The OFFICIAL... Neosd MicroSD Card thread...

[mastamuzz]

It's been acknowledge by the team and they have stated far too many times there is a part of the team focused on still improving things till there is nothing left to improve, they have never left people hanging and I personally have received great customer service from them and I bet GadgetUK too they do constant follow ups for "real Issues" and they don't stop till is resolved, so creating an issue right now is pointless if the user doesn't see it as an issue, as I said they know and they can work on it in due time, at the moment I believe they still have so many people to attend and give customer service cause soooo many people don't want to come here as a noob even though ROT has been doing an impeccable job to keep noobs in the loop even after being rude some of them.

lets keep in mind that our consoles are not brand new and even same revisions have different behavior, as some may recall I had the sync issue on a 3-6 I have two 3-6 units but the other was still open to be modded, guess what? after reassembled that 3-6 has no issue at all, and given my 3-3 had no issue I ruled that as a problem with the console not the cart but not everyone has different consoles to try. so we might see this as an issue with the cart but it might not be.

what I mean with this is that we should let them work on it, having testicular jumper cables attached wont help, the "other product" can come out today or 2025 who knows.

but I am starting to wonder if there is truth to Mitsu's and darksoft words........

[Rot]

but I am starting to wonder if there is truth to Mitsu's and darksoft words........

Please expand on this...

xROTx

[GadgetUK]

I didn't realise this flash time thing had become an area of concern! My first thoughts are there are two possibilities here:-

1) Tolerance difference in the flash ROMs, but 20% seems like a lot - but maybe that is within manufacture specs?
2) Speed of the erase process - I wonder where the ARM gets its clock from? Is it theoretically possible that a clock is too slow or missing some cycles for some reason.

With regards to programming differences:-


It is probably better to separate erase time vs programming time to work out where the bottlenecks are.

Raz, out of interest, have you tried your cart in another MVS to see if the timings are the same? If so, that would point to just tolerance differences between the flash ROMs perhaps. Unless there's some clock thing going on where on some boards the ARM is running faster than on other boards, but I cannot think how, unless the ARM takes its own clock from the MVS, and perhaps there's a difference there. I assume the ARM just "does its own thing" and erases the flash ROM in parallel to the MVS 68K, so the 68K isn't actually sending any erase commands to the board, other than perhaps an initial - "erase these blocks {range from : to}" etc. If the 68K is doing some of the leg work, then maybe the BIOS could have an effect - I am thinking about interrupts etc.

[Razoola]

Hi Gadget, yes, I have tried everything. There is an issue there and its defo related to the NeoSD, three different MVS systems were tried. Then there is also the results yoshi did on an AES with two NeoSDs.

That said neodev is looking at it and has an idea to try given he had some weird situations long ago. Hopefully I'll get a firmware to check at some point later today to test. The exact reason of what is happening is not known and might well be down to issues in some flash chips themself.

The strange thing here is in relation to the times. ATM at least it seems very strange that once the sdcard speed is taken out of the mix the times seem quite constant. The kof2003 erase time being around 37-40 sec vs 50-52 sec is the first pointer the chips are erasing at different speeds. One would normally expect to see times in the middle also and not only at either end, granted though, there are not many postings related to erase times yet. Thats why I still have my money on an issue thay is probably fixable in firmware. Of course I could be wrong.

I have my fingers crossed. btw, I also mentioned another avenue to neodev that may speed up flashing times for everyone. So even if this is not solved another speed increase may be on the cards in a newer firmware (may not be that massive but would be better than nothing).

[Razoola]

Forgot to ask, as a matter of interest, what is the timed amount of the 'erase' message you see with kof2003 before the 'flashing' message appears?

[GadgetUK]

27.14 secs to erase. That was after flashing kof2k3, when shock troopers was previously flashed.

EDIT: And from KoF2003 erased to Metal Slug 5, 23.9 seconds for the erase.

It's going to sound a bit crazy this, but has anyone measured temperature vs flash time? The room my MVS is running in is out door temperatures at the moment - ie. really cold, a few degrees above C. I do wonder if it will slow down a bit when warmed up to room temperature.

[Razoola]

27.14 secs to erase. That was after flashing kof2k3, when shock troopers was previously flashed.

EDIT: And from KoF2003 erased to Metal Slug 5, 23.9 seconds for the erase.

It's going to sound a bit crazy this, but has anyone measured temperature vs flash time? The room my MVS is running in is out door temperatures at the moment - ie. really cold, a few degrees above C. I do wonder if it will slow down a bit when warmed up to room temperature.

