AMD puts its processor lineup back on track

Hark! The long awaited B3 stepping of AMD’s Opteron and Phenom finally made its way to system integrators this week.

AMD made it virtually impossible to obtain any K10-based Opteron processors after the TLB bug caught the world’s attention last December. Desktop Phenom processors continued to ship, though the BIOS workaround for the TLB race condition severely hampered performance on some benchmarks.

The vendor who obtained the B3 sample photographed (right) couldn’t be more ecstatic. “There’s been no Opterons since November. We’ve even been shipping Socket F Opterons to fill AMD orders. This is a big deal,” he tells DailyTech.

“Pre-production” Opterons sent to Torrent search engine IsoHunt last February were later revealed as gray-market B2 stepped processors, which AMD tracked to October 2007 samples.

In addition to fixing the TLB race condition, AMD will finally increase the core frequency of the Opteron series on the B3 stepping. After the initial OEM orders are filled, channel vendors like Newegg and TigerDirect will carry the new Opterons in frequencies ranging from 1.8 GHz to 2.4 GHz. Vendor estimates put this e-tailer ship date in early April.

AMD roadmaps also indicate the Phenom and Opteron lines will reach 2.6 GHz before this Fall on the new B3 stepping. In 2009 both lines will transition from the 65nm to the 45nm process node, codenamed Shanghai, with additional SKUs at higher clock frequencies.

B3 Opterons can be easily identified by the “GH” as opposed to “GD” at the end of the product number. With the exception of Phenom and Opteron SE processors, AMD emphasizes to DailyTech that no vendor should be selling or distributing “GD,” and customers who obtain these older B2 steppings should contact their local AMD distributor.

Remember the rumor we saw floating about a few weeks ago about a Nokia N810 refresh? Well, the lads at The Boy Genius Report are reporting — on a rumor mind you — that Nokia is set to launch its N810 refresh with WiMAX at CTIA in April this year. Interestingly, the name will remain the same, and not be labeled the Nokia N830 as we’d first thought. Yeah, it’s only a rumor at this point, but we’re thinking that giving this radio-less internet appliance a bit more reach is nothing but a good thing.

Check it high-enders. DigiTimes has been milking their Taiwanese motherboard sources for information about Intel’s laptop-class, Core 2 Extreme QX9300 processor. They’ve come away with a Q3 ship date and price of $1,038 when purchasing the quad-core proc in bulk. Digitimes’ own sources had originally pegged the QX9300 for a May release. But such is the life of the muckraker.

Check it high-enders. DigiTimes has been milking their Taiwanese motherboard sources for information about Intel’s laptop-class, Core 2 Extreme QX9300 processor. They’ve come away with a Q3 ship date and price of $1,038 when purchasing the quad-core proc in bulk. Digitimes’ own sources had originally pegged the QX9300 for a May release. But such is the life of the muckraker.

Recently there has been a lot of talk about how piracy affects PC gaming. And if you listen to game developers, it apparently is a foregone conclusion – if a high quality PC game doesn’t sell as many copies as it should, it must be because of piracy.

Now, I don’t like piracy at all. It really bugs me when I see my game up on some torrent site just on the principle of the matter. And piracy certainly does cost sales. But arguing that piracy is the primary factor in lower sales of well made games? I don’t think so.
Is it about business or glory?

Most people who know of Stardock in the gaming world think of it as a tiny indie shop. And we certainly are tiny in terms of game development. But in the desktop enhancement market, Stardock owns that market and it’s a market with many millions of users. According to CNET, 6 of the top 10 most popular desktop enhancements are developed by Stardock. Our most popular desktop enhancement, WindowBlinds, has almost 14 million downloads just on Download.com. We have over a million registered users.

If you want to talk about piracy, talk about desktop enhancements. The piracy on that is huge. But the question isn’t about piracy. It’s about sales.

So here is the deal: When you develop for a market, you don’t go by the user base. You go by the potential customer base. That’s what most software companies do. They base what they want to create on the size of the market they’re developing for. But not PC game developers.

PC game developers seem to focus more on the “cool” factor. What game can they make that will get them glory with the game magazines and gaming websites and hard core gamers? These days, it seems like game developers want to be like rock stars more than businessmen. I’ve never considered myself a real game developer. I’m a gamer who happens to know how to code and also happens to be reasonably good at business.

So when I make a game, I focus on making games that I think will be the most profitable. As a gamer, I like most games. I love Bioshock. I think the Orange Box is one of the best gaming deals ever. I love Company of Heroes and Oblivion was captivating. My two favorite games of all time are Civilization (I, II, III, and IV) and Total Annihilation. And I won’t even get into the hours lost in WoW. Heck, I even like The Sims.

