Pandora Vs Eee Pc Specifications

[ldesnogu]

I was objecting to the blanket statement that SGX530 is a better GPU. For the purpose of 3dmark, it is not "better" by the terms of the benchmarking tool which concluded that the SGX 535 was massively beaten by the GMA900 even with the SGX using a higher performing CPU, which is enough to defeat the sgx530 > gma950 assumption.

Do you have a link to sustain your claim about scores? Not that I doubt you, I just would like to see the context, since 3dmark is dependant on CPU performance on top of GPU performance.

Anyway 3dmark is just a benchmark, it should not be taken too seriously, in the same way the number of polygons, or the number of pixels your GPU can push every second As you wrote, getting q3 scores will be more interesting.

 

[chad78]

"Open system" may be meaningless, but "you can install and run whatever software you want without having to perform any software and/or hardware modifications, lose your warranty and be at the mercy of Sony" isn't.

Be at the mercy of Sony? To do what? And, no, you don't lose your warranty. If you need to send it back - just install the latest official firmware and send it in, and if you did cut open your only battery - time to go buy another one. (Which you can get online for less than $10 - shipping included - or hit up craigslist, or go to an actual store... Either way, you are out - at most - $25. Which, when added to the cost of the PSP, still comes way short of the cost of the Pandora.) So, no, you don't lose anything, and you aren't at anybody's mercy. What will Sony do? Sue me? It's not like it's a criminal act - at the very worst violating an EULA is a breach of contract. No court will convict me, either. I never signed anything. I never even clicked an "I accept these terms" button! I'm in no danger here. ATTENTION GOOGLE - I MODIFIED MY SONY PSP TO RUN THIRD PARTY FIRMWARE! SONY LAWYERS - PLEASE PM ME.

QUOTE

To me, this is a major thing - I don't own a PSP myself, so I don't have any experience with it, but tell me if this isn't substantially correct: To allow any PSP to run homebrew, regardless of software, you need a soldering iron, razor blade, and 7cm of wire. (I read this here on these boards at one point.)


You can do it that way, (although I don't know what the wire is for). Or you can borrow someone else's battery, or buy a $30 pre-made Pandora's battery online, or do a soft-mod.

QUOTE

Or, if you have a supported old version of the firmware, you can run any of a number of software downgraders.



This is a much more common method.

QUOTE

Either way, you lose any claim to support from Sony, as well as running the risk of physical damage in the former case.


Not true, see above, just install firmware and don't send in a Pandora's battery. In other words, unless you installed some sort of modding chip - which is a completely usesless and outdated method of modding a PSP at this point - the process is completely reversable. At worst, you invalidate the warranty on your battery if you cut it open and disconnect the one wire it takes to make the battery hackable.

QUOTE

Now it seems to me that regardless of how easy this procedure may be, the fact that it has to be performed means the comparison is stacked against the PSP from the start. It's not dishonest to cite this as one of the reasons to buy a Pandora over a PSP.



I don't have a problem saying "* - to play 3rd party applications, homebrew, emulators, etc. A PSP must be modded, which can involve soldering connections in the battery, downgrading firmware, or other software means." What I do have a problem with is making it sound like it is some horrible crime, some impossible task, or somehow more evil or dangerous than running emulators on a Pandora. And, again, if you (or whoever) doesn't feel safe desoldering a connection on a battery - DON'T DO IT THAT WAY. I didn't. I borrowed someone else's pre-modded battery. If I could not have found one, I would have either spent the $30 to get a premade one, or done more research on how to softmod.

 

[lubidog]

Chad, calm down! Is it so important? To win the argument?

 

[chad78]

Chad, calm down! Is it so important? To win the argument?

Winning the argument is not important. Accuracy and honesty are. And, yes, it is accurate and honest to say running emus on the PSP requires modding. It is not accurate or honest to say that it can't be done.

And it really irks me when people come in and say "This is pointless! You can't compare the Pandora to____." Yes, yes I can. I can compare the Pandora to a waffle iron if I want to.

What *REALLY* ticks me off is that it is for the benefit of those *EXACT SAME PEOPLE* that bitch about the comparison even daring to exist that this document was started. Look at the subheading to this thread. "Time to end the debate once and for all." You see that? The reason this spreadsheet exist is so that when those "idiot n00bs" show up asking the, (what is to anyone who doesn't live on these boards) obvious and sincere questions "What makes the Pandora better than or different from a PSP?" or "Why should I buy a Pandora? The Eee PC is better." Instead of having to rehash everything, scare away potential buyers, or even bother to roll your eyes in disgust - you can simply point them to the chart. Once the chart is finished, (or finished enough) I can easily see it being included in a new sticky thread... "The Pandora compared to other UMPCs and Gaming Devices, (PSP, Eee PC, DS, etc.)"

