the cereal box comment was for mussels not for u newtekie. I know the dicussion was about the I-ram I just didn't agree with the A64 buggy comment, and the huge performance hit from 2t. it wasn't to all u rest of u wankers who decided to jump into help this guy. when i'm pretty sure he can speak for himself.

And ya I know less than everyone here

i've only been doing this computer thing since I was 12. my first pc was a 80286, first pc I ever build was a 80386. and i'm 25 now.

and please show me in my post when I specifically called someone a newb.

Makavelijimmy he said a 2500+ and a x1900XTK . Guess what there is no 2500+ a64, therefore he was implying a Athlon XP 2500+ which is socket A, there for a AGP motherboard. jesus christ are u guys all blind.

Again, did you ever think that maybe he just made a typo? No one is perfect, and 2 and 3 are right next to eachother on the keyboard, unlike X and K...x1900XTK? And you are talking to him about setups that don't exists, get a damn life.

Besides all that he was making a direct comment about how 1T and 2T command rates to make a difference, and how noticeable it is depends on a lot of factors, and then gave an example of when 1T and 2T command rates would make a very noticeable difference. No tell me do 1T and 2T command rates even exist on Athlon XP systems, since you probably don't actually know the answer I will tell you. No they don't exist, so again, what is more likely? He just made a mistake and typed 2500 instead of 3500, or he really meant 2500 and meant to setup an example that neither applies nor could prossibly exist.

Makavelithe cereal box comment was for mussels not for u newtekie. I know the dicussion was about the I-ram I just didn't agree with the A64 buggy comment, and the huge performance hit from 2t. it wasn't to all u rest of u wankers who decided to jump into help this guy. when i'm pretty sure he can speak for himself.

And ya I know less than everyone here

i've only been doing this computer thing since I was 12. my first pc was a 80286, first pc I ever build was a 80386. and i'm 25 now.

Bull, it wasn't directed at me, but you start of by talking about people posting nonsense, then qouted me, responded to the quote and then asked if I learned my computer knowledge from a cereal box. Right, but it wasn't directed toward me. Nice try jackass, but I really think you should STFU because you are an idiot.

Also how does any your "My first computer was a 286 I have been at computers since I was 12" shit apply here? I don't really care when you built your first computer or what your first PC was, none of that applies to the subjects at hand which you obviosly know nothing about. So far you have said that mussels lyed when you said that 2T can cause a 30% Memory Write hit and that, from what I can only assume is your "experience", it isn't anywhere near 30%, which he then provided proof that it does in fact cause a 30% memory write hit. Strike one for you. Then you said that more then 2GB isn't buggy on A64 rigs, which I then explained that it was in fact buggy and even unstable. Strike two for you.

You are zero for two in the knowing what you are talking about column...so maybe it is you who learned your computer knowledge from the back of a cereal box.

Makaveliand mussels I edit my post for spelling and grammer nothing else. so stop implying bs

But yet you still left obvious typos and gammer mistakes in it...odd.

Makaveliand please show me in my post when I specifically called someone a newb.

Well, with everyone having the ability to edit posts that is kind of hard. I am not saying that you did in fact edit out when you called him a n00b, I didn't see it so I can't even confirm that you originally did. Anyone can edit their posts to say something completely different then what the originally said, so asking something like that simply doesn't work on a forum like this, especially when you have actually gone back and editted your posts.

Besides, implying that the other person is a n00b by jumping all over him when it is pretty obvious to the rest of us that he made a typo and freaking out about how his setup would never even exist is close enough for me to say you were calling him a n00b. You didn't directly come out and say it, but you sure though it otherwise you wouldn't have responded to it and bashed him.

he said it was a 30% peformance it writing to memory. And showed me a everest benchmark. and I said show me a real world applications taking that same memory hit. and he has yet to show me that.

and yes there were still some errors in the post, when I edited it, but I left them as is. I did most of the reply's from work, when I actually was doing something abit more important at the time. I gave the example of when I started cause one of your guys post, keeps implying I don't know what i'm talking about. when i'm been doing this for well over 10 years

and if he doesn't make a correction how am I suppose to know he is talking about a 3500+ your assuming it was typo, until he states that what he wrote stays.

what u presume is close enough to me to me calling him a noob and me actually doing it is 2 different things. I don't give a damn how u thought I wrote. I never called him noob and my post show that. So maybe you should stop putting words in my mouth

Makavelihe said it was a 30% peformance it writing to memory. And showed me a everest benchmark. and I said show me a real world applications taking that same memory hit. and he has yet to show me that.

