing wildly day to day as new information acquired by investors can lead to reaction (and over-reaction) based on a change in perceived future profitability (plus human nature).

The other thing to remember about market cap on any given day is it does not reflect investors in the company as a whole, but instead reflects those who decided to buy or sell shares on a given day. This is in turn influenced by the price and the information available on any given day. The classic example of this is Petrochina which briefly became the world's largest company by market cap at close of trade one day because some idiot (or maybe just some hacker doing it for the lulz) drastically overpaid for the last couple of shares bought that day.

I'm sure that since you have recommended the abortion of C++, the people who wrote Webkit, Firefox, Microsoft Office, Open Office, KDE (and most of its associated apps), Doom3 Engine, Unreal Engine, Call of Duty, Super Mario Galaxy, Halo (and every AAA game ever basically), etc. will realise their incompetence / lack of experience and switch to C, Lua or Python since those are your personal preferred languages "now".

C++ is a heavy-weight language, but it does not generate heavy-weight or poorly optimized code. If that were true it would never have been used for Symbian, Android, every modern game engine, Internet Explorer 4 running on 486s, etc. Last year I was using a modern C++ compiler to target code to embedded 25MHz 386s and no optimization work was required to handle thousands of transactions per second. The C++ compilers today actually generate faster code now than they did when the 386 was in its prime. The core design of C++ is extremely elegant once you have an appreciation of the difficulties involved with balancing type safety, performance, meta-programming and other concepts that become significant when writing large, high performance applications. It does a good job of elegantly solving problems that few if any other languages even attempt.

C++ is not the best tool for every job, and certainly not a useful tool for amateur programmers, but it's a fantastic tool nonetheless. Your personal lack of experience does not make it a bad tool for use by professional programmers who are qualified. Its like trying to claim a heart surgeon's instruments are worthless for saving lives simply because you are not skilled in operating them.

C++ and Java are both widely used internally at Google, Oracle, IBM, and C++ is used heavily at both Microsoft and Apple. When you say java programmers (like those at Google) rarely rise above being mediocre programmers while complaining that C++ is too difficult to learn, your arrogance is matched only by the irony of your words.

I stand by my statement, if your old Microsoft / Sony locked down PC is too slow to run 2007 games at a decent resolution, then its not appropriate to comment on Duke without a disclaimer that you used some terribly out of date hardware. Nobody posted comments about the choppiness if Duke3D on the Genesis and blamed it on the game itself.

I very much doubt most of the sales will be on the consoles because they are out of date now. Your typical office PC now is more powerful than an xbox360 / PS3. I own a PS3 and Wii personally and the last game I bought for either of them was Donkey Kong Country Returns which I only play in the Dolphin emulator on a PC because it looks so much better at Full HD with AA. The game looks awful on the real Wii, but its one of the best looking platformers of the PS3 / xbox360 generation as long as you run it on some modern hardware befitting of the game's 2010 release date.

OK, I understand its cool to bag the game on here before anyone's played the full thing. But seriously, the LOADING TIMES were a problem??? Did you pre-order your computer in 1998? My LAPTOP never took more than 7 seconds to load anything, and I timed my desktop at 4 seconds off the old spinning disk (never thought to try the SSD because it was so fast anyway). The loading times were similar to Borderlands (a 2007 game) on both machines. Seriously if your computer is too slow to run 2007 games please don't bother posting your experience of a 2011 game.

Similarly with the frame-rate. My 1 year old laptop was fairly consistent ~30fps throughout the demo at 1920x1080 high detail (Radeon 5730m) and that gfx card is less than half the performance of a low end $100 desktop card (1/3 the performance of a desktop Radeon 5770 for example). Even my 4 year old laptop can run the game smoothly at 1024x768.

Honestly my only gripe with the demo was the lack of enemies and how ridiculously easy it was on hard. To be fair though, this is probably the very start of the game and I'm hoping it gets better later. I for one will not be cancelling my pre-order.

Obviously you've never worked in a highly competitive mass-market industry. $1 billion R&D / tooling cost is pretty normal for a mass market GM vehicle platform. If you're selling 1+ million vehicles at $50k then $1,000,000,000 is chump change. If you spend $100 million on development / tooling you'll either lose out badly on unit costs, lose out badly on quality or both against someone like Toyota, GM, Ford, Volkswagen, etc. who are plowing the $1+ billion necessary into each platform. This is not news purely because GM went into the volt expecting to lose money the first few years. Its not the million vehicles they sell over the next few years that they care about (that's tiny compared to their pure petrol / diesel volume), its the several million hybrids or all-electric vehicles they expect to be selling every year by 2020 that they're focusing on.

I don't know what is most ironic, the pope's original statement, your post being modded informative or the moderators demonstrating some of the pope's concerns in action by moderating your post as informative.
"The idea of living 'as if God didn't exist' has shown itself to be deadly: The world needs, rather, to live 'as if God existed,
Hence, the Pope urged Catholic journalists to be well-rooted in a decision to always keep God at the top of the scale of values.
The Holy Father exhorted the journalists to "constantly renew your personal choice for Christ, drinking from those spiritual resources that the worldly mentality underestimates, despite the fact they are valuable, more than that, indispensable."

Technology (if not communication technology specifically) will doom us all and repent ye sinners are exactly the interpretations I came away with.

Source: Catholic.net article on the Pope's comments.

Yes of course, because 3D Realms (legal name Apogee Software) have no experience publishing games at all. How old are you, 12?

Duke Nukem Forever is not a case of developers needing a "publisher". Its simply an inside joke. Everytime 3D Realms work on a prototype, they call it Duke Nukem Forever and the artists throw in a few silly Duke Nukem assets for the programmers to work with until they actually have a real game plan.