Log in

View Full Version : samsung cpu


nuggetman
01-22-2004, 03:13 PM
are the samsung cpus in the lower-end ipaqs considered ARM? or something else? what is their performance like compared to say, a 300mhz xscale

dmacburry2003
01-22-2004, 03:18 PM
Yes the Samsung processor does use the ARM architecture. I use an iPaq 1935, which has a 206 Mhz Samsung, and it works great. I can play music in three apps at the same time (without skips) and watch mpegs at 24-30 fps. It does perform well in most games, except in more intense ones (like SimCity 2000, in which the backround music is choppy).

nuggetman
01-22-2004, 03:28 PM
since my axim is broken i was considering an ipaq 1935 since it has more ram and 64 megs, plus amazon is selling it for $187 right now.

but i dont know if i have another $130 for an SD wifi card :?

dmacburry2003
01-22-2004, 05:27 PM
Just a heads up, check here (http://accessories.us.dell.com/sna/productdetail.aspx?sku=A0160823&c=us&l=en&cs=19&category_id=4327&page=external) for the wi-fi card. It is $99. I use this model with my 1935 and it works great.

Jason Dunn
01-22-2004, 05:44 PM
The only Samsung processor being used in a Pocket PC right now is the 266 mhz one, I believe, and performance wise I found it to be quite snappy. Mhz per Mhz, I believe it's faster than the Xscale, but I haven't done any serious testing.

Dave Potter
01-22-2004, 06:48 PM
The only Samsung processor being used in a Pocket PC right now is the 266 mhz one, I believe, and performance wise I found it to be quite snappy. Mhz per Mhz, I believe it's faster than the Xscale, but I haven't done any serious testing.

Actually Jason - this is incorrect. The iPAQ 1930/35 uses a 203 MHz Samsung processor whereas the iPAQ 1940/45 uses the 266 MHz Samsung processor.

Both are very snappy indeed!

Jason Dunn
01-22-2004, 06:56 PM
Actually Jason - this is incorrect. The iPAQ 1930/35 uses a 203 MHz Samsung processor whereas the iPAQ 1940/45 uses the 266 MHz Samsung processor. Both are very snappy indeed!

Ah, thanks for the correction! I don't have a 1930...

dmacburry2003
01-22-2004, 09:12 PM
Whoops! I was 3 Mhz off.

I do think though, but I haven't tested this theory, it is just an opinion, that 266 Mhz is like a cut off to be able to run all apps at regular speed.

Ex: SimCity skips in the music, but I bet with a 266, it doesn't.
Metallion skips in the music, but I bet with a 266, it doesn't.

Anyway, this is just a guess.

dylanbj
01-23-2004, 02:02 AM
The simcity 2000 music is choppy on a 1940, maybe it has somthing to do with WM2003?

dmacburry2003
01-23-2004, 04:40 AM
Didn't know :oops:

Kowalski
01-23-2004, 01:00 PM
both the samsung and xscale are strong arm proccessors. the Xscale is more powerfull but when it comes to software side xscale runs in compatibility mode to strong arm. this makes both proccessors give nearly the same performance.

JustinGTP
01-23-2004, 05:26 PM
So if everyone claims that the 1940 can run just as fast as the 2215, how can they charge so much more for it? Sure, it may run a little faster than 266mhz because of the "architecture" - but I dont see it running 133mhz faster, otherwise they would have advertised it as 400mhz then wouldn't of they? There is not point, I might as well got the 1940!

However - I do disagree, the 1940 is slower than my 2215, my friend and I switched Pocket PCs for a day, and mine has many more programs on it, it still runs faster.

My 2 cents,

-Justin.

Wiggster
01-23-2004, 05:38 PM
If you're a manufacturer, you gotta print the specs as they really are, not as what they're comparable to. And many people will buy the higher processor speed even if it's slower :wink:

I've got a 1935 and my roommate got himself a 4155. There's a noticable speed difference when we're starting up Pocket Quake for a little wireless LAN action, and he gets slightly better latency than me (I blame the latency on my Sandisk SD WiFi while he has integrated), but they run at nearly equal speeds for most applications. Yea, his is faster, but mine's a better deal.