Now who is the one with a selective reading problem? He said "can be as much as 30% on memory write speeds", there is no way to translate that into real world performance, there is no way to show that 2T causes a 30% memory write speed decrease with anything other then a synthtic benchmark. I can't think of any real world programs that let you monitor your memory write speeds...can you? So how exactly is he supposed to show that there is a 30% memory write speed decrease with 2T without synthetic benchmarks?

Makaveliand yes there were still some errors in the post, when I edited it, but I left them as is. I did most of the reply's from work, when I actually was doing something abit more important at the time. I gave the example of when I started cause one of your guys post, keeps implying I don't know what i'm talking about. when i'm been doing this for well over 10 years

Again, how long you have been working with computers has no impact on this discussion. You are still zero for two in the having a clue column...

Makaveliand if he doesn't make a correction how am I suppose to know he is talking about a 3500+ your assuming it was typo, until he states that what he wrote stays.

Simple, you use logic combined with a little computer knowledge, I'll assume you do have some, even if it is very limitted which I am guessing it is.

1.) He was making a direct comment about how 2T affects performance in A64s.
2.) Then he made an example with a 2500+, which if that is really what he meant would not even have a 2T command rate issue because there was no 1T and 2T with Athlon XPs.
3.) The numbers 2 and 3 are right next to eachother on the keyboard.

Hmmm....maybe it was a typo...it might be possible, I mean no one is perfect.

Edit: Oh and Makaveli, since I know you actually don't know this due to your limitted computer knowedge. It is entirely possible to run a x1900XTX with a 2500+. So you are in fact completely wrong on that issue, even if it wasn't a typo.

Makaveliwhat u presume is close enough to me to me calling him a noob and me actually doing it is 2 different things. I don't give a damn how u thought I wrote. I never called him noob and my post show that. So maybe you should stop putting words in my mouth

Did I ever say you called him a n00b? No, I simply said you implied it, which you did. So maybe you are the one that should stop putting words in peoples mouths.

And for Christ's sake are you that starved for attention that you have to make a post for every sentence you write?

Makavelijimmy he said a 2500+ and a x1900XTK . Guess what there is no 2500+ a64, therefore he was implying a Athlon XP 2500+ which is socket A, there for a AGP motherboard. jesus christ are u guys all blind.

WRONG!!!! He said a 2500+... hold on second... I do believe that you can get a Sempron socket 754 CPU can you not? And oh! Look! PLENTY of mobos with socket 754 and PCI-E!!! Now lets put that together: Sempron 2500+ socket 754 and a PCI-E Motherboard means you CAN have an x1900XTX!! Someone is talking out his arse!!! So why don't you actually take a look at what you're saying before posting it idiot. And stop having a go at everyone on these forums. :slap:

no i'm not starved for attention, but I am enjoying this. you guys were the ones that started the name calling. not that it matters to me what u call me.

but I enjoyed the feedback from mussles, just not from josie and the pussycats!

Makavelino i'm not starved for attention, but I am enjoying this.

Nice to see you don't mind being wrong then. :toast:

everyone can be wrong at one point or another. more so in a forum than other ways, it just matters how u play with the words.

the hostility in these forums was the funniest part for me, not to mention the gang banging

Makavelieveryone can be wrong at one point or another. more so in a forum than other ways, it just matters how u play with the words.

the hostility in these forums was the funniest part for me, not to mention the gang banging

Sounds like you were after a different forum if you enjoyed the gang banging.

W1zzardgigabyte just told me it does not use sata-ii. the problem is that the xilinx chip can not run that fast. this is because gigabyte has to engineer a memory controller like in a chipset, without previous experience (like intel, amd or ati) this is very hard to do .. 2gb modules do not work yet either but its definitely planned

Overall, this sounds like a definite improvement over the unit I use (CENATEK RocketDrive) in that it can bootup an OS from itself, that it uses a faster memory type (DDR in init. version & DDR2 in this newer one), & also a faster bus type (SATA vs. PCI 2.2).

The ONLY QUESTION I have about it, is does it maintain state between reboots? I would guess YES, via somekind of CMOS battery, etc. (where the unit I use has an independent powersupply, external hookup).

Anyhow - It's making me want to put the rocketdrive back into my first rig (which is relegated to pure server work here, running IIS 6.x & SQLServer 2005, for work & lab purposes @ home), & replace it in my latest rig (see signature).