===

[EDIT TO ADD]

That said, I'll sit this thread out for a while.

 

[AireTamStorm]

Can't we just add a section to the comparison (preferabley the first item) wihc states "Able to run non-offical software without modification"

My vote goes for this. +1
I'm the author and I completely agree. Consider it added.

I don't know of an emulator that runs every SNES rom I have, its usually a split between ZSNES and SNES9x for me.

Off-topic, but try bsnes. It has high requirements, but you don't actually need a 2.4GHz Core 2 Duo to play a majority of games.


I hate 9x on Linux and I love ZSNES' interface. I'll give it a shot, thanks.

The fact remains, you cannot run homebrew software on a PSP out of the box. It is not an open system and you need to mod it to do so.

The fact remains that the PSP can, does, and will play emulators. And you need to pull your head out of the sand if you think that people will not cry foul if you compare the two devices and claim that Pandora can play emulators and the PSP can't. If you can't beat the PSP in a head-to-head competition, then either take it like a man or don't play the game. Compare apples to apples, here. If I am geeky enough to get a Pandora, I'm geeky enough to mod a PSP. For crying out loud - there's a whole section on this forum where people talk about different ways to mod the PSP, and different "Dev Kits" for the DS --- which is a nice legal way of saying "This cart lets me play roms and eh hem 'Back-Ups'."

Or - if you really want to compare to systems - be sure to add something like "Can play PSP games" "Can play PSP games natively" "Can play UMD movies" "Can play DS games" "Can play DS games natively" "Can play GBA games" "Can play GBA games natively". I mean, if you're talking about a non-modded PSP - that's still something it can do that the Pandora cannot and never will be able to do.


Yeah, but I'd be worried if a PSP couldn't play a PSP game natively. The fact is, you can't simply drag and drop a file on a memory stick, normally you have to go through all sorts of formatting procedures or modify a battery to get past the firmware. The software on the device fights you. The Pandora? No, GP2X? No. PSP? Yes. You cannot do it out of the box, so you should note that it requires modification of the firmware. Enough people here agree, including myself that there should be some note that the PSP is not an open system and should be shown as such. We will also note that there are currently no known planned commercial games coming out for the Pandora at this time. There you go, level playing field.

EDIT: Changes:
Added "Open Platform for Development" under the features and benefits section.
Changes PSP SNES Emulation status to Playable, But may not run some titles.
Added "Requires firmware hack" to the emulation section of PSP. How we word this to work with downloadable PS1 titles on the PSN store (which doesn't require a firmware mod), you figure out.

 

[Knux]

what a joke. you can't say the Eee PC CPU is only slightly faster. And where did you get the DMIPS value for the CPU?

Also why is the Pandora awarded the GPU prize? Sure, it's more advanced, but at maybe half the performance.
Sure, they can be compared, but this is an obvious distortion of a fanboy.

edit: you also got the polygon rate wrong again for the Pandora.

Are you saying that a GPU without even a hardware T&L, let alone pixel/vertex shaders (GMA900) better than one with UNIFIED SHADERS (PowerVR SGX)!!!?? And how else would you measure performace other than the polygon count and fillrate?? Benchmarks aren't accurate anyway since they're designed for mainstream GPU's and not TBDR's like the SGX. Oh and about the polygon count of the SGX, since it's a TBDR, as opposed to an IMR, it increases as scene complexity increases. So there is no definite polygon count, it can range anywhere from 5 million up to 14 million or even higher depending on overdraw. TI only made a rough estimate on the polygon count in terms of ordinary and usual circumstances. So there you have it, the GPU does vertex/pixel shading and leaves the CPU room for other instructions, while the GMA 900 has to rely on the CPU for even transformation and lighting, it'll probably cripple if it were to do vertex/pixel shading operations.


I was objecting to the blanket statement that SGX530 is a better GPU. For the purpose of 3dmark, it is not "better" by the terms of the benchmarking tool which concluded that the SGX 535 was massively beaten by the GMA900 even with the SGX using a higher performing CPU, which is enough to defeat the sgx530 > gma950 assumption. We'll have a better idea when optismised Quake3 benchmarks are compared as they both use OpenGL.

So, incase you missed it again. I'm not saying SGX in Pandora is lame, just that you can't close your eyes and make blanket statements on it's ability to own[/b]. A good example is the Chome9 HC, which has full support for DirectX9.0c and OpenGL 1.5 but will perform like a dog. No one would say that this is more advanced than the GMA900 just because it has features galore.

