Reviews archive at franticindustries - web 2.0, social networking, IT technology trends.

OK, the iPhone is now out, and it has a fabulous UI. I give them that. It looks and works so cool that I want it to be successful - I’ll always root for technology done right, and for technology ahead of its time.

Unfortunately, although the iPhone exceeded expectations in some ways, you can never satisfy everyone’s desires in one product. I’ve seen literally hundreds of positive previews on the net, and I agree with them. But I also cannot overlook some of the features the iPhone is missing. Here’s a simple list of the things I wish the iPhone had, but it doesn’t have them.

1. Longer battery life. 16 hours? It’s not enough. It’s so not enough that I think it will be a dealbreaker for many. Me first. It has to last an entire day. I don’t care how - users won’t care how. But they will want it.

2. UMTS. In Europe, we use it a lot. And we want it. Luckily, it can relatively easily be added to the iPhone, especially since its Europe date is Q4 2007. Which brings me to number 3.

3. I wish it were here sooner. I’ll have to wait almost a full year for it. In a year, its revolutionary features might not be all that revolutionary any more.

4. Exchange support. I don’t use it, but many corporate users do. They won’t be throwing their Blackberries away if they don’t have Exchange support - simple as that.

5. Lower price. I’ve met 10 people in the office today who said: “I’m buying it”. I bet you those 10 people will be holding dearly to their 600 euros (because this is what it will probably cost here in Europe) when it actually comes to the market.

This said, here are also some features that the iPhone lacks and people have been complaining about them, but in my opinion they’re not a big deal anyway.

1. Hardware keyboard - I’ve never used those tiny clumsy keyboards anyway.

2. More storage - come on, 8 GB is enough. It’s the first generation of the product. You cannot expect a 100 GB hard disk. It would be clumsy and it would suck even more battery life. Wanting more storage is nice but let’s be realistic here.

3. Smaller dimensions and lesser weight. Again, reality check. It plays video. It can’t get any smaller. The size is fine.

4. GPS. Sure, it would be nice. If it were possible and viable, I’m sure Apple would have add it. But GPS is still used for pretty specific purposes, and expecting it from the iPhone is just too much.

5. More megapixels. Megapixels aren’t important. The quality of the camera and the chip is. My SE w610 has a 2 megapixel camera, and the photos are worse than on some 1.3 Mp models. I guess we’ll have to wait and see how good the camera is, but I’m sure that a good 2 megapixel camera can make great photos.

I never really bought into the SLI hype. It’s a nice technology, but I was never really sure it makes sense financially. Recently I began considering adding another video card to my (a bit aging) setup, so I did some research about whether an SLI is worth it. The answer = mostly not, unless you have a big monitor.

First, lets cover the one possibility where SLI does make sense (although I wouldn’t call it a good choice). If you’re building a new system, and money is no problem, and you want the absolute high end, then by all means go and buy the best two graphics cards and use them in SLI mode.

Why I don’t think it’s good choice? Well, game developers usually want as many people to play their games as possible, not just the wealthy. So, new games always work perfectly on high-end single (and, in most cases, midrange) video cards. By paying 1300$ bucks for two cards you have a clear case of overkill. Yes, you are futureproofing your rig for a few months, but it’s not too good an investment. You can always buy a top-notch card, and then buy a better one in a few months, and it will always be good enough for even the most demanding games. However, some people like their FPS in the hundreds, so let them have their fun.

In all other cases, SLI doesn’t really fare so well.

First, let’s establish a few facts about SLI in general.

- SLI motherboards are more expensive
- Two cards draw more power, hence you might need a more expensive PSU, and you pay a bigger electricity bill
- Two cards make more noise and generate more heat
- Two cards take more space in your case
- One card is usually easier to overclock than two cards in SLI mode
- SLI can compromise the stability of your system (although it doesn’t have to be the case)
- SLI setups seem to perform better at high resolutions

Now, let’s look at two possible SLI scenarios.

1. You’re buying a new computer.

Let’s check out some lowrange cards. (The prices are taken from Newegg) An XFX 6800XT costs $129. Two cost $258. Also, 7600GS is $119 (x2 = $238).

And what are the single card options for that kind of money? Well, for $185 you can get an 7900GS, 7900GT is $209 and 7950 costs $289.

Now, observe the following comparison. Or switch to any other benchmark, it’s pretty much the same.

7900 GS beats all 6800SLI setups easily in all tests. It’s also very near to 7600GS SLI setups, although it’s about 50 bucks cheaper. 7950GT costs a bit more, but it easily wins against all mentioned setups by a fair margin.

OK, let’s see what are the options among the midrange cards. An XFX 7900GS costs $185. Two of these = $370. XFX 8800GTS currently is $479, and bear in mind that that’s a very bad price, because this card is very new. It will probably go down a fair amount in a matter of weeks. Also, a 7950GT is 289$, and it performs pretty similar to two 7900GS in SLI.

I can’t find a benchmark where two 7900GTs or 7900GSs beat one 8800GTS; in fact, 8800 GTS beats all 7xxx SLI setups quite easily. The price difference here is $100, but when you consider the fact that 8800GTS supports DirectX10, and not only equals the performance of 7xxx SLI solutions but beats them by a fair margin, I’d say that a single card is, even here, a better option.

2. Second scenario: you have a SLI compatible motherboard, a computer with one nVidia card and you are considering buying another identical card for SLI.

Answer: don’t. Whatever card you have, you’re better off with buying a new, better, single card, and selling your old card.

Let’s examine my case. I have a 7600 GT card. If I bought another, it would set me down for $139. However, a 7900 GT costs $209, and it performs nearly identical as two 7600 GTs. I can sell my 7600 GT for at least 80$, which means that I’m getting a 7900 GT for $130, and i get less noise, less heat, and less power consumption (and a new card, which means new warranty). (BTW, if you don’t believe Tom’s Hardware’s charts, which don’t always make sense, here’s another benchmark. In it, you can see the (only?) positive side of SLI - it performs well only if you go to crazy resolutions, so if you have a 24” or bigger monitor, it just might be a viable solution.

If you like, you can do your own math with other cards as much as you like: you can always find a single card solution that’s better than an SLI one. Even if it’s not, you can probably overclock one card more than two cards, bringing the gap closer.

Don’t get me wrong: SLI is nice for hardcore overclockers, benchmark maniacs, showing off, high-end setups, and hardware review sites. My point is that it’s rarely a good solution for the end user. I can only recommend it to users with really big screens, because here is where SLI shines, but taking into account the additional noise, heat, and power consumption, I’d think twice before buying a second card.