6.21.2007

Nvidia's Quest to become a Platform Company


“Keep your friends close and your enemies closer” - Sun Tzu

For years the competitive landscape between the major PC semiconductor companies was quite predictable. NVIDIA was in a somewhat of a horse race with ATI -- a horse race where the winner was the one who didn’t break a leg. The way I saw it, NVIDIA was always responsible for either winning or losing the race. In other words, ATI kept “plugging away,” and whenever NVIDIA “broke a leg” ATI would catch up -- and at least once they passed NVIDIA for a brief period.

Not taking away anything from ATI, but the bottom line is if it weren’t for NVIDIA, gaming would have been boring – NVIDIA pushed the industry to dream bigger.

If I remember correctly there were two major screw-ups in NVIDIA's history. I’m sure there were more here and there, but the significant ones affected the enthusiast space, but that’s pretty phenomenal, and it proves that NVIDIA is very much in control of their destiny. NVIDIA's software drivers were always better than ATI, and their overall execution was usually really good. NVIDIA worked their way into the mainstream by almost always consistently winning on the high end.

ATI, on the other hand, understood power management better, and its notebook platforms were typically more efficient than Nvidia’s. ATI was a much larger company at one point, and their mainstream business was absolutely huge. ATI’s Vista drivers are the best, no question about it, and now that they’re a part of AMD maybe they can start to focus on total platform solutions. AMD has some challenges swallowing the big red pill, but I think the pill is almost swallowed. A few rough patches ahead perhaps, but eventually they will get their legs and come back.

Both NVIDIA and AMD have strengths, but there’s no question that NVIDIA is good at almost everything it touches. Pre-acquisition both ATI and NVIDIA held at least one thing in common – more notably how they leveraged their relationships with AMD and Intel in order to build their businesses. In other words, partnering was fundamentally important for both ATI and NVIDIA in order to be successful.

ATI was/is a pleasure to work with, they understand partnering – and the company understands when to sit back and “listen” to its partners. NVIDIA sometimes takes a more aggressive/strategic approach to partnering, which can be really good ... and sometimes not so good.

Well, the industry has shaken to the core, and the competitive landscape is completely different this year thanks to some very strategic chess playing. Instead of four players in a “doubles tennis match,” it’s now a 3 way chess match. Yes, it’s somewhat odd. It’s hard for me to put to words what’s actually happening to be honest because there are way too many dynamics.

As many of you know, NVIDIA recently decided that they will no longer provide SLI support on Intel chipsets – I wrote a brief article on it over here. This move has profound effects on our space, especially in notebooks. Since NVIDIA flipped Intel the bird, things have been getting somewhat interesting, but somewhat confusing at the same time. AMD, like Intel, has the ability to design total platforms in house. AMD’s strength over Intel is graphics – although I wouldn’t count Intel out of the graphics race.

In the meantime, NVIDIA wants to continue to do graphics but it looks like it has Intel’s Centrino in the crosshairs. It looks to me like NVIDIA wants to beat Intel at its game and become a platform solutions provider rather than just a graphics vendor. This move causes a fundamental shift in the industry, and positions NVIDIA as the official third major island in the highly competitive “platform” space.

The obvious response to this is SLI represents a tiny portion of the market, especially in notebooks – so how is this “profound”? I am quite certain this is only a start, and there will be more changes as the year progresses. NVIDIA management knows they have something special, and they also realize in order to grow their business they need to increase their scope.

Whether or not NVIDIA wants to beat Intel at the Centrino game, they certainly want to keep AMD in check. Not only do all three companies compete against one another, but they partner with each other as well.

On a similar note, I think it is risky going head to head against a giant like Intel, and it could prove extremely rewarding or somewhat disastrous for NVIDIA.

Sun Tzu also said, “The skillful employer of men will employ the wise man, the brave man, the covetous man, and the stupid man.” I hear Intel is hiring in the graphics department – which man are you?

Read More...

6.18.2007

Staying Close to your Roots part 2?


This is an extended version of the article I wrote for print this month while we were at Computex. It's a bit different from the normal topics, but something I've been wanting to write for awhile.

---------------------------

We are here in Taipei during Computex, I’ve been here since Saturday enjoying the torrential rain and sweltering heat. I would have to say this is the most productive Computex ever and I only spent one hour on the show floor.

By now you may have heard that OCZ recently acquired PC Power & Cooling. PC Power & Cooling is known for building high quality power supplies while OCZ produces performance components for PCs. If OCZ can leverage PC Power & Cooling’s expertise in design and scale the manufacturing without changing an iota of the quality it could be good.

