Buying a new computer, clueless on some factors

[CmdrFly]

Mobo I'm looking at is the Asus A8R32-MVP Deluxe based on the ATI Xpress 3200 (RD580 northbridge / ULi M1757 southbridge). Board will be married with an Athlon X2 4400 and OCZ PC4000 Platinum 3-3-3-8 as dual channel (2x1gb sticks).

This motherboard is grossly overclockable. All the reviews I have read put it in the top 5 of modern overclockable boards. Fine and dandy of course, if I had any idea what the hell I'm doing with overclocking.

What I'm mostly lost on, is the purpose / capability of the HTT (HyperTransport Thisletterrepresentsaword) speed:

One of the most interesting new features in overclocking the RD580 is that the chipset is designed for high-speed operation. In most cases, current NVIDIA and ATI chipsets need to have Hyper Transport frequency adjusted to 3X around a 260 to 270 Clock frequency to keep HT speed at or below 1000-1100. The RD580 chipset was designed for higher HT speed. We consistently found that the board could easily handle HT speeds in the 1450 to 1550 range. For most overclocking this means, for the first time, that it is not necessary to even adjust the 5X HT setting for most overclocks. It was only where we were exceeding around 1500 (300 Clock Frequency) that we had to drop the HT one notch to 4X. Overclockers should be very happy with this new feature of the RD580 chipset.

WTF does that mean? Is HTT a measurement of some clock × some multiplier? Tied in with mem timings / command delays?

And just where is Carmen San Diego?

 

[Guiller]

Alienware Ftw!!!!!!!!

 

[tauromachine]

I'd stay away from OC'ing unless you know more or less what you're doing. It's become much easier in the past couple of years, but I'd still avoid it.

I personally look for stability, though, which is kind of the opposite end of that spectrum.

 

[Bustaballs]

Over-clocking is usually pretty stupid reguardless of whether you know what you're doing or not. I don't even trust any of the crap that's meant for OCing.

 

[nofriends]

Asus boards are VERY reliable for OC'ing, and with their A.I technology, they can take a lot of the guess work out of it for you. With their bios, you can safely OC' 10-20% (assuming you have the proper cooling) and the system will make all the necessary changes for you. Ive NEVER had an issue with any ASUS board ive bought for my own personal gaming rigs.

 

[CmdrFly]

I never said anywhere that I was overclocking... I was asking about HTT. There really isnt anything scary about overclocking if you understand exactly what you're doing.

 

[nofriends]

To be honest, I dont know much about HTT, but from what youve said I think you have the general idea down though.

OC'ing isnt all that dangerous in reality as long as you do it in progessive steps. Dont crank it up to 50% over stock and see what happens, turn it up in baby steps and test stability and temps and continue on. If something isnt working correctly, back it down to an acceptable level. As long as you have proper cooling, OC'ing does not hurt the life of your computer as much as people think.

 

[tauromachine]

Oh, my bad. I didn't quite read the OP completely. It does sound like you've got at least the principle of it. Everything I'm seeing seems to indicate it's a multiplier and would need to be adjusted with core speeds. I'd bet, though, that if you're not OC'ing, it's not something you'd need to worry about. At the same time, I'm just a guy on the internet, so make of it what you will.

 

[Raziella]

WTF does that mean? Is HTT a measurement of some clock × some multiplier? Tied in with mem timings / command delays?

Sounds about right to me. I'm not familiar with the technology, nor do I overclock vidcards very often, to know for sure. I guess HTT timing has to be taken into consideration when overclocking the GPU from what that description says.

I haven't watched it yet('cause I'm at work right now), but this might help explain things: http://www.mindshare.com/knowledge/?section=HyperTransport|TM



and btw:

Alienware Ftw!!!!!!!!

They are owned by Dell now, for starters, and they are gaudy, overpriced systems to attract gamers with more money than brains... "The yuppy's gaming machine" They can brag about the equipment (and how much it costs) even though they probably haven't a clue what half of it does (or use it to it's full potential). Sorry, but Alienware FTL.

 

[CmdrFly]

I decided to put in an ounce of effort for a change, and discovered this:

AMD's A64 platforms have abolished the ‘Front Side Bus'. The ‘Front Side Bus' was essentially a data bus that carried data to and from system components and the CPU (usually by connecting the CPU to the north and southbridge chipsets). These chipsets provide connections to other buses, such as the AGP and PCI bus, and many other system components. All current Intel platforms, and pre-A64 AMD chips follow this basic model.

AMD decided to do things a little differently with the A64, and adopted ‘Hyper Transport Technology'. Hyper Transport is a high-bandwith, low latency computer bus that replaces the aging FSB. The Hyper Transport bus does essentially the same thing as the FSB, only much faster. Many people find themselves still calling it FSB, but for the sake of correctness, we'll call it the HTT bus.

There is a common misconception that Hyper Transport Technology is a proprietary AMD technology. HTT was developed by the ‘HyperTransport Technology Consortium'. Hyper Transport, sometimes called LDT (Lightning Data Transport) has been used by many vendors, including nVidia and Cisco Systems. You may recall that nVidia used HTT to provide high bandwidth, low latency communication between the north and southbridge chipsets in their older socket A ‘nforce' platforms.