So now you want evidence? the stuff I said are basic facts that everyone should already know. I don't think any evidence should be presented to prove my points. Either way, i can point you to links to factsheets and whitepapers about the SGX line of GPU's if you want.

 

[hicksman]

That's great. I just finished reading another review of the EEE 901 and wondered if anyone thought to compare it to the Pandora. Good thread.

 

[icurafu]

what a joke. you can't say the Eee PC CPU is only slightly faster. And where did you get the DMIPS value for the CPU?

Also why is the Pandora awarded the GPU prize? Sure, it's more advanced, but at maybe half the performance.
Sure, they can be compared, but this is an obvious distortion of a fanboy.

edit: you also got the polygon rate wrong again for the Pandora.

Are you saying that a GPU without even a hardware T&L, let alone pixel/vertex shaders (GMA900) better than one with UNIFIED SHADERS (PowerVR SGX)!!!?? And how else would you measure performace other than the polygon count and fillrate?? Benchmarks aren't accurate anyway since they're designed for mainstream GPU's and not TBDR's like the SGX. Oh and about the polygon count of the SGX, since it's a TBDR, as opposed to an IMR, it increases as scene complexity increases. So there is no definite polygon count, it can range anywhere from 5 million up to 14 million or even higher depending on overdraw. TI only made a rough estimate on the polygon count in terms of ordinary and usual circumstances. So there you have it, the GPU does vertex/pixel shading and leaves the CPU room for other instructions, while the GMA 900 has to rely on the CPU for even transformation and lighting, it'll probably cripple if it were to do vertex/pixel shading operations.


I was objecting to the blanket statement that SGX530 is a better GPU. For the purpose of 3dmark, it is not "better" by the terms of the benchmarking tool which concluded that the SGX 535 was massively beaten by the GMA900 even with the SGX using a higher performing CPU, which is enough to defeat the sgx530 > gma950 assumption. We'll have a better idea when optismised Quake3 benchmarks are compared as they both use OpenGL.

So, incase you missed it again. I'm not saying SGX in Pandora is lame, just that you can't close your eyes and make blanket statements on it's ability to own[/b]. A good example is the Chome9 HC, which has full support for DirectX9.0c and OpenGL 1.5 but will perform like a dog. No one would say that this is more advanced than the GMA900 just because it has features galore.

So now you want evidence? the stuff I said are basic facts that everyone should already know. I don't think any evidence should be presented to prove my points. Either way, i can point you to links to factsheets and whitepapers about the SGX line of GPU's if you want.


I'm well aware of the SGX capabilities, and I don't disagree that it is more advanced. obviously you didn't read my post regarding performance.

 

[aranson]

I'm really getting sick of all the people saying "Comparing the Pandora to the ___ is stupid! It's not a ___ it's a ____." Well damn. I guess the Pandora is in a class by itself. It is incomparable! It's freakin' GOD!

I'm pretty sure that's how so many cars get the "highest safety rating in thier class". Just make up a new class where only you qualify, and you can say what you want.

Seriously, this is borderlining on argumentative masturbation. If you want to declare a PSP or Asus better for your needs, then do you really need to do it so violently and publicly?

 

[chad78]

I'm pretty sure that's how so many cars get the "highest safety rating in thier class". Just make up a new class where only you qualify, and you can say what you want.

That's precisely *my* point. Everyone keeps saying "You can't compare the Pandora to that!" - Well, what the hell are you going to compare it to? Can you only compare the Pandora to the GP2X and the GP32? Then, it absolutely wins every time. My point is that there are many devices that are similar enough to the Pandora to bear a fair comparison. These devices include the Nintendo DS, the Sony PSP, the Asus Eee PC, the Nokia N810, and others.

QUOTE

Seriously, this is borderlining on argumentative masturbation.



Are you saying that because I said "Let's add Can it play PSP Games, GBA games, DS games, etc.?" I was proving the point that you said. People wanted to make "Is it open?" a criteria. Well, then the jack off arguement applies. Of course only the Pandora is technically open. But if you going to add such a biased category, let's add all the biased categories - and that includes native commercial gaming, which the Pandora cannot do.

QUOTE

If you want to declare a PSP or Asus better for your needs, then do you really need to do it so violently and publicly?



I don't want to declare anything of the sort. Personally, I already have a Nokia N800 and a Sony PSP - but I still want a Pandora. My point is not to discredit the Pandora - quite the contrary - to declare the Pandora the winner - but in a fair fight.

