Benjamonous

I've seen that the UX50's processor clockspeed tops out at 123MHz. How does this compare to other Palm Powered handhelds that run at higher clockspeeds such as 144MHz (Tungsten T), 200MHz (NX80), and even 400MHz (Tungsten C)? How does the performance compare? Surely Sony wouldn't bring out a handheld with inferior processing capabilities.

What is everyone's take on this issue? Can anyone shed any light on the matter?

Sneezy

Well, this has been debated on multiple threads, but bottom line...you can't go by MHz in comparing processor speeds. Also, it's uncertain if the benchmarks that have already been performed are accurate given the processors ability to "adapt" depending on workload. Lastly, the UX50 has still got to be fast despite it's clockspeed--it can handle video at 30fps, which I'm not sure that other PDAs with higher clockspeeds can handle.

EdFrmBrighthand

I'll repost what I said on this earlier.

My friend Eugene, who lives in Tokyo and has access to the UX units on display there, ran yCPUBench on the new models. This is a benchmarking app designed to test Palm OS 5 handhelds and runs a variety of tests. I'm not spending 20 minutes putting in all the results so you'll have to accept that this is representative. It is a floating point math intensive benchmark. It calculates, using double precision math, a Mandlebrot set (depth of 100) for a 32x32 size monochrome bitmap.

UX50:
Time (mS/Loop) = 796
kFlops/s = 190

NX80V:
Time (mS/Loop) = 512
kFlops/s = 296

This shows the UX50 is slower than the NX80V. No surprise, as we know the NX80V uses a 200 MHz XScale processor, while the top speed of the Sony Handheld Engine is 123 MHz.

He also ran Speedy 2.6 on it. This performs a processor speed test, a memory speed test, and a graphical speed test. The model for comparison is a Palm Vx, which has a score of 100. The UX50 got a score of 2174. The NX80V got a score of 3125.

Again, this shows the UX50 is slower than a NX80V.

Of course, this begs the question of why Sony used a processor it knows is slower. It has repeatedly emphasized that the Handheld Engine uses very little power. Plus it has the built-in Digital Signal Processor (DSP) and graphics accelerator chip, allowing it to show video at 30 fps at 320 by 480 pixels.

__________________
Editor-in-Chief
www.Brighthand.com
News, Views, and "Hands-On" Reviews of Mobile Devices of All Types

NJL!2016

correct me if i'm wrong, the UX's processor is slower than any other OS5 device, but sony has included some other hardware to optimize video performance?

__________________
NJLx016
HP iPAQ rx3715
Motorola v710
Creative Zen Micro 5GB
Nintendo DS

Omnitron

The processor is supposed to be variable speed - using only as much power as warranted. Granted you're going to need more horsepower for watching a video than reading an ebook. Trick is, how does it know when to shift gears. And giving that, does the benchmark software give the appropriate signals to put it in the appropriate mode to run such an app?

If we have multiple processors at work, the combined effective speed may be greater than the simple mhz rating -- we don't know yet. I'm not going to put my faith in only one test.

Granted I'd hate it if I'd have to get a patch or set a setting to determine optimum processing speed for each App. But that may be the case For all we know they skipped a step and went straight to running the program above and as such, its running in 'low gear'.

__________________
--PhotoAccess Gallery--Daily Gadget * "Life thru a UX50" Photo blog *Personal PDA History* *Palmforums*

thooeu

Yes, Sony has done some optimisation with the Handheld Engine. It is a all-in-one chip that includes the ARM core, Graphics, DSP and other I/O interfaces (such as camera, MS, etc). It also has 64Mbit DRAM built in. This chip allows Sony to use a higher bus speed (128 bit). You will recall that Intel's XScale PXA255 is faster than the PXA250 because it has a faster bus architecture.

No doubt that CPU clock speed is important, but pls bear in mind that to perform a task, many other aspects are involved. This is also true for a PC. With GHz CPU, the bottleneck may happen somewhere else.

In Japan, among other things, UX is marketed as having the capability of smooth video playback and stamina (long) battery life. Bear in mind that the battery of UX is 30% smaller in capacity compared to other Clies. If it can really play mp3 for 16 hours, I will be very pleased. My first Clie N700C managed abt 13 hours with the hold button on.

I have tested some mpeg 1 videos on the UX50, TG50, NX80V, NX70V and NZ90. The best video playback was on the UX50. One of the videos was encoded with the following: 320 x 240 30FPS, CBR at 1150kbps. The 128 bit bus architechture does play a part for video playback.

Benchmark test could be tricky for this CPU as it is a varible speed CPU. We do not know how the CPU reacts to the benchmark program compared to other fixed speed CPUs. So far, I have not observed any program running slower than my NX70V.