:)

I do know 1 thing, from real-world practice & experience using these things (pagefile.sys on first partition here, & webpage cache, %temp% ops location (apps & OS), logging (apps & OS), print spool location, %comspec% location, & even running SETI from it on 2nd partition) that things DO speedup... I tested this out (not benchmarks, just "feel" performance of the system w/out the solid state disk in place) here, & there was a diff.

Just not hearing my disks "grind" paging/logging/webpage caching/temp ops etc. was proof enough (as we ALL know that memory is 1000's of times faster than HDD's are, & just common-sense if applied properly), as well as just getting crisper overall response from my system using a slower model of solid-state disk here for various things!

FOR REAL WORLD TESTS THOUGH? Well, YEARS AGO (circa 2003) I DID A REVIEW FOR CENATEK, "An independent users review" on their front page of their site, & it noted many things improving (synthetic tests, to ones with WinZip etc.).

* Things in this field just get BETTER, all the time, & it makes you want to burn dollars! That's for sure... oh well, that is good too (albeit not on a personal finance level, lol), because it keeps the economy running!

Makavelithe hostility in these forums was the funniest part for me, not to mention the gang banging

I don't know you @ all, & you may be a heck of a nice guy...

HOWEVER, you DID startup with the "cereal box" thing (lol, which was funny, but it got the people you ribbed on, on your case now)... but, the point is, you DID kick up the mess iirc from my skim of this post.

Fact is? I think these forums are pretty cool, & folks are less prone to startups of fights... @ least so far, & I am new here myself.

Heh, & I may be the LAST person who should say "don't flame war" here, lol, because I have been known to be quite "notorious" for it myself!

APK

P.S.=> Guys! Enough with the debating/name tossing already... both sides have points imo.

Makavelli does in 1 regard - that nothing "real world application-wise" shows bennies from DDR2/AM2 iirc, @ least not yet (other than raw memory benchmarks tests/synthetics)... & the others did catch him on various points via exceptions. Anyhow, it dragged this way off-topic & started up a "flame war"...

Interesting read for a flame-war though, many interesting points were made on the hardware level & enlightened me in fact on many fine points (I just skimmed, so cut me some slack if I messed up on some points I noted above), so I guess it wasn't ALL that bad then, eh?

(Compared to many of you on the hardware-side? I am a relative "noob" when it comes to current/state-of-the-art equipment, so I can gain by your discussions... even when they are name-tossing flamewar ones!) apk

Jimmy 2004Sounds like you were after a different forum if you enjoyed the gang banging.

lol point and case I said it was funny, not that I was looking for it :rockout:

Alec§taarOverall, this sounds like a definite improvement over the unit I use (CENATEK RocketDrive) in that it can bootup an OS from itself, that it uses a faster memory type (DDR in init. version & DDR2 in this newer one), & also a faster bus type (SATA vs. PCI 2.2).

I didn't realise Sata was faster than PCI :cool: . But anyway, surely (like I think someone said earlier on) DDR ram is faster than SATA II anyway so DDRII won't give all of the benefit.

Jimmy 2004I didn't realise Sata was faster than PCI :cool: . But anyway, surely (like I think someone said earlier on) DDR ram is faster than SATA II anyway so DDRII won't give all of the benefit.

Well, I think so (on the bus speeds part & you are right-on as far as DDR/DDR2 speeds, @ least as far as AM2 chipsets by AMD have shown, thusfar @ least)!

Anyhow: on bus speeds -

PCI 2.2 (which my CENATEK rocketdrive & the HyperOS III use as slot bus insertion to mobo types) does what? 133mb/sec??

SATA does what?? 150mb/sec???

(So, iirc, & I have my "facts" straight - even SATA 1 blows it out by 17mb/sec transferral rates, albeit, those may be BURST rates).

Now, SATA II does 300mb/sec. (& according to Wiz, who spoke to them & I quoted it, SATA II's not in use by Gigabyte for this IRAM solid-state disk (yet))!

* Still, remember: The rates we're citing? They are purely theoretical... rarely equating to that in practice.

In any event - it has my 'itching' to invest in one of these... they DO sound great!

NOW, What I would like to see, is a PCI-e implementation of these, especially via the x1 & x4 sockets, as they have a faster transferral rate than PCI 2.2 for sure, & possibly, SATA!