However, the Pandora isn't for everyone, and having a full featured, and fair, comparison chart is a great tool to let people decide if the Pandora is right for them.

 

[gaterooze]

That's precisely *my* point. Everyone keeps saying "You can't compare the Pandora to that!" - Well, what the hell are you going to compare it to? Can you only compare the Pandora to the GP2X and the GP32? Then, it absolutely wins every time. My point is that there are many devices that are similar enough to the Pandora to bear a fair comparison. These devices include the Nintendo DS, the Sony PSP, the Asus Eee PC, the Nokia N810, and others.

You can compare anything to anything, obviously. What I meant by these sorts of comparisons not being terribly useful is that with some devices there is often one prime element that defies comparison and which renders all other specifics moot.

For example, the ability to play PSP games. If you want to play PSP games, then the comparison with a Pandora is useless. Another example is the ability to stick it in your pocket. This rules out an EEE, so comparing the CPU or RAM is meaningless. These should be common sense things that you don't need a chart to tell you. On the flip side, the number of polygons a Pandora can push is pretty meaningless information if there are no games on it to take advantage of that power.

It's all well and good making charts, and more power to you guys, extra information is always excellent to have at hand, but please don't be so upset if other people realize the only true basis of comparison is: "Will I get more fun/use from this?" And only the individual can answer that -- to further emphasise this, consider that to you a PSP = emulators, but as you can see from other people who have posted, they couldn't be arsed messing around modding one, so it doesn't have that level of utility to them. You see what I'm saying? Some thing just don't have black & white criteria to definitively compare.

 

[aranson]

That's precisely *my* point. Everyone keeps saying "You can't compare the Pandora to that!" - Well, what the hell are you going to compare it to? Can you only compare the Pandora to the GP2X and the GP32? Then, it absolutely wins every time. My point is that there are many devices that are similar enough to the Pandora to bear a fair comparison. These devices include the Nintendo DS, the Sony PSP, the Asus Eee PC, the Nokia N810, and others.

You can compare anything to anything, obviously. What I meant by these sorts of comparisons not being terribly useful is that with some devices there is often one prime element that defies comparison and which renders all other specifics moot.

For example, the ability to play PSP games. If you want to play PSP games, then the comparison with a Pandora is useless. Another example is the ability to stick it in your pocket. This rules out an EEE, so comparing the CPU or RAM is meaningless. These should be common sense things that you don't need a chart to tell you. On the flip side, the number of polygons a Pandora can push is pretty meaningless information if there are no games on it to take advantage of that power.

It's all well and good making charts, and more power to you guys, extra information is always excellent to have at hand, but please don't be so upset if other people realize the only true basis of comparison is: "Will I get more fun/use from this?" And only the individual can answer that -- to further emphasise this, consider that to you a PSP = emulators, but as you can see from other people who have posted, they couldn't be arsed messing around modding one, so it doesn't have that level of utility to them. You see what I'm saying? Some thing just don't have black & white criteria to definitively compare.



Exactly! Hoo-rah!

 

[chad78]

You know what? I see your point. How about something a lot less detailed, but more to the point. I still think there are reasons to give people comparative info about the systems, so how about this?

http://spreadsheets.google.com/pub?key=ptN...hTQtPpIhwyqmAOA

[EDIT] BTW, I didn't check any of the stats, so the storage, for example, could be off by a lot, except for the Eee PC, which I am 99% sure about. [/EDIT]

 

[gaterooze]

That seems like a pretty sensible summary, all told.

 

[aranson]

Excellent comprimise.

 

[AireTamStorm]

You know what? I see your point. How about something a lot less detailed, but more to the point. I still think there are reasons to give people comparative info about the systems, so how about this?

http://spreadsheets.google.com/pub?key=ptN...hTQtPpIhwyqmAOA

[EDIT] BTW, I didn't check any of the stats, so the storage, for example, could be off by a lot, except for the Eee PC, which I am 99% sure about. [/EDIT]



Added to the beginning of the thread. Consider merging this into the top of the existing document, so that you can see simple information, and scroll down for more complex information. Awesome job.

 

[Knux]

what a joke. you can't say the Eee PC CPU is only slightly faster. And where did you get the DMIPS value for the CPU?

Also why is the Pandora awarded the GPU prize? Sure, it's more advanced, but at maybe half the performance.
Sure, they can be compared, but this is an obvious distortion of a fanboy.

edit: you also got the polygon rate wrong again for the Pandora.