So at Computex it was somewhat interesting for me to see a technology that I’m 95% sure that I’ve seen before being showcased at OCZ’s booth. It was some “brain scan” controller that you wear and attempt to play games. It sounds really lame, but it’s somewhat neat actually. Although it’s somewhat embarrassing when you wear it, the concept of using your jaw to “fire” and moving your head seems novel to some I’m sure. Personally I hate “wearable technology” when playing video games; it just doesn’t feel right to me. That said, it sure is far from OCZ’s roots, oddly enough. It seems as if they’re casting a wide net and seeing what they catch, hopefully that works for them.

This makes me reflect on a story I wrote awhile ago about how in business it’s important to stay close to your roots in order to be successful long term. Well, one company in the memory space has remained pretty close to their roots and it seems to have done them well.

I believe there is no better enthusiast memory manufacturer than Corsair. Corsair is known for their quality and consistency, and when we experience a problem with their memory they do everything they can to get us up and running in short order. We have tried all types of performance memory, including Crucial, Kingston, OCZ, and many others – and there have been times when we tried to switch – but we ended up returning to Corsair.

I remember one time we were using Crucial Ballistix memory and almost every single freaking chip we sold ended up being returned, sometimes more than once. It was a total disaster, and there was really no explanation for it – although we believed the memory was unable to handle the high voltage they were designed for. Props to Crucial for honouring the warranty, but the cost incurred by Voodoo was insane and unacceptable; although we probably wouldn’t have switched to another manufacturer so quickly if they had not tried to convince us that we were the only ones experiencing such problems.

Since then we switched back to Corsair and haven’t looked back.

There are also many commodity memory manufacturers who feel that by slapping a fancy heat spreader on their chips they can “fool” the enthusiast community to believe the memory is “designed for gaming”. In fact, I found one manufacturer who copied the “Dominator” series heat spreader from Corsair – almost like they are using the same machine to make them!

Let me be the first to tell you that there is much more substance to enthusiast memory than a fancy heat spreader on a PCB. In fact, very few manufacturers really understand what it takes to commit to creating low latency, consistently high quality chips. Corsair spends the time to batch test their modules, they also employ real enthusiasts who understand the complexities of performance.

Corsair works with people from Nvidia, Intel, AMD, and other companies to ensure they deliver to their promise. Most importantly Corsair does *not* switch chips/components without informing us – and their review parts performance on par with their production parts. This is NOT consistent with some of their competitors - and thus one of the big reasons we use them.

Perhaps Corsair will be a target for acquisition from someone like Micron, Kinston, or Samsung. Although there are only so many quality chips an enthusiast manufacturer can hand pick and produce before they become just another commodity.

Let’s hope that if Corsair does get acquired they don’t screw up what they do well, and equally important they should stay close to their roots.

Read More...

6.13.2007

...a look into the future of gaming



Here's an interesting look into the future of gaming coming from HP Labs. We have this technology called "Mscape", where you can create your own location based games based on our technology by visiting http://www.mscapers.com/. Feel free to check it and and be among the first to be broadcast on our site. Crank up the volume before you watch it!

Read More...

6.11.2007

Ripping the Competition a New 20"



This blog was originally written by me for Dailytech. I wanted to start this article off to update you on what’s happening lately and I also wanted to speak a bit on the much anticipated HP 20” HDX.

As you may know, at HP we have three master brands, because no brand has endless elasticity and we want to serve multiple markets. We have Compaq, HP, and Voodoo. Compaq has been refreshed and is as lively a brand as ever – check out the new CQ logo, it’s pretty cool. We are witnessing great success in many markets with the Compaq brand. A typical Compaq is an aggressively priced no-nonsense series with straight forward configurations and top quality parts.

HP is positioned slightly higher than Compaq, with more configurability, more personalization, and certainly a broader range of quality devices. You have all probably seen the “Computer is Personal Again” campaign, and you’ll certainly see more of this as time goes on. Voodoo is at the top of the pyramid offering high touch configurations, a high level of personalization, fanatical customer service, and incredible performance.

The Voodoo brand allows us to take real innovations out of HP Labs and elsewhere within HP and integrate them into our devices quicker than normal. This brand positioning will start to make more sense in the coming months. Eight months after the acquisition of Voodoo things are rolling along intensely. Internally we have created a ton of excitement around our yet to be released product lines, and as we continue we are tweaking the business to allow us to move quicker.

These things take time, let’s not forget the official close date of the acquisition was October 1st, and we promise to start launching products in 2007. The best way I can describe this process is when Lamborghini and Audi got together they made some significant changes at both companies. As they leverage each others’ strengths they are creating some of the most fantastic cars in the world.

