What Nokia should really do: Parallels with Apple in 2000

F

FRIENDLY TROLL (WS)

New member
The whole litany behind Nokia all sound strangely familiar.

It has so many parallels with Apple at the start of the decade.

The stock is tanked.

Too many models on an architecture that is being outrun.

Banking on a savior OS in the horizon.

Being asked to shift to Windows and become another Dell.

So what are the lessons Apple has in store with Nokia?

1. Stuck to your OS guns. MacOS never had as much apps as Windows, and still doesn't today. But it all survived and prospering. The whole 2000s also demonstrated the gradual decreasing importance of apps in lieu of the internet and web apps, as developers migrate to the Internet. The same long term phenomenon may happen on mobile with HTML5 technologies. Where Meego has to particularly deliver is on a fast HTML5 browser.

2. Consolidate your hardware line to a few killer models. This doesn't need explanation.

3. Do not be afraid of self disruption. MacOS 10 was an entirely new OS based on NeXT and Unix. Only the UI seems familiar, but it was definitely a clean break from MacOS 9 whose OS heritage stems back to the original 128k Mac. Apple definitely pissed off some loyalists at that time for the incompatibilities that presented.

4. Overhaul your hardware architecture. Apple went from the Motorola 68k architecture, then to the Motorola-IBM PowerPC architecture, and then surprisingly, went Intel. By abandoning the PowerMac line, Apple saved its Mac line. Its clear that Nokia should start chucking the old ARM11 architectures and shift to new powerful ones.

5. Transform the company. Apple turned from a hardware company to a media company. Can Nokia do such a profound transformation? It used to be a lumber company. One time, it even made rubber boots. Nokia needs to sign up and promote content. Music companies. Movie companies. Book publishers. Game publishers. Game developers. In turn, they form another revenue stream.

6. Management efficiency. Nokia needs to do something drastic. It employs over 300,000 people. Apple has less than 50,000. If an organization is cost efficient, it is naturally your costs will be lower and your profits rise as a result. Business basics. Steve Jobs is extremely clear eyed on this --- one of his attributes that are vastly overlooked. Look at your margins.

I would say 1 to 4 are the easy part actually. 5 and 6 is the hard part.
 
interesting...
at least Nokia has the unique situation of being the industry leader but "failing"....
Apple was at the bottom in computer sales....

EDIT- I don't know about #2, that's what makes them and Samsung the two leading manufacturers...
the portfolio has been reduced already...
maybe less at the high end, which has already been done....
 
You're definitely right in most regards, or perhaps I just think that this is the case because I agree with you in most regards. Just a couple of things.

1. Nokia employs about 130K people, not 300K. 65K are employed by NSN and 5K are NAVTEQ. That leaves about 60K on devices and services.

2. I don't think that your architecture argument makes sense. Apple switching from the Motorola 68000 to the PowerPC to x86 were fundamental shifts as the architectures were completely different in each case. ARM Cortex 8 and 9 chips use the same fundamental instruction sets as the ARM 11 chips, though. It's more like a Pentium 4 versus a Core 2 Duo or something like that and just a matter of selecting parts to suit the task at hand as opposed to spending on parts that aren't needed. Now switching to an Intel chip would be a major shift, but one that I think the Qt API makes much less painful. I have no doubt we're going to see Nokia phones with Intel processors within a year.

3. I don't think the few killer models approach would work for lower-end devices, and Nokia doesn't even have a demonstrated deficiency in developing hardware. On the contrary, it's very capable. I do agree with you in terms of smartphones, though, where it makes a lot of sense. Even if you can make a wide array of differentiated higher-end devices, you probably shouldn't.

I really think Nokia has a solid software strategy in place, and yes that includes Symbian. They just fail to execute. Some of the changes that you suggest could address this, but there's more that needs to be done.
 
Good analysis and a nice write up.

I'd say one of the main differences is the management style. While Nokia is a bureaucracy, almost like a mini-Parliament, Apple's management structure is very much 'my way or the highway' among its top executives.

Also, Jobs had a background building up a media company: he bought a quirky offshoot from LucasFilms that specialized in computer graphics and animation....and turned it into Pixar. The story is that he was still heavily involved all the way through into Toy Story 1, Finding Nemo and Bug's Life, sitting in on storyboard conferences, animator meetings and such.

In comparison, I can't fathom Nokia becoming a media clearinghouse. I think it's too much of a drastic change for them.

They have a hard enough time getting non-buggy software out the door.
 
ARM11 chips use the ARM v6 instruction set. Cortex A8 uses the ARM V7 set. Also the ARM11 is scalar, while the Cortex A8 is super scalar. To put this in context, scalar means the fastest it can do is to execute an instruction on one clock cycle, though its a mixed bag some instruction are executed in two or more. This means a scalar CPU directly scales with clock speed. Super scalar means it can execute multiple instructions per cycle. So clock speed no longer directly reflects a processor speeds. A superscalar CPU with a lower clock speed can execute much faster than a scalar CPU with a higher clock speed.