So when it comes time to make a game, I don’t have a hard time thinking of a game I’d like to play. The hard part is coming up with a game that we can actually make that will be profitable. And that means looking at the market as a business not about trying to be “cool”.
Making games for customers versus making games for users

So even though Galactic Civilizations II sold 300,000 copies making 8 digits in revenue on a budget of less than $1 million, it’s still largely off the radar. I practically have to agree to mow editors lawns to get coverage. And you should see Jeff Green’s (Games for Windows) yard. I still can’t find my hedge trimmers.

Another game that has been off the radar until recently was Sins of a Solar Empire. With a small budget, it has already sold about 200,000 copies in the first month of release. It’s the highest rated PC game of 2008 and probably the best selling 2008 PC title. Neither of these titles have CD copy protection.

And yet we don’t get nearly the attention of other PC games. Lack of marketing on our part? We bang on the doors for coverage as next as the next shop. Lack of advertising? Open up your favorite PC game publication for the past few months and take note of all the 2 page spreads for Sins of a Solar Empire. So we certainly try.

But we still don’t get the editorial buzz that some of the big name titles do because our genre isn’t considered as “cool” as other genres. Imagine what our sales would be if our games had gotten game magazine covers and just massive editorial coverage like some of the big name games get. I don’t want to suggest we get treated poorly by game magazine and web sites (not just because I fear them — which I do), we got good preview coverage on Sins, just not the same level as one of the “mega” titles would get. Hard core gamers have different tastes in games than the mainstream PC gaming market of game buyers. Remember Roller Coaster Tycoon? Heck, how much buzz does The Sims get in terms of editorial when compared to its popularity. Those things just aren’t that cool to the hard core gaming crowd that everything seems geared toward despite the fact that they’re not the ones buying most of the games.

I won’t even mention some of the big name PC titles that GalCiv and Sins have outsold. There’s plenty of PC games that have gotten dedicated covers that haven’t sold as well. So why is that?

Our games sell well for three reasons. First, they’re good games which is a pre-requisite. But there’s lots of great games that don’t sell well.

The other two reasons are:
Our games work on a very wide variety of hardware configurations.
Our games target genres with the largest customer bases per cost to produce for.

We also don’t make games targeting the Chinese market

When you make a game for a target market, you have to look at how many people will actually buy your game combined with how much it will cost to make a game for that target market. What good is a large number of users if they’re not going to buy your game? And what good is a market where the minimal commitment to make a game for it is $10 million if the target audience isn’t likely to pay for the game?

If the target demographic for your game is full of pirates who won’t buy your game, then why support them? That’s one of the things I have a hard time understanding. It’s irrelevant how many people will play your game (if you’re in the business of selling games that is). It’s only relevant how many people are likely to buy your game.

Stardock doesn’t make games targeting the Chinese market. If we spent $10 million on a PC game explicitly for the Chinese market and we lost our shirts, would you really feel that much sympathy for us? Or would you think “Duh.”

You need a machine how fast?

Anyone who keeps track of how many PCs the “Gamer PC” vendors sell each year could tell you that it’s insane to develop a game explicitly for hard core gamers. Insane. I think people would be shocked to find out how few hard core gamers there really are out there. This data is available. The number of high end graphics cards sold each year isn’t a trade secret (in some cases you may have to get an NDA but if you’re a partner you can find out). So why are companies making games that require them to sell to 15% of a given market to be profitable? In what other market do companies do that? In other software markets, getting 1% of the target market is considered good. If you need to sell 500,000 of your game to break even and your game requires Pixel Shader 3 to not look like crap or play like crap, do you you really think that there are 50 MILLION PC users with Pixel Shader 3 capable machines who a) play games and b) will actually buy your game if a pirated version is available?

In our case, we make games that target the widest possible audience as long as as we can still deliver the gaming experience we set out to. Anyone who’s looked at the graphics in Sins of a Solar Empire would, I think, agree that the graphics are pretty phenomenal (particularly space battles). But could they be even fancier? Sure. But only if we degraded the gaming experience for the largest chunk of people who buy games.

The problem with blaming piracy

I don’t want anyone to walk away from this article thinking I am poo-pooing the effect of piracy. I’m not. I definitely feel for game developers who want to make kick ass PC games who see their efforts diminished by a bunch of greedy pirates. I just don’t count pirates in the first place. If you’re a pirate, you don’t get a vote on what gets made — or you shouldn’t if the company in question is trying to make a profit.

