NForce2 / KT400 questions

[airspirit]

Other than a waterblock, since I didn't design my current system to account for the extra draw from the chipset. I only have 1/2" (then to 3/8") flowing to that particular PC. Splitting it off to a chipset would be hazardous to the CPU.

[airspirit]

I strapped a fan to a jury-rigged frame, blowing on the current NB sink, and I'm able to clock to the mid 1700s. When I put the Vantec on it, I wonder what I'll be capable of then?

WITH THE NFORCE2 CHIPSET YOU NEED EXTRA COOLING ON THE NB IN ORDER FOR IT TO FUNCTION CORRECTLY. TAKE THIS INTO ACCOUNT WHEN LOOKING AT A NEW BOARD. IF IT DOESN'T HAVE A GOOD HSF (NON-PASSIVE) ON IT, EXPECT TO BUY A HSF AS WELL!!!

[jtroutma]

Simple enough for me, I have 3 Blorbs sitting around doing absoluety nothing

[gmat]

Already got a chipset waterblock anyway...
Does the southbridge get hot at all btw ? That way I could recycle that NB heatsink...

[airspirit]

No.

[bigben2k]

Quote, from the December 12 OC news page:

"Epox 8RDA+ I have a couple copies of emails sent to me from Epox technical support saying that the board 1) doesn't have a PCI lock and 2) has a maximum PCI divisor of /5. So that's the big secret."

Click here

[bigben2k]

Here's yet another review of the A7N8X (in french no less !).

Note the test results on memory latency of NForce2 versus KT400.

[airspirit]

My board became infinitely more stable when the Vantec Copper Iceberg cooler was installed. I thermal taped a copper heatsink onto the southbridge for good measure, as well, though this wasn't a problem ... I just was feeling froggy and had an extra one laying around.

I am currently running my AGOGA/XMS2700 512MB at 10x188 at Cas2-2-2-5 at 1.9V/2.9V with no stability issues.

A new BIOS was released today allowing FSB speeds up to 250. This release and the 2C06 releases make the board infinitely more stable, with many of the no POST errors resolved.

I'm going to have to give a recommendation to this board. While the original BIOS is absolute crap and should be instantly flashed into oblivion, the latest ones seem to keep the thing stable. I'm using the onboard sound and it is lovely, even compared to Audigy 1/2 solutions. The onboard LAN is also superior to most LAN cards I've used.

One cautionary note: runninig a multiplier of less than 10 causes this board's memory access to dramatically reduce in speed. It is unknown whether this is the case with other boards, but you are better off with this board on a Palamino to run 10x18# FSB than you are to run 9x2## FSB.

That said, I've had my XMS up to over 215 FSB on this thing at 9x multiplier, and it ran sweeeet. Since over 200 FSB has little to no gains, stick with the 10+ multiplier.

I would highly recommend the NForce2 platform to anyone who is looking for a new board. $120 for a new EPoX + $20 for a chipset cooler gets you one hell of an overclocker. This also will get you a significant boost in your 3dmarks, which just blew my old ones away (I gained 500 points). The layout of this board is good, as well, with plenty of room around the ZIF and NB for cooling applications. The SB is cramped, only allowing for smaller heatsinks, though as I said, it is really not necessary unless you have a stick on model handy.

BEN: the EPoX 8RDA+ board will get you the 233+ FSB you crave, if you don't mind the NB cooling.

If you have any questions about it, just let me know.

Further: there are no 1/5 or 1/6 mults on this board. The PCI is locked at 33.3r Mhz, even if you change your AGP out of spec for some reason (benchmarking?). Voltage on CPU goes to 2.2V (calls out for a Prometia), RAM to 2.9 (dissapointing), and AGP to 1.8. This board should satisfy the most ardent overclocker among us.

Even though I had severe problems with mine at first due to a bitch of a BIOS, the new BIOSes are finally mature enough to use. I have swapped from an AVOID rating to a RECOMMEND.

[edit]: bad grammar

[jtroutma]

ACK!!! I NEED MONEY NOW!!!!!!


airspirit:
There is one question I do have about the Epox but more about the nForce2 chipset in general. Do you run 2K or XP? If you run 2K, what type of system does 2K detect your board as? As in when you go into the Device Manager and you look under the "Computer" header at the top of the list.

