Poor memory bandwith on K7n ILSR

Surpised that SIS Sandra reports my K7n ILSR board to have a memory bandwidth of 2370 mb/s, when compared to a P4 that manages 4000mb/s odd!
Any tweaks in the bios that will improve this figure, or is the bandwidth just very poor on this board?
I am using PC2700 DDR 512mb sticks x 2 '333 dram clock
Thanks for any help!

you better read up on the limitations of the amd cpu chip then
the board etc does better than that but thats all the cpu can handle
onboard vga can use the full bandwidth thats been chucked out but not the amd cpu and its what the cpu uses sandra measures
i assume you dimms are in 1.3 or 2.3

Similar Messages

FX5500 T128 Memory Bandwith

Hi, where can I find the memory bandwith for MSI FX5500 T128? I've tried MSI site, but this data does not there. Thanks

ironcoque,
It is probably 128bit, but that will not help that card for playing today's games. The GPU just does not have the horsepower for today's games and no memory bandwidth will help it.
Take Care,
Richard

Poor memory performance

Hia all you out there
So I bought few days back 857P Neo - mobo. All is okay, but the memory performance of this board is not as good as I expected it to be. I wanted to have the best performance! I've tested this board with SiSoftSandra 2004 and Pc-Mark2002 and they both give a poor result.
I had a ASUS 865PE board before this one and according to tests it performed better . Oh why oh why!
Well here are my specs:
P4 3.2 Ghz (800 mhz)
875P Neo (bios 2.9, newest chipset drivers)
2 x 512 Mb Twinmos 400 mhz ddr-ram
Fortron 350 W powersupply
2 x 120 Gb Maxtor SATA HDD's (no raid)
2 x CD-rom drives (DVD & CD-RW)
ATI Radeon 9800 XT Gfx-card
Audigy player sound card (newest drivers)
I have the memory chips in dimm 1 (Cha A) and dimm 3 (Cha B). So it is a dual channel set up. The sound card is in the 2nd lowest pci slot and other pci slots are free. Pat is on in bios and DOT is set to colonel now. I had it disabled too, but regardless the setting the memory performance was lower than with my ASUS 865PE board. This affects my game play too, not to mention 3dMark scores.
Any help I woul appreciate greatly!
Best wishes, Tatu

Tatu,
You have not informed us as to what your memory benchmark results are/were on either of your platforms.
There are a few generic things to do to get optimal performance from your memory susbsystem:
A) have enought juice - Note 350W PSU is likely to be streached check many postings in this forum on PSUs.
B) Dynamic OverClocking – Disabled
C) Performance Mode – Ultra Turbo
D) DRAM Frequency - Auto
E) Spread Spectrum - Disabled
F) DDR Clock(MHz) - G) Configure Timing by SPD – On (except when timing is too tight)
H) 1:1
From another thread note the following....
Quote
Yes this Depends Heavily on How Fast your Memory is Rated to Run, as Most PC-3200 Modules, with a P4 (C) 800MHZ Bus Speed CPU Usually Score Between 4,600MB/Sec and 5,200MB/Sec (Also Depending on the Speed of the CPU)..And Most PC-4000 Rated Modules can achieve between 5,200MB/Sec and 5,800MB/Sec, My ABIT i875 with PC-4000 XMS-PRO and a 3.2GHZ Extreme Edition P4 can Achieve 6,120MB/Sec with a PBS (Processor Bus Speed) of 225MHZ, And you will only Reach your Max Mem. Bandwith while Running at a 1:1 Ratio (Synchronous) with the Mem. and CPU..............Sean REILLY875

ITunes 7 Poor memory management!

Huge iTunes 7 memory problem.
Upon initial launch, it will use around 40-50mb of real RAM.
If I scroll through CoverFlow it will peak at around 1.5 to 1.6gb of real RAM usage! Going back to normal list view will bring it down to 700-800mb..
I have to quit and re-launch iTunes to get back down to less than 100mb.
If I try this on a computer with less than 2gb of ram OR I am running any other program, iTunes will start to skip because it ran out of physical ram.
I have been able to replicate this on multiple machines.
Very poor QC..