The reason why we don’t put copy protection on our games isn’t because we’re nice guys. We do it because the people who actually buy games don’t like to mess with it. Our customers make the rules, not the pirates. Pirates don’t count. We know our customers could pirate our games if they want but choose to support our efforts. So we return the favor – we make the games they want and deliver them how they want it. This is also known as operating like every other industry outside the PC game industry.

One of the jokes I’ve seen in the desktop enhancement market is how “ugly” WindowBlinds skins are (though there are plenty of awesome ones too). But the thing is, the people who buy WindowBlinds tend to like a different style of skin than the people who would never buy it in the first place. Natural selection, so to speak, over many years has created a number of styles that seem to be unique to people who actually buy WindowBlinds. That’s the problem with piracy. What gets made targets people who buy it, not the people who would never buy it in the first place. When someone complains about “fat borders” on some popular WindowBlinds skin my question is always “Would you buy WindowBlinds even if there was a perfect skin for you?” and the answer is inevitably “Probably not”. That’s how it works in every market — the people who buy stuff call the shots. Only in the PC game market are the people who pirate stuff still getting the overwhelming percentage of development resources and editorial support.

When you blame piracy for disappointing sales, you tend to tar the entire market with a broad brush. Piracy isn’t evenly distributed in the PC gaming market.

Blaming piracy is easy. But it hides other underlying causes. When Sins popped up as the #1 best selling game at retail a couple weeks ago, a game that has no copy protect whatsoever, that should tell you that piracy is not the primary issue.

In the end, the pirates hurt themselves. PC game developers will either slowly migrate to making games that cater to the people who buy PC games or they’ll move to platforms where people are more inclined to buy games.

In the meantime, if you want to make profitable PC games, I’d recommend focusing more effort on satisfying the people willing to spend money on your product and less effort on making what others perceive as hot. But then again, I don’t romanticize PC game development. I just want to play cool games and make a profit on games that I work on.

I wrote this article for Custom PC Magazine in the U.K. – it’s the latest issue, and it’s awesome. A warning though, this article is pure speculation! There’s my disclaimer before people start freaking out – this is pure “hookah smoking, sitting around with a bunch of friends and chatting” speculation…! I will write an extended version of this later.

A few months ago I wrote an article on some of my own personal thoughts about AMD’s strategic position in the market. I included some ideas around AMD potentially creating strategic partnerships with companies whose pieces fit snugly in their holes (can you guess who?). In the process of doing this they would also need to lighten up some of the assets, and scale the business back enough so profitable growth can be sustained.

Its funny how things seem to make sense when you’re involved in a business for more half your life. If you put all the pieces on the board and stare at them long enough they start to tell a story.

Going further into the potential strategies for AMD, I think they could try to spin of their CPU, chipset, and graphics technology businesses into a separate entity. At the same time they could spin the foundry business into a separate entity or division. I continue to believe that somehow they would need to be linked in order to maintain their X86 license among other things.

Now imagine if they did both of the above; their value on the foundry side would be cut, but on the technology side they may get some great valuations with higher multiples, and thus the IP side would become a great target for acquisition for Nvidia.

Now the question is who in their right mind would want to own shares in a foundry business? Perhaps they could work a deal so existing shareholders get a share of one and half of another, and perhaps if Nvidia decides to come out and play they could also offer some share benefits to existing shareholders.

In my mind it would be easier dealing with two or four giants, then the three header four armed abomination we’re currently contending with. It’s just too bloody confusing.

Carbon nanotubes look to step in and pick up where copper trails off.

The scientists at the Rensselaer Polytechnic Institute, located in Troy, New York, have been busy playing with carbon nanotubes (CNT) for the past few years. Their research has brought us the possibility of paper batteries, remote-controlled disease killing bombs, and the blackest material in the world.

Saroj Nayak, associate professor at Rensselaer’s Department of Physics, Applied Physics and Astronomy, recently led a team on a research project to compare the conductive properties of copper nanowires with that of carbon nanotube bundles. The conclusion probably won’t be much of a shock: CNT bundles came out on top.

Rather than comparing the empirical data between the two subjects, Nayak’s team used the world’s fastest university based supercomputer, Rensselaer’s Computational Center for Nanotechnology Innovations, to study their quantum mechanical properties. While empirical measurements are fine for research at a normal scale, the interaction of molecules, atoms, and some of their building blocks are more accurately measured below the macro scale using quantum mechanical observations.

The team’s ultimate goal was to learn which material would be better for microchip interconnects. Copper interconnects are quickly coming to a choke point as chip cores continue their downward spiral. The current 45nm technology is not predicted to be the final blow, thanks to things like high-k metal interconnect gates. Some research suggests cores built on 15nm technology are more than feasible.