Of course, even with my frequent visits to the Sony show room, I could not possibly test the claim on battery life. Will try that when I get my unit this Sat.

For tasks that require raw CPU speed, I will think UX is slower than all the other OS 5 Clies, after all 123MHz is slower than 200MHz (or 400MHz for that matter). In a perfect world with unlimited battery life, of course, the higher the processor speed, the better. However, given the requirements of being a multimedia and mobile wireless device, Sony tries to strike a balance among the various factors. I think this will not be the last Handheld Engine from Sony. When conditions permits, there will be a Handheld Egine with higher clock speed.

For those interested, benchmark tested with the Speedy 2.6:

UX50

• Calc Test: 0.41 sec (abt 244 MHz)
• Memory Test: 0.05 (abt 2000 MHz)
• Graph Test: 0.23 (abt 435 MHz)
• Total: 0.69 sec (abt 435 MHz) / 2174% of Palm Vx

NX70V

• Calc Test: 0.24 sec (abt 417 MHz)
• Memory Test: 0.12 (abt 833 MHz)
• Graph Test: 0.13 (abt 769 MHz)
• Total: 0.51 sec (abt 588 MHz) / 3061% of Palm Vx

NX80V

• Calc Test: 0.24 sec (abt 417 MHz)
• Memory Test: 0.11 (abt 909 MHz)
• Graph Test: 0.13 (abt 769 MHz)
• Total: 0.48 sec (abt 625 MHz) / 3125% of Palm Vx

TG50

• Calc Test: 0.25 sec (abt 400 MHz)
• Memory Test: 0.12 (abt 833 MHz)
• Graph Test: 0.14 (abt 714 MHz)
• Total: 0.51 sec (abt 588 MHz) / 2941% of Palm Vx

TT

• Calc Test: 0.51 sec (abt 196 MHz)
• Memory Test: 0.17 (abt 588 MHz)
• Graph Test: 0.42 (abt 238 MHz)
• Total: 1.1 sec (abt 273 MHz) / 1364% of Palm Vx

TC

• Calc Test: 0.14 sec (abt 714 MHz)
• Memory Test: 0.06 (abt 1667 MHz)
• Graph Test: 0.07 (abt 1429 MHz)
• Total: 0.27 sec (abt 1111 MHz) / 5556% of Palm Vx

Note:

• You can see that the memory speed test for UX is really fast.
• I ran the test myself on the UX50, NX70V and NX80V. The numbers for TG50, TT and TC were from the benchmark recorded in Speedy program itself.

Benjamonous

Thanks Thooeu,

That's very intertesting. So, am I to understand that although the UX50 has a lower top clockspeed, it can still outperform other "faster" Palm Powered Handhelds in many areas? That's awesome it can play fullscreen videos at 30fps.

Are the benchmark tests biased at all? The UX50 adjusts its processor speed to tackle the task at hand, which can throw off the benchmark test results. It sounds like the current methods of benchmarking aren't capable or optimized to test the full capabilities of the UX50. Could this be the case?

Also, I read somewhere that it will have a version of OpenGL to help process graphics. So, really this machine should be able to handle graphical games much easier than other handhelds, right?

thooeu

Benjamonous, UX outperforms other faster devices on specific tasks such as video playback. I am sure there are other tasks where the other faster devices will outperform the UX50 especially those that require the raw CPU speed. Sony designs and markets the UX with specific tasks in mind. For those tasks, UX should perform well.

I am not a big fan of benchmark and do not have much experience with it. I included it here for those people who are curious to find out with the current benchmark programs. It may well be true that the benchmark "under rated" UX performance. But if I use the Memory test as a guideline, it seems that the benchmark gives a reasonable indication. For example, I would assume that the Calc Test will concentrate on using the CPU for calculation. UX performs as a 123MHz machine does. But it did really well on the memory speed test which we know that that is the strength of the Handheld Engine (high bus bandwidth). As I said, I am no expert in this area. I am more interested in using the machine and observe the performance (such as the video test).

Yes, UX50 supports the OpenGL 3D graphics API. The 3D launcher that comes with the UX50 was developed using the OpenGL API.

Benjamonous

Does anyone know how the UX50 compares to other Clies in reading from and writing to the Memory Stick and Memory Stick Pro? The NX models have a faster processor, but the UX50 has a faster front side bus, thus enabling higher bandwidth. How does this affect its interaction with the Memory Stick compared to other Clie models?

cykalan

thanks thooeu, that's a very informative review/reply...

__________________
Alan

Now you are really flying.
www.cathaypacific.com