The coverflow implementation is indeed using a ton of memory. It kind of has to if you think about it. In order to quickly browse through all of your albums, it has to preload all of your album covers, which are of decent resolution. On my MacBook Pro with 2Gb of ram, it flows effortlessly, but on my MacMini (intel) of which I plundered the ram for the MBP (now only 768mb), the coverflow isn't very useful. It quickly runs out of ram, and starts paging in and out. However, it does seem to release it's memory quite quickly, even if I just change the from the coverflow view to the standard view.
Apple should probably be clear about the memory requirements of this new feature, as a lot of people will just think it's a bug with iTunes.
Berndt

Kingston HyperX Memory for K7N2 Delta-ILSR

Hey will my K7N2 Delta-ILSR support the Kingston HyperX serises DDR400-512MB Ram?
Im looking to buy 2 maybe 3 of these and really want to be sure theyll work.
http://www.ec.kingston.com/shop/partsinfo.asp?ktcpartno=KHX3200A/256&promo=hx2shp
I checked the supported ram list for the K7N2 Delta-ILSR and they wernt on there so i just figured it may be out of date?
So thanx in Advance for info!
p.s. It will support 3x 512MB right?

Has anyone noticed how the Hyper X went from the poor-man's Corsair to being priced equal to OCZ, and usually higher then Geil, as well as usually being very close to Corsair, if not the same? I think only Mushkin beats it out now price-wise...

MSI-6368 ver 2.1 very poor memory performance

Hi i recently recapped a very old MS-6368 mainboard after which it came up. Previously it used to freeze just before loading OS. The complete system specs are as follows:
P3 8.5C @ FSB 100 = 850 MHz
2 Sticks of PC133 SDRAM (Kingston 512MB & Non-branded 256 MB)
MSI 6368 ver 2.1
Onboard Trident Graphics
Cooler Master 350W
Seagate 20GB HDD (ST320413A), UATA-5
BIOS - optimal settings, Memory Clock - HOSTCLK+33MHz(133 MHz)
WinXP SP3.
I'm getting a memory transfer rate of just 97 MB/s in Memtest and in AIDA (memory read, write, copy & latency as 150MB/s, 306MB/s, 189 MB/s, 220.7ns). I have installed latest drivers (VIA 4 in 1) and still no improvement in memory throughput. Memory timing has been set by SPD.
a) One thing i have observed is "In order Queue Depth value" is 1. Is it normal?
b) I'm planning to use memory interleave enabler for this VIA chipset.
I will be grateful if someone who had this mainboard earlier can share the BIOS settings.

hi, i'm trying to install a pci graphics card(Zotac GT520 conventional PCI) into this old system to play some classic games. It is not getting selected as the primary display device by BIOS during boot though "Init Display" is set to PCI. So i have to boot into dos and by tweaking PCI configuration registers of the graphics card, i'm able to get it to work under windows.
But the performance is really poor probably due to incomplete manual tweaking. I was just assigning BAR's to the discrete card. System information utilities report onboard's video BIOS and not that of the discrete card. I guess the problem may be due to incompatibility in pci versions. Motherboard supports old pci spec 2.1 or 2.2 and the new graphics card may be complying to version 3.0. Further the graphics card is not an actual PCI card. It is a PCI-Express card with a PCI to PCI-E reversible bridge. However the specification states that the PCI-E should be compatible with PCI at software level.
Has anyone in this forum had a working system with MS-6368 mainboard with PCI graphics card? If so, did a normal pci graphics card(not the reversible bridge type as mentioned above) become the primary display device during boot?
thanks,
gandhi.

Poor memory or poor documentation?

This might be no more than a failure of my memory, but I can't figure out what the up statement in the following does:
     IF :System.Record_Status = 'NEW' THEN
          vRecNum := :System.Cursor_Record;
               up;
     END IF;
I can't seem to find any documentation of this as a valid PL/SQL statement, but it obviously is-this compiles without error (although the form as a whole generates an error, possibly caused by whatever this statement does).
thanks.