This is where the major jump begins. That is what I meant by a major architecture jump. Nokia for some reason keeps holding on to ARM11, and the most plausible reason for that is that Symbian kernel specifics are tightly coded to the ARM11. I see a similar situation here with the Blackberry OS. Both OS can be made to run on ARMv7 architecture, like the way Symbian ran on the Omnia HD and Vivaz, but they're not optimized to get the most of it. Ever realized why Blackberry is positioning the QNX Playbook OS as the eventual replacement of Blackberry OS?

There is enough of a difference that Adobe declared that Flash will only run on ARMv7 devices. That stretches across multiple ARM based platforms from Android to WebOS.

I honestly don't think Nokia has a good hardware strategy in place by continuing to use ARM11.

Comparing ARMv6 to ARMv7 ala ARM11 to Cortex A8 is more like comparing 386/486 to Pentium. Comparing Pentium M to Core 2 Duo is more like comparing Cortex A8 to Cortex A9. There is an extreme gap between the processing power you can find on something like the Motorola Atrix vs the Nokia E7 which still uses an ARM11.

To give you an idea, using the Quadrant benchmarks on Android 2.2,

Phones with 500 to 600Mhz ARM11 SoCs might be scoring just over 300.

Devices with the first generation Snapdragon 8250 are scoring about 1300 to 1400

Devices with the 2nd generation Snapdragon 8255 are scoring between 1800 to 2000

DEvices with the Tegra 2, dual core, are scoring up to 2600.

Benchmarks tend to have little or no OS overhead.

That's quite a leap.

With regards to Nokia's software strategy, fundamentally it seems sound. Just the software alone. But that's only 1/2 of the equation. Remember, when your hardware is deficient, people will also blame it on the OS. In order to make your OS look good, your hardware has to be good too. Remember those out of memory errors Symbian S60 phones tend to make? Its not Symbian itself causing that, but the fact the phones only have a meager 128mb of RAM. But people will blame that on Symbian anyway.

Just remember, Android itself didn't really took off until the Motorola Droid came with Android 2.0. That was the first piece of ARMv7 Android rode upon. The second was the Nexus One. The rest was history. The Nexus One itself didn't sell well, but models that copied it as a template exploded in the market like the EVO, Droid Incredible and the HTC Desire.
 
The one problem I have when I hear people say "but symbian runs well on minimal hardware" is that we no longer live in a world where that necessarily matters. Throw in that extra RAM just because you can.

The content we view on our phones keeps getting bigger and bigger. Websites get larger and more complex, video media online keeps getting higher quality and better resolution, 720p, on and on. So it doesn't make sense to try and use limited efficient hardware just because its efficient, when all the tasks we use phones for is growing drastically. The two philosophies are opposite each other.

I'm not saying to be irresponsible and use 5 gb of RAM just for the hell of it. But content has changed and the minimum efficient hardware philosophy that Nokia still tries to push no longer makes any sense whatsoever.

Arguably the n8 STILL could be said to not have enough RAM. After Nokia's history with this type of thing, It shouldn't even be close to too little RAM. They should have put so much RAM on the n8 to be almost obnoxious about it. That's my opinion anyway and I'm sure I'll get blasted for it:-).
 
After Drillbit explained the Armv6 and Armv7 differences it seems clear that Arm11 will be pushed down to lower and lower spec'd devices which will be able to run the new Symbian. Then they will do dual core and other higher spec'd devices for the higher end Symbian devices.
 
The Nokia N8 has like 256mb of RAM. Today, some low end Android phones like the LG Optimus One has 512mb of RAM.

Another problem is Nokia's attitude in design. Lets do the body first, the exterior finish and the camera. And then nickel and dime on the electronics to fit the budget.

The Koreans have the opposite approach. They invest heavily on the electronics first, the screen, processor, RAM, memory, battery. The works. Then to fit the budget, they would need and dime on the exterior.

The result is Nokia produces all these wondrously beautiful devices with inadequate hardware. Samsung or LG on the other hand would produce such powerful superphones but the body feel somehow feels cheap, though well engineered to be sturdy. And its working for Samsung.

HTC? They're unique. They would spend on both the innards and the exterior. You don't get these glossy plastic cased phones like the Galaxy S; HTC phones are built on aluminum unibody frames that has to be carved out with a highly sophisticated multi axis machine tools. Where they would nickel and dime is on the battery, which itself is one of the most costliest components in any phone.

In my view, Nokia needs to find a manufacturing model where it can afford to sell large 4" plus screen touchphones with Cortex A8 and A9 cores, and yet preserve traditional Nokia attributes. Chances are you may have to cut down on a few other things like the obsession on the camera.

The less hardware resources an OS has to run, the more gimped it tends to be on the UI and feature set. A lot of what Symbian has become is due to a history on running in gimped memory environments. UIs are particularly memory demanding. You want a smooth fast, rich UI,, you have to give it enough processor and memory headroom to run it.
 
One other thing I like to say is that Nokia needs to cover the tablet area. This is still a wide open and virgin field. It still easy for a new comer like Meego to break in there.

Nowadays, to have a truly sustainable ecosystem, you need both smartphone and tablet sharing the same pool of OS resources and developers. Palm and HP is right bringing WebOS to tablets. The fact that WP7 doesn't have a tablet is a strategic mistake. Google was right canning ChromeOS from tablets and going with Android Honeycomb. For Meego to be successful it needs a combination smartphone and tablet strategy, which includes outsourcing them to other manufacturers.
 
Back
Top