I am asking because everytime I have installed 2K on VIA or SIS chipsets, it has always defaulted to "Advanced Computer Power Interface" and all my devices on my PCI/AGP bus share a single interupt. I have ALWAYS had a problem with that and I have to manually setup my system as a "Standard Computer" so my IRQs dont conflict. Many people dont have a problem with this but all my configs have.

I am hoping that the nForce2 will detect as a "ACPI - Uniprocessor PC" or something that will allow individual IRQs to devices and NOT share them.

Thanks for your help.

[airspirit]

I run XP ... sorry. The NForce2 chipset ... at least the EPoX, will do the IRQ assignment itself. You have the option of letting the mobo do it, or the OS do it. It's a BIOS setting. Let me flash into BIOS real quick and check ... brb.

[airspirit]

You can either let the OS do the IRQ/DMA assignments, or you can have the mobo do it. It is under the PnP/PCI Settings tab in the BIOS. Check out the manual online. Win2K should let the BIOS handle the assignments.

Does that answer your question?

[jtroutma]

Close.......

I think I know the setting that you are refering too but will need to see the BIOS screen to make absolutely sure.

Thanks for you help though.


Now I just need to justify getting a new MB..........<deep sigh>

[HelpImNewbish]

LOOKS like it is PCI locked now, at least in the 12/10 BIOS.

http://www.amdforums.com/showthread....5&pagenumber=2

If not, than an automatic 1/6 divider, or MAGIC that allows it to be pushed way beyond 200fsb, stably.

[airspirit]

jtrouma: I'll try to get the exact wording from the BIOS for you tonight.

newbish: there are absolutely no dividers in the NForce2 platform. It is an impossibility. Hypertransport doesn't use them when communicating with the PCI components. The PCI is locked at 33.3 Mhz, and that's that. AGP can be clocked up, and some boards may ALLOW you to overclock stock PCI settings, but standard settings at any FSB are going to be stock AGP and PCI speeds.

I have a toy HDD that corrupted at 204 FSB on a 1/5 divider on my MSI, and it runs strong at 215 on this with no bumps at all. The reason is that PCI is locked.

This will be confirmed by testing all over the net shortly, you wait and see. Any mention from EPoX that they use dividers and not PCI locks is absolute rubbish, because it is impossible on this chipset.

[bigben2k]

Originally posted by airspirit
BEN: the EPoX 8RDA+ board will get you the 233+ FSB you crave, if you don't mind the NB cooling.

If you have any questions about it, just let me know.

Further: there are no 1/5 or 1/6 mults on this board. The PCI is locked at 33.3r Mhz, even if you change your AGP out of spec for some reason (benchmarking?). Voltage on CPU goes to 2.2V (calls out for a Prometia), RAM to 2.9 (dissapointing), and AGP to 1.8. This board should satisfy the most ardent overclocker among us.

Even though I had severe problems with mine at first do to a bitch of a BIOS, the new BIOSes are finally mature enough to use. I have swapped from an AVOID rating to a RECOMMEND.

Glad to hear about the BIOS update!

Are you assuming that because there is no 1/5 or 1/6 option in the BIOS, that the PCI is locked? Because it could be an automatic divider. How did you verify that the PCI freq. is 33 MHz? Software? Oscilloscope?

[airspirit]

I didn't verify myself, however many places on the web have people that have independantly verified this on every available NForce2 board. Further, Hypertransport spec doesn't require dividers to function, as the chipset itself, from NVidia, has seperate clock generators for the FSB, AGP, and PCI buses. These are each independantly locked, with the BIOS itself giving control over which can be tweaked. I can verify from results on that POS HDD that the bus must be locked, or at 215 FSB that drive would have instantly corrupted. It is an ancient Maxtor 3.6 GB drive, and is not the best overclocker out there (understatement alert).

As I said, verification is being done all over by both real evidence from PC Geigers, and by the theorist who comb the spec sheets.

I'm glad my initial doom and gloom due to manufacturer (particularly EPoX) driven FUD has been proven wrong.

[airspirit]

Quote for AOAForums:

Having trouble understanding how your Nvidia Motherboard keeps time? It's done with: posted by Daniel <- GHEY!!!

"ARGH! 8RDA has too many clocks! quote:

Lets really examine this nForce 2 board, the EPoX 8RDA+.