Are you saying that a GPU without even a hardware T&L, let alone pixel/vertex shaders (GMA900) better than one with UNIFIED SHADERS (PowerVR SGX)!!!?? And how else would you measure performace other than the polygon count and fillrate?? Benchmarks aren't accurate anyway since they're designed for mainstream GPU's and not TBDR's like the SGX. Oh and about the polygon count of the SGX, since it's a TBDR, as opposed to an IMR, it increases as scene complexity increases. So there is no definite polygon count, it can range anywhere from 5 million up to 14 million or even higher depending on overdraw. TI only made a rough estimate on the polygon count in terms of ordinary and usual circumstances. So there you have it, the GPU does vertex/pixel shading and leaves the CPU room for other instructions, while the GMA 900 has to rely on the CPU for even transformation and lighting, it'll probably cripple if it were to do vertex/pixel shading operations.


I was objecting to the blanket statement that SGX530 is a better GPU. For the purpose of 3dmark, it is not "better" by the terms of the benchmarking tool which concluded that the SGX 535 was massively beaten by the GMA900 even with the SGX using a higher performing CPU, which is enough to defeat the sgx530 > gma950 assumption. We'll have a better idea when optismised Quake3 benchmarks are compared as they both use OpenGL.

So, incase you missed it again. I'm not saying SGX in Pandora is lame, just that you can't close your eyes and make blanket statements on it's ability to own[/b]. A good example is the Chome9 HC, which has full support for DirectX9.0c and OpenGL 1.5 but will perform like a dog. No one would say that this is more advanced than the GMA900 just because it has features galore.

So now you want evidence? the stuff I said are basic facts that everyone should already know. I don't think any evidence should be presented to prove my points. Either way, i can point you to links to factsheets and whitepapers about the SGX line of GPU's if you want.


I'm well aware of the SGX capabilities, and I don't disagree that it is more advanced. obviously you didn't read my post regarding performance.

Yes in fact I did read that. But once again, we really don't know how the SGX compares to the GMA or any GPU for that matter in terms of performance. Benchmarks can't be trusted all the time, and btw, it may well be that TI optimized the SGX 530 to perform even faster than the SGX535 that intel is using, that again, we don't know. Yes, so you are right that the SGX 530 is more advanced than the GMA 900, but about the SGX being slower in performance than the GMA 900, i can't comment on that yet.

 

[ldesnogu]

I'm well aware of the SGX capabilities, and I don't disagree that it is more advanced. obviously you didn't read my post regarding performance.

I am sorry, but your post regarding performance is useless without links providing information about how the numbers you quote were obtained.

So where did you get this info from ?

I do know that a 3dmark test between a menlow iwht SGX 535 and GMA950 @133 on a McCaslin, show the GMA950 with a significant polygon per seond icnrease.

 

[AireTamStorm]

I'm well aware of the SGX capabilities, and I don't disagree that it is more advanced. obviously you didn't read my post regarding performance.

I am sorry, but your post regarding performance is useless without links providing information about how the numbers you quote were obtained.

So where did you get this info from ?

I do know that a 3dmark test between a menlow iwht SGX 535 and GMA950 @133 on a McCaslin, show the GMA950 with a significant polygon per seond icnrease.

If you read the actual Document this thread is about, you do see that the GMA has the peak theoretical of the SGX outspec'd.
I have both theoretical and synthetic benckmarks. If you start using OpenGL 2.0 instructions (which are done via software instead of on the GMA900 itself) the SGX will likely outperform the GMA.
(not that many applications will make heavy use of those instructions over OpenGL 1.3)

See that here:

QUOTE

Intel GMA 900 OpenGL Compatibility
http://www.intel.com/products/chipsets/gma900/index.htm (AireTamStrm)
More Intel GMA Info, Pixel Fillrate and Shader Info
http://en.wikipedia.org/wiki/Intel_GMA (AireTamStrm)
Confirming the 1333MPixel/s fillrate via synthetic benchmark
http://www.xbitlabs.com/articles/chipsets/...ay/i915g_4.html (AireTamStrm)
Intel GMA 900 Raw Rendering Power, synthetic benchmark.
http://www.xbitlabs.com/articles/chipsets/...ay/i915g_5.html (AireTamStrm)
PowerVR SGX Raw Rendering Power, peak theoretical
http://focus.ti.com/lit/ds/symlink/omap3530.pdf (Icurafu)

 

[sindbad]

I'm sorry, but I fail to see the point of this GPU comparison. They're both powerfull enough and the SGX is very powerfull for a handheld. Playing games on the eeepc is unlikely, as it isn't made for that.