I have an open PCI-Express x1 slot on my mobo now, & it would be NEAT to get a solid-state disk that took advantage of it & its transfer rates, providing they are superior to SATA 1...

APK

P.S.=> Correct me where I am "off" guys, because I can stand to gain via correction as much as the next guy can (especially in hardware know-how etc.)... apk

Makavelino i'm not starved for attention, but I am enjoying this. you guys were the ones that started the name calling. not that it matters to me what u call me.

but I enjoyed the feedback from mussles, just not from josie and the pussycats!

Nice try, but you were actually the first person to start insult others in this thread.

And you enjoy being proven wrong every step of the way? What did you read the "Getting Started" manual that came with your Dell and now you think you are some kind of computer expert?

No makaveli, you are. He said "if you were to pair up something like" not 'if i stuck an XTX in my old AGP system'

He did NOT directly state that it was AGP. you assumed that. He did not even mention it was a remotely feasible system, it was an example of an old CPU and a high powered video card. YOU are blind, because like everything in this thread, YOU have missed the entire point.

Stop nagging over minor issues and pay attention to what people are saying, not ways you can harass them over the mose pointless things.

Practical applications & results of Solid-State RamDisks

See subject line, & this URL:

www.avatar.demon.nl/cenatek.html

:)

* That's a review I did for CENATEK in 2002-2003 showing results & gains you get by using solid-state diskdrives, & for home-use purposes!

(Another one I did for SuperSpeed.com/EEC systems years before it, lends itself MORE to how you can use these things in industrial environs/server purposes, especially for database work - iirc, the link for that is also on that page as well)... but, if it's outta date, moved, etc.? Here tis:

www.superspeed.com/desktop/faq.php

A decent read imo & possibly one that may help you decide on whether or not you could use one of these units (which have only gotten better imo, than the 1 I have here...)

APK

while dated, that info is a good indication that just speeding the page file alone is a good boost to everyday apps.

2-4GB of I-RAM pagefile goodness seems pretty appealing, as XP doesnt care if it has to re-create the page file on each boot.

Have cenatek released any products like that lately? It's not a name i've heard here in aus.

Musselswhile dated, that info is a good indication that just speeding the page file alone is a good boost to everyday apps.

Dated doesn't matter - it's ALL principals, they never age... you gain no matter what really, and with any type of solid-state drive... It all really depends on how you apply them.

(However, this newer type using faster bus types than the PCI 2.2 model I have, faster RAM types, & also just all in all better/newer technologies will only do you right, even moreso, imo! Especially since you can bootup from it...)

Mussels2-4GB of I-RAM pagefile goodness seems pretty appealing, as XP doesnt care if it has to re-create the page file on each boot.

True, but if you see my signature below, or my earlier posts? There is a great deal MORE you can do & move to it, to get even better performance out of your system using these units!

MusselsHave cenatek released any products like that lately? It's not a name i've heard here in aus.

Not that I am aware of, which is WHY I am very interested in getting one of these units, and then placing this one back into my older system.

APK

I'm not denying anything you've said, i agree with it. I just prefer a way that wont break my system should something go wrong with the drive :P

if its just a pagefile, everyone can do that and theres no issue. Browser cache idea i like too. but windows temp i'm not so keen on, in case files went missing that were in use/needed.

(i've emptied temp files before and broken some rather badly coded apps)

i want 10GB ramdisk drives already :D

MusselsI'm not denying anything you've said, i agree with it.

Cool, because it all works (very well I have to add). There are "catches" though, like what I mention below (UPS)... IF you demand that level of "uptime" & keeping state on them, that is.

Most of what I do on mine, if you note HOW I use it? Doesn't demand that...

Only in DB work would I start to worry... but, that's where DB backup/mirror is something most DBA's practice.

MusselsI just prefer a way that wont break my system should something go wrong with the drive :P

Nothing really ever does here @ least... there's no moving parts & they tend to last (e.g.-> Mine's been going strong for 4 years now).

Having a working, reliable, backup UPS helps a LOT though, if you are doing work on them that demands keeping state.

Musselsif its just a pagefile, everyone can do that and theres no issue. Browser cache idea i like too. but windows temp i'm not so keen on, in case files went missing that were in use/needed.

That's what the UPS is for, but temp ops are just that: TEMPORARY.

(You can always redo an operation if needed).

Mussels(i've emptied temp files before and broken some rather badly coded apps)

That'd be the app's design downfall then, imo @ least, & it seems, yours also.