However, replacing the copper that is currently used for interconnects with a more efficient material would be a boon to chip makers and designers, possibly allowing them to even further shrink the process.

Though CNT bundles look to be a promising new material for microchips, there are still some ramifications to be dealt with before mass production could start. An economical way to grow the bundles, as well as a method to ensure the tubes themselves are 100% metallic will have to be found. A more thorough understanding of the electrical properties of CNTs as interconnects will be needed as well.

Nayak’s groups’ research will be featured in the March issue of Journal of Physics: Condensed Matter.

Researchers simulate childhood thought process, further blurring the line between artificial intelligence and biological intelligence

While some skeptics, such as Apple-cofounder Steve Wozniak, dismiss artificial intelligence insisting that robots will never be able to reach a human level of thinking process and behavior, the reality is that artificial intelligence is fast approaching human level thought process. Battlefield robots are making life and death decisions, and an international panel recently met to discuss whether robots could be tried for war crimes.

In vehicles, DailyTech witnessed firsthand the GM-sponsored DARPA robotic driver navigate a complicated course with efficiency matching or surpassing that of a human. Meanwhile, SRI National works to create DARPA funded robotic assistants which learn and organize thoughts in a human-like fashion.

As robots become more and more human-like, we face the duality of the result. On the one hand, in creating something that is human-like we learn more about what makes us human; on the other hand, by creating a replica of man, the line between human and machine becomes more blurry. As we enter the future, reality in the virtual world and real world is merging into one. Scientists already demonstrated the first “mixed reality” systems — systems in which a virtual and a real world device were indistinguishable.

Continuing along the path of convergence between biology and the digital world, researchers at the Rensselaer Polytechnic Institute (RPI) are developing complex artificial intelligence to control characters in the popular online game Second Life. These characters will be able to have beliefs, distinguish human and AI characters’ beliefs, and manipulate the behaviors of human and AI characters based on these beliefs.

The team unveiled their first creation, a 4-year old child avatar dubbed “Eddie”, at an AI conference. The avatar not only follows the aforementioned intelligence goals, developing beliefs, but also behaves psychologically like a human child. Researcher Selmer Bringsjord explains the creation process, stating, “Current avatars in massively multiplayer online worlds — such as Second Life — are directly tethered to a user’s keystrokes and only give the illusion of mentality. Truly convincing autonomous synthetic characters must possess memories; believe things, want things, remember things.”

You won’t be seeing a character like Eddie walking around on the street for a little while explains Bringsjord — Eddie’s complex behavior requires the processing power of a supercomputer. The processing power is leverage to combine traditional logic-based artificial intelligence with computational cognitive modeling techniques.

Understanding, predicting, and being capable of manipulating the behavior of humans is one benchmark of intelligence, and the principles behind how this works in the human mind is known appropriately as the “theory of mind”. The RPI team’s research marks one of the largest efforts to date to engineer based on the principles of the theory of mind. The researchers, implementing the part logic and part math theory, impart on the AI-controlled avatars an understanding of such “human” concepts as betrayal, revenge, and evil.

Similarly, they employ human-like stages of cognitive development. For example, Eddie behaves correctly in a false-belief test. In a typical false belief test a person observers an object, in this case a virtual teddy bear. When the person leaves the room, another person moves the object to a different location. Upon the return of the first person to the room, the adult observer expects them to look in the old location of the object, knowing that they don’t have knowledge of the move. However, a child four years old or younger will think that they will look in the new location, not understanding that they couldn’t see the move. In an example of a case where it’s right to be wrong, Eddie correctly believed in the “false” location, the proper “human” behavior for a child.

Eddie can also be digitally switched to have adult-like reasoning and make the correct decision. The reasoning is accomplished by an automated theorem prover. An interface takes conversational English in Second Life and turns it into formal logic, which is processed by the prover. A video clip of Eddie in action can be viewed here.

The RPI research is sponsored by IBM. The RPI team’s final goal is to place humans in a Star Trek-like holodeck filled with projected virtual characters with human-like behavior. The researchers say that they could accomplish such a simulation in theory by leveraging the processing power of RPI’s Computational Center for Nanotechnology Innovations (CCNI) and the Experimental Media and Performing Arts Center (EMPAC).

With over 100 teraflops of computing power, the CCNI is the most powerful university supercomputer in the world. It is composed of massively parallel Blue Gene supercomputers, POWER-based Linux clusters, and AMD Opteron processor-based clusters. And soon, it may be thinking, just like humans, if the RPI team continues in its success.