You guys have probably read the stories on the internet about the upcoming much anticipated HP 20” HDX notebook. I have personally seen this monster mature from when it was a wooden model to after it was born. First let’s start by saying that this notebook is absolutely kick ass, and whether they like it or not it’s going to rip the competitions odd looking 20” a new hole. There is no doubt in my mind that they will be oversold.

The design is lean and very compelling – it was designed for the higher end mainstream space, the unique hinge allows for a more natural feeling when using such a large display. The mechanicals are complex, yet so well designed that it’s somewhat amazing that the team pulled it off! The HP notebook team has done some amazing things in the last few years. Lately they have stepped out of the norm and started to create some extremely compelling designs that meet high form and function together.

The HP 20” HDX was originally conceived by the notebook design team before the gaming business units was even setup! So to be clear, the gaming business did not design this notebook, which is pretty wild considering this notebook is perfect for gaming (I should mention that the HP 20” HDX plays games better than 98% of the gaming notebooks on the planet (with MAX MAX MAX settings) and it seems to be priced to compete). Yes, the media speculated, but the bottom line is notebooks take much longer than 8 months to complete.

This should give you a good indication of where we’re going, since the gaming group is fortunate enough to be able to knowledge share with the notebook team.

Since the acquisition we have been provided with resources to allow us to design compelling devices that we feel our customers will fall in love with. As such we are working to create some pretty masterful machines which you will see released in the future. Our systems are somewhat complex – and they are anything but conventional.

We are placing some big bets in this space, and believe me you will know when we launch a product. There will be absolutely no mistaking our creations for anything else. Of course I cannot explain why yet, but dreams will become reality very soon. Yeah I know, “talk talk talk”, but we think it’s better that we remind you that we’re still here – and that’s not going to change.


....for specs on the HP 20" HDX check out this walk through that Gizmodo posted.

Read More...

6.10.2007

Nvidia gives Intel the finger...


"We cannot enter into alliances until we are acquainted with the designs of our neighbors. "
- Sun Tzu


We are here in Taipei during Computex, I’ve been here since Saturday enjoying the rain and sweltering heat. It’s like hell in a shower and I need a glass of ice water. I would have to say this is the most productive Computex ever and I have yet to set foot on the show floor. In the meantime I’m waking up at all hours of the morning and like a schmuck I’m working instead of sleeping.

So it’s time for a story on Nvidia – they continue to do well in spite of their driver issues with Vista. People love their chipsets on the desktop side, and they continue to improve on the notebook side. Of course they are still doing very well in the graphics space, and they won the PR war for the last year. As they improve Nvidia realizes that they want to become less of a “graphics only” company and more of a “platform company”.

Based on what’s happening on the competitive landscape it would seem that Nvidia didn’t have much of a choice but to focus all their future SLI resources on nForce platforms. That means, no other competing chipset will support SLI – and Nvidia may no longer provide “bridge chips” to make it happen.

Some might say this is aggressive “world dominating” behaviour, but if you look at it from Nvidia’s point of view it makes business sense. Nvidia has something that people want – nForce is pretty awesome.

It’s so awesome that last year Nvidia almost took on too much, they spread themselves somewhat thin and in the process they did a poor job of executing effectively with Windows Vista. The good thing is they realize this and I believe they have decided it’s important to have laser focus on their own branded platforms. If you were Jen-Hsun what would you do if two giants were bearing down on you?

Nvidia has always held a dominant position in the enthusiast space with their nForce platform, but they like everyone else have had somewhat of a tough time penetrating Intel’s dominance on Centrino/notebooks.

Well, no longer the case, Nvidia is using SLI, among other very cool very innovative features as their deck of trump cards.

The landscape has changed immensely this year – more consolidation has made them the third giant in the industry. They need to protect what makes them strong, and they don’t want to pimp graphics all day long. Nvidia wants to sell entire platforms while remaining completely agnostic on CPU choices.

It sounds like a good business model to me, and in my mind they aren’t “limiting choice” because we could still choose AMD graphics as a partner if we wanted.

So is Nvidia really giving people “the finger”? No. I think they have a valid explanation as to why they need to pursue this path. Do I agree with it? That’s a tough question, I guess time will tell - Nvidia may find out sooner than you will whether we agree with it or not.

The chess board keeps changing, no doubt about it. I wonder if Intel will try to buy Nvidia now... That could be expensive. Nope, not gonna happen.

...but whether AMD or Nvidia like it or not Intel will likely be the third major graphics vendor in less than two years.

small print: I did not break this story, the head of PR for Nvidia has been seeding it around, and I didn't click publish until someone else posted it.

Read More...