:System.Record_Status and :System.Cursor_RecordThese are Oracle Forms internal system variables. A search of the Forms Help system would have given you all the information you need about these variables. For example; Start the Forms Help system, Click on the Index tab and type "System" in the search field...
SYSTEM.RECORD_STATUS System Variable
Syntax
SYSTEM.RECORD_STATUS
Description
SYSTEM.RECORD_STATUS represents the status of the record where the cursor is located. The value can be one of four character strings:
CHANGED - Indicates that a queried record's validation status is Changed.
INSERT - Indicates that the record's validation status is Changed and that the record does not exist in the database.
NEW - Indicates that the record's validation status is New.
QUERY - Indicates that the record's validation status is Valid and that it was retrieved from the database.
Usage Notes
Both SYSTEM.RECORD_STATUS and the GET_RECORD_PROPERTY built-in return the status of a record in a given block, and in most cases, they return the same status. However, there are specific cases in which the results may differ.
SYSTEM.RECORD_STATUS can in certain cases return a value of NULL, because SYSTEM.RECORD_STATUS is undefined when there is no current record in the system. For example, in a When-Clear-Block trigger, Oracle Forms is at the block level in its processing sequence, so there is no current record to report on, and the value of SYSTEM.RECORD_STATUS is NULL.
GET_RECORD_PROPERTY, on the other hand, always has a value of NEW, CHANGED, QUERY, or INSERT, because it returns the status of a specific record without regard to the processing sequence or whether the record is the current record.
SYSTEM.RECORD_STATUS Examples
Assume that you want to create a trigger that performs a commit before clearing a Changed record. The following Key-CLRREC trigger performs this function.
IF :System.Record_Status IN ('CHANGED', 'INSERT') THEN
   Commi t_Form;
END IF;
Clear_Record;Related topics
GET_RECORD_PROPERTY built-in
SET_RECORD_PROPERTY built-in
SYSTEM.BLOCK_STATUS examples
Craig B-)
If someone's response is helpful or correct, please mark it accordingly.
Edited by: CraigB on Jan 7, 2011 2:24 PM

EMac has poor memory management

I use an eMac and a G5. I would expect the G5 to be much faster than the eMac but here's a funny thing: I find that the G5 also manages memory much better than the eMac! For example, when I run OmniWeb for a long time on my eMac and really put it through its paces, loading up multiple pages and tabs, I find that OmniWeb starts running slower. Not surprising, but the rest of the computer starts running slower too! For example, creating a new folder may take several seconds.
Even stranger: after I quit all of my programs, the sluggishness continues. My eMac does not recover until I completely restart. Maybe OmniWeb has a memory leak or something, but this doesn't happen on the G5! I can put OmniWeb through its paces on the G5 and launch twice as many programs, including Photoshop, with no noticeable slowdown.
Now here's the weirdest thing of all: both computers have exactly the same amount of memory (768 MB) and both have plenty of free hard drive space (over 40 GB). All the memory on my eMac is Apple brand. I have even clean-reinstalled the OS on my eMac, to no avail.
Is it normal that memory management would be so much better on a G5 than an eMac?

The improvements in each OS generation in memory management involve swapping tasks in and out of memory more efficiently, not using less memory --- invariablly more memory is actually used (to do more tasks). It would've been clearer had I noted that each OS generation does more with the memory and places more stringent demands on the RAM timing. Memory chips tha were good enough under 10.1 failed when 10.2 came out; ditto with 10.3/10.2 and 10.4/10.3
See what Console and/or Activity Monitor tell you is using CPU cycles.

Memory Timming for K7N@

I need help with setting my Kingston Memory timming.... I know that Kingston is not the perfered brand....
There are 4 settings in Bio that need to be set... Tech support said to set them to 3,3,3 that is only 3 settings what about the fourth one. I know that the CASL should be at (3). Any help would be appreiciated.
T-(RAS)
T-(RCD)
T-(RP)
CASL

Mine (Kingston and K-Byte...I know, I know) works great at 2.5-3-3-9.
I tried T-ras @ 7 through 11, and 9 gave me tyhe best performance.

OSX Poor Memory Management.

Hi!
As a former windows user and developer I was happy with its memory management ( Windows 7).
Now I decided to change, and give Mac a chance... and the only thing I have to say.. Im felling as using the old versions of windows... My machine has 8GB of ram and using it for about 2 days running 1 VM (2 GB of ram) and internet browsing gives me 15MB of free memory and 2.5GB swap file!! The system is slow as ****... Btw page outs is about 25% of page ins.
What can I do to stop that ?
Do I have to keep my ( pseudo ) Unix system rebooting? Thats a shame!
Thanks!