We know that the SPP (Northbridge) and MCP-T (Southbridge) are connected together via HyperTransport, rather than the more traditional PCI. If you look closely at the HyperTransport bus in the nForce 2 implementation, it operates at 800MB/sec. We also know that is a point-to-point bus, meaning that it only connects two devices together. It's also packetised, so instead of handling individual bytes, it handles packets of data, rather like most PC parts. (Hard disk, Ethernet, SCSI, floppy and so on). This is partly why the nForce can provide a feature called "StreamThrough". "StreamThrough" just means that the HyperTransport bus can guarantee dedicated bandwidth for particular data transfers. This means that the packets of data containing video information coming from the network connector will always be able to travel across the bus when they need to. They don't get pushed around by other devices such as the hard disks, or USB devices. Cool huh?

In addition, the memory can run asynchronously from the FSB. All this is important! Running things asynchronously means that each device run asynchronously needs it's own clock to keep it's own timing.

Before, when everything was running synchronously, there was a master clock that regulated everything. Each different frequency required was simply a division of the master clock. In one fell swoop, all this has changed!

The FSB and memory bus can run asynchonously, so each must have a separate clock. We also know that the AGP slot can run at a fixed clock - this is asynchronous, so it too needs it's own clock.

So far, we have 3 clocks, all independent of each other, and we haven't looked outside the SPP yet! Remember that HyperTransport bus? Well, that needs it's own clock too to regulate the transfers across it. In fact, for every 8 bits, HyperTransport needs a seperate clock. An 8bit wide HT bus needs one clock, a 16bit wide HT bus needs two clock, and a 32bit wide HT bus needs 4 clocks! nVidia's implementation only uses 8 bit wide transfers, so it only needs a single clock!

We're up at 4 clocks now, and none of them are generated from the same source.

Lets move down to the southbridge. This is where information gets a little more sketchy. We already know from the SPP that there's no master clock. Where does the SPP derive it's clocking from then? It would make little sense to derive it from the FSB, as the MCP-T doesn't connect to the FSB, nor is the MCP-T's connection to the SPP regulated by the FSB. It would make more sense to derive the MCP-T's clocking from the HyperTransport bus, as there is only one clock in this case. If the nForce2 chipset had a wider HyperTransport, then it would be more difficult, as there would be more clocks for the HyperTransport bus. An HT bus clock runs at 100MHz

Let's look at what needs a clock on the MCP-T. Hmm, this could be a nice place for a list...





Dual IDE interface (266) ?

PCI bridge (133) (33)

Audio Processing Unit (150) ?

Audio Codec Interface (1) ?

Ethernet (25) (20)

LPC Bus (Low Pin Count) - The remnants of the ISA bus. (1)

Dual USB (50) (20)



Remember, ALL of these are integrated onto the MCP-T. The first number in brackets illustrate the approximate bandwidth needed to run the various parts flat out. The second number illustrates the clock rate needed.

With each of these devices needing different speed clocks, it makes little sense to tie them all together into the non-existent clock generator. Remember that the SPP doesn't need the clock generator, as it has 4 separate clock generators in silicon just to handle it's own interfaces and keep things running asynchronously.

A lack of clock generator, and a radically different architecture to the standard Northbridge/Southbridge system calls for a lot of differences. This includes the MCP-T having it's own clock generator that runs independently of the SPP. Independence means that the audio, PCI, IDE and Ethernet all run in spec all the time. There's no question about PCI locks or an IDE lock, as that would require a complete redesign of the nForce2 chipset! Items like Ethernet and Audio are highly clockspeed sensitive. If you changed the clock for the Ethernet, your machine could no longer communicate with other devices on the network. The audio should be fairly obvious what would happen!

All in all, this points to all the peripherals running in spec regardless of what the FSB, Memory bus or AGP bus do.

As far as items like IDE disk corruption go, there's more to this than meets the eye. People are already beginning to trace some of this strange corruption through to memory, rather than hard disk. Why this should be is odd, but may be related to the caching all modern OSes use. Modern OSes rely heavily on data cached in main memory. This cache can be a sizable chunk of memory (On my 512MB W2K system, the cache is nearly 200MB in size!). Any slight corruption of the cache will cause nasty side effects as this corrupt data gets used. As this data usually includes the various control structures on the disk (FAT for FAT, MFT for NTFS, I-Nodes for EXT2FS and FNodes for HPFS), this can be very dangerous. Couple this with the fact that the control structures cached will be accessed EVERY time the OS wants to do something with the hard disk, and these become heavily used parts of memory.