You need to flash kof2003, then flash it again and time that. I suspect it will then it will be around 37 seconds. As a matter of interest can you also time the total time (erase+flash)

I have tested with tempreture. In my case the warmer it is, the slower the speed, thats starting from about 18c roon tempreture.

Raz

[GadgetUK]

You need to flash kof2003, then flash it again and time that. I suspect it will then it will be around 37 seconds. As a matter of interest can you also time the total time (erase+flash)

I have tested with tempreture. In my case the warmer it is, the slower the speed, thats starting from about 18c roon tempreture.

Raz

A fraction over 27 secs to erase! It's freezing in here though!

EDIT: Total time 2 mins 48 secs.

[greatfunky]

About temperatures , i use my neosd mvs in the same condition as Gadget : cold room , near outdoor temp , however i use my neosd aes in the house (20°C) and i get a very very close timing on both neosd (more or less one or two seconds difference).

[GadgetUK]

About temperatures , i use my neosd mvs in the same condition as Gadget : cold room , near outdoor temp , however i use my neosd aes in the house (20°C) and i get a very very close timing on both neosd (more or less one or two seconds difference).

That's good to know, so sounds like we can rule out temperature. My guess is its just differences between chips. If I remember correctly, one of the things about nand flash is there's sort of ECC block and an error mapping system (hence how these chips can be flashed so many times, they are not only high quality but have a large amount of additional space to deal with errors. It's possible that because of that error management process that's why they vary in their timings. Because some chips could have a different distribution of errors than others, and that distribution pattern could affect timing when erasing and writing? I could be talking crap, but I've seen similar timing differences on Everdrives before now, although when you are comparing two programming operations that take 4 or 5 seconds the difference is a lot less noticeable.

[greatfunky]

That's good to know, so sounds like we can rule out temperature. My guess is its just differences between chips. If I remember correctly, one of the things about nand flash is there's sort of ECC block and an error mapping system (hence how these chips can be flashed so many times, they are not only high quality but have a large amount of additional space to deal with errors. It's possible that because of that error management process that's why they vary in their timings. Because some chips could have a different distribution of errors than others, and that distribution pattern could affect timing when erasing and writing? I could be talking crap, but I've seen similar timing differences on Everdrives before now, although when you are comparing two programming operations that take 4 or 5 seconds the difference is a lot less noticeable.

Yes , chips differences seems more likely , just think about overclocking on identical CPUs , some are more stable than others and could be overclocked at a higher speed , sometimes with huge differences.

[shaneboc]

@ neodev

Hi neodev,
just got my aes cart and wanted to thank you and your team again for the great product. I have had no issues with the cart and wanted once again to say thank you for such a great product and over the top fantastic service that you have provide the neo geo community.

[GadgetUK]

The other thing that supports it being the chips themselves - if you do the math, mine works out approximately 50% faster at erase, but only 8% faster on programming. So there is a difference in prog speed too, not just erase speed.

When all is said and done (and I know its easy for me to say because mine seems fast at the moment), when you are looking at 2 mins 48 vs 3 mins 25, its approx 20%. It's really not a big deal (well to me at least). I know that's easy for me to say, but what's an extra 30 seconds when you've waited 3 mins anyway. If you had known in advance that it could be + or - 20% would you be bothered, I know I wouldn't - I still would have bought one regardless.

I also think we should let neodev look into it more when he has time, maybe a firmware update could tweak this - maybe its not the problem we think it is, and maybe its something that can be tweaked in firmware, who knows. It's also worth considering that before we had the 'faster flash erase' update, it took longer for everyone. So if anything, we are all better off than at the start - if that makes sense? I am not trying to dismiss this, I am just trying to change perspective on it. Lets keep thinking about it, but not get overly stressed about it.

[yoshi41]

Some more testing with erase time...

AES UniBIOS 3.1 + NeoSD AES Firmware 1.05

Sandisk Ultra microSDHC UHS-1 (32GB, Class 10) SDSQUNC-032G-GN6MA

KOF 2003 - Erase - 0:38
KOF 2003 - Total - 3:22

Omega CMVS UniBIOS 3.3 + NeoSD MVS Firmware 1.05

Sandisk Ultra microSDHC UHS-1 (32GB, Class 10) SDSQUNC-032G-GN6MA

KOF 2003 - Erase - 0:32
KOF 2003 - Total - 2:49

PS: Both NesSD carts are totally great. Just interested to see if this can be fixed somehow.

[TMOSteel]

Question: the Neosd only supports up to SDHC card format, not SDXC, correct? SDXC default format is also exFAT, and I believe the Neosd only supports FAT32, is that right?