Personally, I've never had a problem in that capacity here, & have been at this stuff for quite a while now, with a pretty fairly wide range of apps (relative term, this varies person-to-person in this field of course).

Power outages CAN floor you though, but, they can do the same to an "ordinary setup" too!

However, please note, that if you do the things I do (mostly folders oriented), you create a batchfile to immediately recreate that folder structure & poof - you're "back-in-business".

That, or be sure to have a UPS!

Once you've formatted them, that is... & that? Heh, only takes SECONDS, literally.

(Try that with a std. HDD! So do defrags... unreal speed!)

ON THOSE "BADLY CODED APPS":

Can you recall what apps those were?? I'm curious on this note, because I will either avoid them, or write their author with a potential fix in his/her design.

Musselsi want 10GB ramdisk drives already :D

This one's close, iirc, the IRAM current model can do like 8gb, iirc, from the reads on it (possibly more if "spanned or striped").

The CENATEK unit I have, if striped for example, can go up to 16gb, but the cost of this is prohibitive (needs 4 of these boards & the RAM too - NOT CHEAP!).

APK

Heh, on my wondering if there is going to be a PCI-Express implementation of these types of cards (solid-state ramdisk drives) here earlier?

Well, I got a "PM" in my UserCP inbox some of you may wish to be made aware of!

Very cool, check it (along with my reply to him):

DDRdriveAlec§taar,

I read your post concerning the Gigabyte i-RAM, and wanted to make you aware of a PCI Express based solution that will be released later this year.

DDRdrive X1 Pre-Release Specifications:

* PCI Express x1 based plug-in card.
* 8GB capacity / 4 DDR 184 pin DIMMs (Bootable).
* Custom and upgradeable high performance Memory Controller.
* External power jack with an included switching AC adapter.
* Industry standard hard drive activity LED connector.

More details will be released in due course.

Best regards,
Chris
www.ddrdrive.com

WoW... NIFTY! Thanks for that info... bigtime!

DDRdrive*** I remember your review of the RocketDrive!

:)

Ha, thanks for that!

APK

P.S.=> Is that "smokin" or what? Folks into this arena are on it as far as PCI-e usage, already, & watching user concerns on forums like these, which is cool imo!

Plus, they are also working on JUST what I need too!

That is because I have an open PCI-e 1x slot left on this mobo (& iirc, PCI-e 1x is FAR better "bandwidth" than PCI 2.2 is, & possibly more than this latest SATA one from Gigabyte in their IRAM GC-Ramdisk as well, but not sure on THAT latter account here though - I'd have to double-check!):

I did just check, & oddly, from my other post about Gigabyte's products, here are the bandwidth specs of each current bus type (NOTE THE BOLDED ONES - for comparison's sake (SATA/SATA II vs. PCI-e 1x slot mounted type noted by the person who wrote me in PM)):

--------------------------------------------------------------------------------

Common Buses and their Max Bandwidth

PCI 132 MB/s
AGP 8X 2,100 MB/s
PCI Express 1x 250 [500]* MB/s
PCI Express 2x 500 [1000]* MB/s
PCI Express 4x 1000 [2000]* MB/s
PCI Express 8x 2000 [4000]* MB/s
PCI Express 16x 4000 [8000]* MB/s
PCI Express 32x 8000 [16000]* MB/s
IDE (ATA100) 100 MB/s
IDE (ATA133) 133 MB/s
SATA 1 150 MB/s
SATA 2 300 MB/s
Gigabit Ethernet 125 MB/s
IEEE1394B [firewire] 100 MB/s

PCI Express is a serial based technology (transmitting data in packets over the first four layers of the OSI model in fact afaik & even does packet retranmission retries if any are dropped - couple that with SATA 1/2 crc32 checks (iirc, it does that) & you get better reliability as well & even packet reprioritization (QoS) so that streaming media like video get priority over other data types for smoother/faster processing).

Thus, data can be sent over the bus in two directions at once.

Normal PCI is Parallel, and as such all data goes in one direction around the loop.

A 1x lane in PCI Express transmits in both directions at once serially, simultaneously.

In the table above, the first number is the bandwidth in one direction and the second number is the combined bandwidth in both directions.

Note that in PCI Express bandwidth is not shared the same way as in PCI, so there is less congestion on the bus.

--------------------------------------------------------------------------------

Convenient & cool for me @ least, possibly for you others as well w/ PCI-e capable mobos... because this SMOKES the SATA/SATA II bandwidth possibilities even! apk