Now I decided to change, and give Mac a chance... and the only thing I have to say.. Im felling as using the old versions of windows... My machine has 8GB of ram and using it for about 2 days running 1 VM (2 GB of ram)
Most virtual machines take the amount of memory you give them away from the Mac for the life of the virtual session. So now you are most likely down to 6GB of RAM.
and internet browsing gives me 15MB of free memory and 2.5GB swap file!! The system is slow as ****...
Since you do not indicate what kind of applications you are running (besides the 1 VM), it is difficult to guess what is consuming your memory. Some applications demand more memory than others, especially graphic/video related apps where there is a lot of data to process.
Also having 15MB of free can be considered a good thing, as the Mac is actually using all that memory you paid for. "Inactive" memory is also available to the OS. Its main difference from "Free" memory is that the operating system knows what is in the inactive memory and if the original owner asks for that memory back, the OS can give it to the original owner without requiring an expensive disk I/O.
Btw page outs is about 25% of page ins.
Since the operating system will page a program into memory, it is very typical that you will have more pageins than pageouts.
You can use tools such as
sar -g 60 100
to monitor your pageout activity once a minute for 100 minutes and see if you can correlate heavy pageout activity with the use of a specific application.
You can also use Applications -> Utilities -> Activity Monitor to monitor your programs. If you think you have found a heavy memory consumer, you can select that app, in Activity Monitor, then click on "Inspect", then "Statistics" and that will show you the paging activity.
You can also use Unix commands such as 'top' and 'ps' (see "man top" and "man ps"), as well as 'vm_stat', 'iostat', etc...
Another thing. A lot of people have reported slowness when it turned out their disk was getting lots of errors. This is unlikely, but it has been know to happen.

K7n ILSR dead?

Just bought a new ILSR. When I put it into my system, the system will power on, display the MSI splash page, then sit for about 8 second and power down. This happens whether I use the onboard video or an external card. Proc is Athlon XP 1700, 512 MB DDR 2100 RAM. PSU is 350W PowerMan.
Any thoughts or than I got a dead board shipped?

dougec,
Quote
Kingston ValuRAM, 256mb
Kingston RAM is hit or miss. That could be your problem. Try some other RAM if you have some.
As for the power supply, those numbers are good enough on yours.
Check the Heatsink on your CPU. Make sure it is mounted correctly. Also if you reused it, did you put new Thermal Grease on it before mounting it again?
Take Care,
Richard

Limited Memory Options For ILSR?

Hi all
I'm about to go out and purchase a K7N2-ILSR but I'm in South Africa and my memory choices are only Hynix, Transcend and Samsung.
I've been reading all the posts about memory issues with the ILSR and was wandering if someone can tell which is the better brand of the 3 to go with.
I will either get an XP2600 or XP2800 with the board.
Thanks
John

Our exchange rate isn't too healthy and I wouldn't even know where to begin to look for ram on the net that would actually ship to here.
I thought about the KT600 but from what I understand the NF2 performance is better and the features on the ILSR outway the KT6-LSR.
I can get the board/memory straight from the distributors so maybe I should talk to a technician and see how many returns they've had.
The one distributor sells MSI boards/Hynix and Samsung memory and the other MSI boards and Transcend memory.
My last option is to get ram (Kingston) from another distributor in another city but I noticed that Kingston and ILSR don't get on.
Would I notice the performance difference between ILSR and KT6-LSR?

'Not enough memory' + 'Internal error' x 'Could not render the database' = Parting of company :(

