UMPC processors benchmarked: Ultra Mobile PC Tips
One advantage to owning multiple UMPCs or having friends that own UMPCs? You can benchmark each under relatively similar conditions to determine performance differences. That’s exactly what Frank Garcia (aka: Ctitanic) did with four UMPCs, each with a different CPU. The premise here was to use a program that strictly measures processor performance, regardless of other device attributes. That’s a tricky situation since memory, bus speeds and other hardware properties among other things can influence a test. Frank settled on chess program that has a function to specifically measure how many chess moves the CPU can process in a single second. Using a 1 GHz Pentium M as the 100% baseline he’s sharing the relative results found with a 900 MHz Celeron, 800 MHz A110 and 1 GHz Via C7-M. No surpises in the results to me and one of the key reasons I’ve kept my Q1P as a primary mobile device. For folks new to the UMPC space or interested in another measurement of relative performance, the results are well worth the look.
Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
The only surprise for me is how the “new” Axxx series is even slower than the older Celeron, which backs up what we’ve been saying here that the processor lacks punch.
Now, I might be missing something there, but seeing how I can only find the Dothan (723) Pentium M as 1.0 Ghz, with the same or faster FSB and four times as much cache; the Celeron M having as much or twice as much cache as well as a similar bus speed;as well as the fact that the A100 series is a derivative of the Pentium Ms, why *should* we expect the a lower-clocked A110 to outperform a higher-clocked Celeron or Pentium?
good for you Kev… you keep that much bulkier, 800×480, duller screened, poor joystick, barely 2hour battery life, thumbboardless (which is at least good for hotkeys
in exchange for those flaws being fixed now, i just have to wait a few more seconds for my apps to load on my Q1U
tell me something, how much time is lost shutting down/hibernating your device to swap batteries compared to me just waiting a few more seconds occasionally? since i get nearly 4hours with mine compared to your 2, i dont have to swap.
look at it like this, your car goes 20mph faster than mine, but mine gets better gas mileage. how much of your life do you genuinely drive your car at max speed? not much, most of your life is spent driving in the slow lane just like mine, in which your device is NO faster at all… unless you are playing videogames nonstop & constantly maxing out your CPU.
MSvp,
Still, though, when you’ve had a UMPC for over an year and want to UPGRADE, you would assume you would gain some speed. I will be upgrading soon, but it probably won’t be the same form factor as my Q1. It’s like why get a Q1U when the Q1 isn’t all that much different and end up with lower performance. Besides the screen, that doesn’t seem much of an upgrade. But obviously, you’re more about battery life than performance.
Your analogy with cars doesn’t fly. Sure, I don’t go the max speed on my car, but you better believe, I go 100% CPU usage on my UMPC all the time.
A better analogy is scooters. I’d rather have one that goes 40MPH so I can ride all over town rather than one that goes 20MPH where I’m limited to residential streets. Sure I can get more mileage on the 20MPH scooter, but I’d have to take the long way using side streets to get to where I’m going.
he gave the all reasons for the upgrade & there were alot. the Q1 & Q1U are radically different devices. VERY few devices have many differences between generations, but the Q1 series has had alot.
actually the gas mileage analogy makes perfect sense. it’s your scooter analogy makes absolutely no sense at all. so why is that? because these the Q1/Q1U perform the exact same tasks, 1 faster than the other, 1 for longer than the other. your analogy consist of 2 devices completely taking separate pathways, which is wrong. what you should have said instead is, scooter 1 will get you to the mall slightly faster, but scooter 2 will get you to the mall AND then to grandma’s house also (without needing to refuel)
hate to tell you, but you dont use your device at 100% “all the time”. that’s just an absolute lie, because if you actually did you would have already replaced it with something faster a long time ago. everyone who uses a PC knows that a PC constantly locked at 100% is useless.
That is the MAIN REASON why I still KEEP MY SONY U750P…
Well you got me there. But I’m not lying about the 100% CPU usage. I didn’t say it’s 100% usage 100% of the time, but whenever I’m doing something, the fact that it even hits 100% CPU onusage made me think of scooters since you can only go so fast.
Rod’s scooter analogy was completely wrong, that would be like comparing a PocketPC to a PC. but the Q1 & Q1U perform the exact same tasks, 1 faster & 1 longevity.
the ONLY reasons for faster processors are for very few tasks that max out the CPU such as, videogames, video editing, HD video, etc.
besides that, then there is NO difference in speed at rudimentary tasks. most of a PC’s life is spent near idle & not even close to it’s full potential.
a Ferrari going 50mph wont get you to the store any faster than a Prius going 50mph. ONLY when they both go at max speeds will the Ferrari begin to pull away & get you to the store faster.
so if there is any kind of noticeable speed difference between these 2 generations then people need to start looking elsewhere besides the CPU, because that’s not the problem. what about the RAM, HDD, GPU, and something extremely critical but almost always overlooked, the drivers. i wont even begin to get into the numerous OS configurations that could be causing the issues.
i really thought most enthusiasts (bloggers & forum posters) were smarter than this, but it’s becoming painfully apparent that most of you have no clue. it gives the whole community a bad name.
Ok the way I looked at the cars analogy was, buying a faster car but only driving 50MPH is just like have a 2GHz processor put throttling it down to 500Mhz to save gas.
Yes I know 100% usage for a constant time is usually reserved for tasks like encoding or games. But I’m talking about doing normal tasks. Like say load iTunes. It’ll hit 100% CPU usage. Maybe only for a few seconds (a lot more), but still, there’s that upper limit where you just can’t go any faster to load it.
If I’m going to upgrade from a slow scooter to a slower scooter with better handling, it’s still a slow scooter. So in that case, it’s just not worth upgrading.
How would the Via in the Oqo 02 compare to these benchmarks do you think?