icic

Anybody wanna post up screenshots? I need more to prove the theory. Thanx

* antonio_zth, me still waiting for yr new screenshots

A question...

Refering to this diagram, on the HT bus system, where would the potential bottleneck be, given all the parts are the highest-end ones currently available in the market?

if all highest end device that available in the market put into a system, i think the processor should be the bottleneck of a system. The proc cant really optimize the HyperTransport speed yet.

Evogenix

i'll try to contribute some HTT comparison screenies for nF3 s-754. my board with 338mhz ht can run with 2.5x or 3x LDT.

but what benchmark should i run? 3dmark01 ok? shy shy want to run 3dmark03 coz me only haf fx5200

//maianeh

imo, the limitation would be the latest component, as 10 pieces of Raptor running at peak transfer rate can't even saturate the 16-bit HTT bus. Besides, RAM is another limitation too bcoz the on-die memory controller is actually working a lot more faster than RAM.

the HTT is able to transfer up to 4GB/s, and 10 Raptors might give u around 2GB/s in a short burst. Other I/O devices such as LAN, Sound and etc doesn't require too much bandwidth too. The data transfer into GPU at PCI-e is not very big too as modern graphic cards have plenty of VRAM. Most of the textures rendered are stored in the VRAM and the processor usually sends and receives control signals.



//soulfly

thanx dude, you can read post #2 for the benchies, but I'll post here la

Benchies: Sandra CPU, RAM and SuperPI. CPUz reading on 'CPU' and 'RAM' need to be in the screenshot too.

Waiting for u

Test bed:
Sempron64 2800+ @ 2704mhz, 338MHz(HTT) x 8
DFI nF3 250Gb
TCC5 @DDR600 CL2.5-4-3-7-1T
WinXP SP1

*all tests were using the same setting, except for LDT multiplier
*each test ran one by one
*tests for LDT 3x was done first, then reboot before proceeding tests for LDT 2.5x
*A64Tweaker shown just for guide on memory timings

LDT 3x

SiSandra2005 CPU Multimedia benchmark


SiSandra2005 Memory Bandwidth benchmark


SuperPi 1M

LDT 2.5x

SiSandra2005 CPU Multimedia benchmark


SiSandra2005 Memory Bandwidth benchmark


SuperPi 1M

This post has been edited by soulfly: Sep 27 2005, 01:56 AM

Moved back to Hardware

yesterday i saw in a forum said HTT * LDT multiplier cannot more than 1k
izzit true?

Yup, it is suggested to keep it below 1000MHz (not 1k ), so that the system is stable enough.

refer to link info on hypertransport one of the chapter in the athlon 64 bilble. Hope this info is useful. thks

http://66.218.71.231/language/translation/...8428%26page%3d4

by refering to the pic here. yea, i think the Hyper Transport Technology connect the proc to the mobo or (system). Of its I/O communication.

u guys heard of multiplier_less/ratio_less HT on newer RD580 chipsets?



how does it work?

Source: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2609&p=5

the 1000MHz limit has been broken..? or its irrelevent now coz of the new chipset?

This post has been edited by jinaun: Nov 21 2005, 04:52 PM

I'm not really a big fan of the HT (coz I don't own a system with it) but I guess motherboard and system makers have followed the HT Specs, started off at 800MHz, and now 1GHz. It definitely follows a ratio, but mathematically ratio is a simple thing to implement.

FSB runs 180MHz, we make HT runs 5x that value.
So what chipmakers do (i think) is just to implement a circuit that makes the HT 5x faster than some reference value (could be FSB, CLOCK, whatever).

It really comes down to chipset makers to decide whether should they create a circuit which follows certain ratio or create a circuit totally independent of ratios. We have seen this happen in the past with something called asychronous access; with the RAM running at different speed in respect to the FSB, totally disrespecting whatever ratio between the RAM speed and the FSB.

Maybe it's along the same line for HT

finally after a month and half... some1 replied

But one must understand, 1000Mhz is the optimum requirement for the A64 proc. Beyond that is saturation which is overkill unless AMD decide to increase it's internal abillity to take davantage of the 800 Mhz increase. I may be wrong, but that's my current understanding

Saturation = 100% usage. You probably mean underutilization.

Anyway, i don't think it's limited by the CPU. It's limited by the I/O devices themselves. They just don't use that much bandwith.

As ikanayam mentioned, we (the desktop PC user) don't need 1000MHz, maybe 500Mhz is more than enough.

However, industry or server computers need very high I/O capability. This is where Hyper Transport shines.

hrmm. interesting, what i read at DFI-Street mentioned.....



so now, which is which?

That's bi directional bandwidth, because the HTB has 32 tracks, 16 for up and 16 for down, so people have a habit of putting the speed of both directions together . But generally it's understood that we factor in only the base frequency which is 1000 Mhz, something like DDR lah

ahh. that explains it better. cos i'm used to using the actual LDT x HTT figure.