Phew.. Hopefully that'll give a bit of a better idea about why the nForce2 architecture is so radiacally different from VIA and other's implementations. This means that things you used to understand may be totally different! Fortunately, as time goes on, people will understand the nForce2 better and better.

Áedán"

[bigben2k]

Originally posted by bigben2k
Quote, from the December 12 OC news page:

"Epox 8RDA+ I have a couple copies of emails sent to me from Epox technical support saying that the board 1) doesn't have a PCI lock and 2) has a maximum PCI divisor of /5. So that's the big secret."

Click here

They've since revised it, to state that the PCI is locked.

Here's another review comparing the Epox 8K9A2+ with the 8RDA+. Again, the PCI has been observed as being locked.

In short: the 8RDA+ mopped the floor clean with the 8K9A2+.

[jtroutma]

YEEEEHHHHHAAAAAAAAAAAAAAAAAAAAAAAA!!!!!!!

Thanks to my new roommate, I am now able to afford a new Epox 8RDA+ !!!!

I am going to 2nd day Air the sucker and get it here just before I head out of town.

Sometimes I LOVE my work!!!


BTW Airspirit: how warm does your NB get with that Vantec on it?

[airspirit]

It isn't that warm at all, anymore. That OEM sink is a piece of crap, and there is only a pea sized bit of thermal tape (!) to hold it on, leaving the rest of the NB to cook. With the Vantec, it gained me over 200 Mhz (I'm over 1900+ on my AGOGA Palamino now stable), and the sink is just warm to the touch instead of scorching hot. This board is a candidate for watercooling, if one is so inclined.

I also found as I've clocked it up that the southbridge gets warm as well. With the passive sink that came with my Vantec, it is now almost as warm as the northbridge. This needs to be addressed when cooling the NB, but a fan doesn't appear to be necessary (and due to component arrangement on the board, only a stick-on sink can be applied to it, even though there are holes, unless you want to physically mod your stock HS for it with a dremel).

You will not get even decent performance unless the NB cooler is replaced at least with a blorb, and you paste something to your SB.

Answer your question?

[jtroutma]

My first instinct was to replace the NB heatsink with a Blorb. The SB I will probably leave alone because I have a case fan around the same higth that will provide some airflow for it.

Sidenote: I contacted the company today to see what their lead-time for getting the board out of the shop was and they didnt even have a record of my order?!?!? Also they couldnt even guarranty that they could get it out the dor for 3 days......

Contacted another place and they said that they would ship it out today with Overnight delivery for HALF the price I was going to pay the other place for 2nd day air !!!! It came out CHEAPER to get the board for a slight bit more and OVERNIGHT than cheaper board with 2nd day air and no guarranty that it would make it to me before the end of the week. (I am going out of state on Saturday so I need it in my hands before then )


Few more questions: 1800+XP and the EPOX both support on-die thermal sensor, yes?
On a Palimino, do I still need to close the L1 bridges in order to use the Multiplier settings on the EPOX?

Done for now!

[airspirit]

Pally will need to be L1 unlocked by you. There is a diode reader on this board, but it only used for thermal overload testing. MBM and BIOS reports off a thermistor (crap, huh?). The overload sensor is controlled by core temperature, shutting off when it gets too hot (mine is set to 60C).

[bigben2k]

Here's a comparison of the A7V8x and the A7N8x, from Legion Hardware.

[bigben2k]

Here's yet another review of the A7N8X, by OC NZ.

They say that it doesn't use the on-die diode for temp measurement, but that a lot of NForce2 boards don't either. Good. I wanted to use pHaestus' mod anyways!

[bigben2k]

Here's a roundup of Asus' A7N8X, Leadtek's Winfast NCR18D-pro and Epox' 8RDA+.

It seems that the Asus and Leadteks won't automatically unlock a CPU. Note how this poor fellow didn't unlock the CPU for an OC roundup of tests.

A7N8X is the only one offering raid SATA (but who cares?).