[Rot]

Some more testing with erase time...

AES UniBIOS 3.1 + NeoSD AES Firmware 1.05

Sandisk Ultra microSDHC UHS-1 (32GB, Class 10) SDSQUNC-032G-GN6MA

KOF 2003 - Erase - 0:38
KOF 2003 - Total - 3:22

Omega CMVS UniBIOS 3.3 + NeoSD MVS Firmware 1.05

Sandisk Ultra microSDHC UHS-1 (32GB, Class 10) SDSQUNC-032G-GN6MA

KOF 2003 - Erase - 0:32
KOF 2003 - Total - 2:49

PS: Both NesSD carts are totally great. Just interested to see if this can be fixed somehow.

I think neodev said... no...

There are restrictions with the AES side...

Important thing... does it work?... Answer... Yes...

Do you swap games every 30 minutes?... IF... the answer is Yes... then you're a tard...

xROTx

[Takumaji]

Do you swap games every 30 minutes?... IF... the answer is Yes... then you're a tard...

Damn straight.

[Razoola]

Yea, there are some differences with the AES compared to MVS but thats something seperate to this issue.

[Rot]

Yea, there are some differences with the AES compated to MVS but thats something seperate to this issue.

I really don't care about this... as long as the device works... a 20% difference means very little in the great scheme of things...

On the other hand if people came out of the woodwork and said the device wasn't working and the Neosd couldn't fixit... then i'd get interested...

I am here viewing the overall makeup of our forum... the forums sense of entitlement to make issues out of what i think as non-issues... is sorta grating on my nerves somewhat...

This is because... the device works... I DO NOT care if there is a variation of 20%... I DO NOT care if people think it's important because they swap games every 30 minutes...

The only thing I care about is the overall care of our community... not individuals... and this is why I am where I am..

xROTx

[Razoola]

Yes, this is a technical thing only. Its clear for you this is no biggie. Given this thread here is about sdcard / load times, its important the issue is discussed. Espicially given how it seems the sdcard is looking less likley the deciding factor governing load times and more about if one is lucky to have a neosd that is better proforming than the next off the product line. I'm hopeful through discussion about this situation and sharing of information and data, a solution can be more easily discovered.

The times posted by Gadget for example are intresting given his erase times appear faster than the spec.

[benjiedude]

Finally got around to timing my cart. I'm within the norm for AES as others have reported. I have a stock, no Unibios, US AES in the 13,000's. SanDisk class 10 16gb.

KOF 2003- 3:27
Magician Lord- 24.2 secs

[Rot]

Technical... Schmehnical.... So there's a variation... 20% as stated...

Does it affect the gameplay?... nope...
Does it affect your enjoyment of the games?... nope...

I remember when the AF 3 affair was happening... and certain guys were viewing that it as a technical masterpiece...

...When infact the bigger picture was far... FAR... more complicated...

I view this case as such... I may not have the tech brains but I don't view the forums as a tech hangout...

This point maybe important to some... but in reality in the great scheme of things... it's not even close to the top of whats truly important for this device...

Gameplay...
Hardware Compatability...
Ease of Firmware upgrade... etc etc...

THESE are some of the things that are important... not a measly 20% on a working device....

Geez... get a grip here...

xROTx

[Yodd]

When I was doing testing for my review video and the follow up firmware video, I noticed that flash and erase times were not consistent. I would flash the same games, in order, rotating through the list. Sometimes the flash times would be identical, other times vastly different. This was especially true when I used the expensive Sandisk card. With the cheap class 10 they were more consistent, though some times it varied.


So, in my experience the cheaper cards deliver a more consistent (and usually faster) flash time.

[massimiliano]

Rot, what's the point of this thread if not discussing timers? I mean, if trough Raz's analysis, we are gonna end with a clear picture about SD cards and factors influencing loading times, I guess it should be welcome..or not? (at least in this very thread, I 100% agree being anal about this i.e. in the general thread, would be just meh-sturbation)

[Razoola]

Yodd, yup the previous flashed game indeed plays a role in the flash time but that is not part of this issue. Its why its important to flash the game twice and only time the 2nd run. Its a weird situation.

It has only come to light becase neosd mentions a time of 2:42 to flash kof2003 and in my effort to hit the same time on the neoSD, I can only hit a time of 3:25 (various mvs tried). Thats using an sdcard neosd sent me which he verified flashed in 2:42 before he sent.

We are just trying to gather info to find the cause and hopefully if possible a fix that can be applied via firmware.