I can’t believe the amount of concern / disappointment / frustration spread across every associated Fireworks forum re: CS5’s
"An internal error occurred"
"Could not render the database"
"Not enough memory"
"Crash without notification"
Etc.,
We installed the CS5 trial – being more than wary about Adobe’s past releases – and lo and behold, all of the above beared ‘true’!
This is nothing new of course, we’ve all been experiencing this since CS3 – but hoping against hope - a newer release would solve the poor memory management and general ‘bug-ridden’ code; alas - as per usual, Adobe has not responded with any pro-solution based action, but successfully furthered our frustration with a couple of fresh gimmicks without strengthening the core software.
As an avid fan of adobe software […and a Fireworks freak] working in a design house that has many different employees with widespread software tastes – I eventually said ‘enough is enough’ after the umpteenth crash [as of 3 weeks ago] and have revisited the Rebel Alliance; ‘yes, CorelDraw’! No I’m not going to go into some tirade about how much better Corel is etc. - as it has its own strengths and weaknesses, too […but without the hourly crashes] – so to be honest, we figure the time increase in some projects due to using Corel […time is diminishing with each project’s acquired experience] are negated by the downtime of Fireworks; so far, this is holding true.
This may seem drastic, but it has been a long time coming – that is, implementing a move from our decade invested workflow to a ‘somewhat’ new schema, but due to the disappointing aforementioned, eventually principal / expenditure comparisons / sanity / lack of support all culminate to such, and if there’s no support for the competition, well – then there is ‘no’ competition; a luxury Adobe has taken for granted way too long.
I truly hope Adobe turns around and fixes their ways, until then – we’ll be supporting those that do – and hopefully along the way, just maybe, with the added funds from disgruntled adobe x-pats – the software will far exceed what I used to love and adore […how I miss macromedia]; honestly, it would be near on impossible to argue which suite was better ‘either way’ – so it may not be such a distant future. Besides, with the market-door Adobe is opening due to such poor software, the new player sniffing around the edges will be welcomed by many with open arms; I know my/our allegiance will go straight to the company with the greatest software stability and sound support, whomever that maybe.
So here’s to hoping no more, and actually doing! Very sad...

I am also getting this error message. It happens randomly. Adobe help told me to close and reopen FW and then it stopped happening. Does this sound like a good solution to anyone?

C771US (C700 Series) Maximum Memory

If I install 4GB of ram in this laptop (I know HP doesn’t support it), the memory will test ok in the BIOS but Vista x86 will blue screen with PFN_LIST_CORRUPT.   If I reduce the memory to 3GB (1 GB in slot closest to MB and 2GB in the other bank), Vista will run stable. Happens with G-Skill F2-5300CL4D-4GBSQ or Kingston KHX5300S2LLK2/4GR RAM.
I've tested this with BIOS versions  F.34 (Current) and F.33. It would be nice if HP Compaq would update the Insyde BIOS to correct this issue.
I know that only 2.8-3.3 GB of memory is available for use by a 32-bit OS. A 32 bit OS has 4GB of address space available. Approx .7 - 1.2 GB of addresses are used for other hardware thereby leaving around 2.8 - 3.3 GB of addresses available for main memory. 64 bit OS would make all 4GB of RAM available to the OS though. I’m going to test the system with 4GB of RAM and Vista x64 this weekend to see if it blue screens as well.
Regards,
Tom
Message Edited by ThomasWilson on 11-18-2008 07:59 PM

After further testing, I found that the 32-bit and 64-bit versions of Vista and Unbuntu 8.10 will not run on this model notebook if 4GB of RAM is installed. Vista blue screens with a PFN_LIST_CORRUPT message, the 64 bit version of Ubuntu crashes with an ACPI error and the 32-bit version of Ubuntu crashes with a large list of problems.
I have tested all of this memory with Memtest86 and have found zero errors. As such, I believe this limitation to be a result of a BIOS bug or a poor memory timing/voltage configuration setting in the BIOS that is not accessible to the consumer.
All of the operating systems will run if 3GB is installed (1GB in the slot closest to the mainboard and 2GB in the other slot - Note: OS will crash if 2GB is in slot closest to Mainboard).   I have three of these notebooks each having 3GB of ram that run rock solid stable. I am using Kingston Hyper-X KHX5300S2LLK2 DIMMs.
While I understand that HP officially only supports running this notebook with 2GB of RAM and as such is under no obligation to address this issue, it would be nice if their engineers would fix it in a future update of the BIOS.
If your running a 32-bit OS, 3GB is close to maximum amount of memory that the OS can use (any thing more is a waste). However, a 64-bit OS can take full advantage of all 4GB.
Regards,
Tom

LabVIEW: Memory full

Hello guys, look I have this situation:
I was running a executable file (I don't have acces to the code) and this error appeared: LabVIEW: Memory is Full VI"Nameofmyvi.vi"was stopped at unknown""at call to"Nameofmyvi" I've read that its possible that this happens when we deal with large data set, at a possible solution is manipulating the code, but since I told you guys before: I don't have access to it.
So, I'll be very pleased if you guys can help me.
Cheers

As already said, the problem is likely with the code having poor memory management. Without the code, there's not much you can do.
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines

Poor memory bandwith on K7n ILSR

Similar Messages

Maybe you are looking for