After testing a multitude of boards my "best guess" is they are not locked. On each board i test i start clocking till the system becomes unstable with my Raptors then i simply swap a Maxtor ide drive in their place and keep going. i can run the raptors at 280 fbs on my canterwood board and they wont run past 220 on any athlon board reliably.
Now at first i thought this was surprising, but really its not think about it. They lock the multipliers on every cpu except the FX so that the enthusiast are forced to by that processor, well that would not work to well if we just upped the fbs to make up the performance delta now would it? I mean they admitted they locked the multiplier on purpose to make the enthusiast community buy the FX so why not think they made sure they limited out FBS clocking efforts ( which is the natural next logical step since we cant adjust the multiplier) otherwise we would still just get a 3400 and clock it up using fbs till we hit the cores limitation which would put it faster than the FX, But AMD doesnt want to anger the community so they dont come out and admit it. The only other viable option is that not a single A64 chipset or vendor is capable of producing a pci lock for the cpu because they are all imcompetant and dont want to sell more motherboards. Something is fishy with all this notice how theres no talk of it from the manufactures either.
Sigh and the question to a definitive answer on pci lock continues.
M*XI
__________________
Abit Fatal1ty AN8-SLI w XP90c
2x1gig Corsair PC3200
2 Raptor 74gig SATA 10krpm(raid 0)
2 Maxtor 300 gig 7200 rpm 16Meg(raid 1)
2 200 7200rpm 7200rpm 8Meg (HotSwap)
1 200 gig Maxtor External
2xGeForce 6800 GT 400/1100
Audigy2
Pioneer Slot loader DVD 120s
Plextor 708a DVD burner
NEC 3500 DvD burner
Logitech Z 680 5.1
Windows XP SP1
Last edited by Maximus[X-D] : 02-16-2004 at 02:56 PM.
what video card? because i my old maxtor eide drive will do 245 on the boards ive tested but thats with a Nvidia card in there.
man i wish you had a Raptor or really any serial ata drive to test.
heres how i think about it there is a crystal or something setting a divider either with memory as you suggest or with clock speed or whatever. If you can set a divider you can set a lock yet no lock.
The only boards I've seen do 300FSB(PCI lock?) or better are the Gigabyte and Shuttle. If you want to test yours, drop your multiplier, set your RAM to 4:3, set your LTD to X2 or 400, and let 'er rip.
I noticed the same thing. In DDR400 mode I'm limited to 225fsb but when I drop it in ddr333 I can go all the way to 237fsb. You might be on to something...
Which comes back to my point that 1. PCI lock is achievable since dividers are.
2. if 1 is correct or even if not what would be a ulterior motive for not having a pci lock since no one seems to have brought one to the table. i mean think about the competition for market share in the aftermarket motherboard arena. Don’t tell me that if they could or probably more accurate "ALLOWED" to that someone looking to gain market share would not implement and aggressively market a board boasting pci/agp lock. In fact not only would it be a huge selling point said hypothetical vendor would make sure the reviewers knew about it and pointed out not only their boards ability to do this (make it a shoe in with the enthusiast crowd) but their competitors inability to provide a pci lock. i mean the fact its quite from the vendors, reviewers, and AMD along with the blanket problem it seems to be across all chipset platforms would seem to indicate a power at work greater than the sum of any individual vendor or chipset design team. I.E. AMD themselves. It really makes simple logic why lock a multiplier on a chip to admittedly keep us from overclocking and then permit us to gain that back with FBS? I mean why would they do that? given the already admitted purposeful locking of multipliers to prevent us from overclocking it, thereby keeping the FX as the top of the line cpu, then why would they allow us to drop the multiplier and up the fbs which would let us gain ground, performance wise, even more quickly than adjusting the multiplier would, given the inherit benefits in hypertransport and memory bandwidth that would come from clocking the entire bus.
I mean to say we already KNOW they do not let us adjust the multiplier (preventing easy overclocking) to keep the FX as the top of the line cpu and keeps its premium price.
why would they then leave in place a tool such as pci lock that would allow us to make up that difference even quicker. when you think about it they would be dumb given they are trying to maintain the FX superiority to allow us to overclock the vanilla A64. that coupled with all the buzz in the community yet noone has come forward with a statement as to what the real deal is? Just dont sit right.
Think about it why do we all want pci lock? So we can (without undo stress to all out system componets) lower the multiplier and clock the system to as high as our ram and/or Hypertransport bus will allow ( depending on performance tradeoffs) then up our multiplier till we hit the wall of the cpu. Once we are done with that do you think it will perform better than an Unlocked Cpu running at stock but just upping the multi till you hit the cpu limits would? I know the answer and so does AMD they blocked 1 why not the other? I mean why not?
He gets his shuttle AN50R (nf3 chipset tho) up to 240ish fsb with no issues also, using clockgen.
Screens from clockgen seem to indicate that the AGP/PCI are always set to 66/33, and the author does remark that with some fsb's the agp/pci are changed, but clockgen can reset them to the proper frequencies.
I've been trying to figure out for the longest time if these boards do in fact have AGP/PCI lock...hope this helps.
Processors
Motherboards
Chipsets
Memory
Graphics Cards
Storage
Cases and Cooling
Mobile
Systems
Displays
Shows and Expos
Linear Mode