Kati Compton
01-23-2004, 05:41 PM
So if everyone claims that the 1940 can run just as fast as the 2215, how can they charge so much more for it? Sure, it may run a little faster than 266mhz because of the "architecture" - but I dont see it running 133mhz faster, otherwise they would have advertised it as 400mhz then wouldn't of they? There is not point, I might as well got the 1940!

However - I do disagree, the 1940 is slower than my 2215, my friend and I switched Pocket PCs for a day, and mine has many more programs on it, it still runs faster.
You may find there to be a speed difference - I haven't played with either.

But remember that MHz ratings should really only be compared within a specific processor design. Equating MHz across processor models isn't accurate (such as even PIII vs. P4, and they're both Intel) There is much MUCH more to what makes a given processor "fast" than simple clock speed. Otherwise, there'd be a lot of computer architecture researchers out of a job. ;)

Anyway - back to the beginning of your post. It's not a question of the 266MHz processor actually running faster than 400MHz, or the 400MHz processor running slower than 400MHz. It has to do with how many instructions that they can each execute in a given time, how well they cache memory, and a few other things that are even more involved.

So while the CLOCK speeds might be very different, the instruction THROUGHPUT could be very close. And it's throughput that users perceive as "speed".

Janak Parekh
01-23-2004, 05:44 PM
So if everyone claims that the 1940 can run just as fast as the 2215, how can they charge so much more for it?
Don't forget that the 1940 has only an SD slot, while the 2215 has a CF slot as well.

Not only that, there are other factors to speed. First off, as Kati mentions, different processors can't be compared MHz to MHz. Another good example of this is Athlon vs. P4. AMD invented the "+" notation because their processors are much more efficient at a given clock speed. Second, there is much more to a PDA than just a CPU, there's the graphics controller, memory, etc.

From what I've anecdotally heard, the 1940 is similar in performance to the 2215 for most everyday applications. You might notice some differences for specific setups, but I wouldn't use that as the only important metric for deciding which of the two to get.

--janak

Kati Compton
01-23-2004, 05:55 PM
So if everyone claims that the 1940 can run just as fast as the 2215, how can they charge so much more for it?
Don't forget that the 1940 has only an SD slot, while the 2215 has a CF slot as well.
Doesn't the 2215 have a much better battery life with the XScale?

Janak Parekh
01-23-2004, 06:02 PM
Doesn't the 2215 have a much better battery life with the XScale?
Hard to say, because one would also have to compare battery sizes. ;) I don't have enough data to say.

--janak

Kati Compton
01-23-2004, 06:26 PM
Doesn't the 2215 have a much better battery life with the XScale?
Hard to say, because one would also have to compare battery sizes. ;) I don't have enough data to say.
Same size - 900mAh. Compaq claims 12 hours for the 2215, 8 for the 1945. While I think those numbers themselves may be a bit exaggerated, if we can assume similar exaggeration for both, the 2215 does have a longer battery life, though I'm not sure *really* by how much. I wonder if real-use follows those percentages and the 2215 has 50% longer battery life than the 1945.

Kowalski
01-25-2004, 08:51 PM
If you're a manufacturer, you gotta print the specs as they really are, not as what they're comparable to. And many people will buy the higher processor speed even if it's slower
totally agree! the same thing happened when intel announced mobile pentium. for years they said the faster the clock speed the faster your computer but now they are downgrading their proccessors! what a pitty :mrgreen:

and about the battery runtimes: compare the axim x5 with a pxa250 and pxa255. you ll find the answer i think

Andy Whiteford
01-26-2004, 12:17 PM
I've gone from a h1940 to h4150 and testing both units together, I feel the h1940 'feels' a little bit more snappy although not by much. After a soft reset and say starting Wordbook 3 for example, the h1940 started this app up slightly quicker than the h4150. It also seems to close apps slightly quicker as well. Once in an application however, both devices feel about the same speed although I think CPU intensive activities such as doing a phonetic search on the Wordbook dictionary file is slightly quicker on the h4150.
All in all, there doesn't seem much to choose from although I have read a lot about emulated games such as SNES roms running faster on the PXA 255 devices such as the h4150 and this is a good indication of pure processing power.

I guess it's like comparing desktop Athlon XP against Pentium 4. Clock for Clock the Athlon is faster however certain types of processing are catered for better on the Pentium 4 such as Miultimedia even though both chips are based on the same x86 architecture.

It's fair to say both the Intel and Samsung chips are very competent performers.