The Hyper Transport bus operates using a multiplier system to derive its overall speed. The ‘base' or lowest HTT frequency that the HTT Consortium defined is 200MHz. The overall operating clock speed that the HTT operates at is simply a multiple of that 200MHz base or ‘reference' frequency. Many other clock frequencies are also derived from this 200MHz reference clock, such as CPU clock speed. Most Socket 754 A64s, for example, operate at an HTT speed of 800MHz. A clock multiplier of 4x was used to obtain this. So 200x8 = 800MHz. You may be asking why AMD lists an HTT speed of 1600MHz in the specifications for these processors. HTT is a ‘double pumped' or ‘double data rate' technology, much like DDR RAM. So, simply double the final result. 200x8 = 800MHz x 2 = 1600MHz. I'll get more into A64 Mathematics later on.

So that's it. HTT = the new FSB. Now I feel really dumb for ordering meek 400mhz ram, but I hear there's more benefit to buying ram with good memory timings over a higher max bus speed. Can anyone comment on that?

 

[Raziella]

So that's it. HTT = the new FSB. Now I feel really dumb for ordering meek 400mhz ram, but I hear there's more benefit to buying ram with good memory timings over a higher max bus speed. Can anyone comment on that?

I think as long as your memory speed matches your FSB/HTT, the timing should be fine. I think it becomes more of an issue when your overclocking because the speed is native to the max speed's timing. But again, I'm not a big overclocker, so take that with a grain of salt.

 

[Eulore]

A64's stock FSB is 200MHz anyways.

And mem timings are nice, I use 2-2-2-5 and they run like a dream

 

[Langly]

CAS 2 memory has a warm spot in my heart.

 

[Bustaballs]

=/ I don't have the money to stay up to date with you guys. I'm still stuck with my PC2100 DDR 512MB stick. I barely meet Oblivion's minimum req ; ;

The gaming world is going to leave me behind.

 

[CmdrFly]

I think as long as your memory speed matches your FSB/HTT, the timing should be fine.

Eulore's right - stock HTT on an AMD X2 is 200mhz, so the DDR400 works just dandy. Regarding this quote though, you can actually crank your HTT way the f up and downclock your memory back to 200 (or whatever you choose) using memory dividers.

Example, say you overclock your core HTT to 300mhz. Use a 3:2 memory divider, and your HTT will run at 300mhz while your memory will run at 200mhz. By default, divider is set for 1:1.

And mem timings are nice, I use 2-2-2-5 and they run like a dream

I always used to think that lower timings across the board were better, and as such was looking at some 2-3-2-5 sticks. But then after reading this article (see the first diagram on that page), it's apparent that a tRAS of 8 is better than 5.

My current curiousity is whether or not I'll be able to run 2x1gb at 2-2-2-8 on a 1T command delay. Everything I've read thus far says that I'mma have some troubles getting 1T stable with dual large sticks.

 

[Raziella]

Eulore's right - stock HTT on an AMD X2 is 200mhz, so the DDR400 works just dandy. Regarding this quote though, you can actually crank your HTT way the f up and downclock your memory back to 200 (or whatever you choose) using memory dividers.

Example, say you overclock your core HTT to 300mhz. Use a 3:2 memory divider, and your HTT will run at 300mhz while your memory will run at 200mhz. By default, divider is set for 1:1.

Well I'll be damned. That's pretty slick. I'll have to keep an eye out for that next time I'm in the market for a new mainboard.

 

[masterorion]

i figured teh HTT was the same as FSB. one thing ive noticed lately is alot of new computers that you buy off the shelf come with a bus speed ranging in the 1000-2000 range. 2 years ago, 800 used to be a massive amount when it came to bus speed.
thing with the bus speed is, its going to be limited by one thing or another. if the processor maxes at 200 but the board maxes at 800, the most it can run at is obviously 200. and that is also reversable as in if the processor runs at 800 bus speed but the board only has 200 or whatever, the processor can only go as fast as the board can handle.
most of you know this, but i figured i would throw it out there just as useful info to those that didnt know

 

[CmdrFly]

thing with the bus speed is, its going to be limited by one thing or another. if the processor maxes at 200 but the board maxes at 800, the most it can run at is obviously 200. and that is also reversable as in if the processor runs at 800 bus speed but the board only has 200 or whatever, the processor can only go as fast as the board can handle.

Not quite the same with newer CPUs. Description below is fairly common for both AMD and Intel, but keep in mind AMD's memory controller is on-die nowadays, while Intel for some inexplicable reason still stores their controller outside of the physical CPU chip. They really need to fix that, but back to the topic:

Where FSB was the bus for traffic from CPU to memory and Northbridge, HTT is the bus for traffic from CPU core to CPU core, for CPU core to memory controller, and for CPU core to northbridge. The increase in density of CPUs today completely stomps the density of older layouts, and as such the density of HTT lanes stomps the density of FSB lanes.

Stock AMD HTT bus speed = 2ghz (200mhz DDR = 400mhz @ stock 5.0x multiplier = 2000mhz). Flat. Going from core to core, core to mem controller, or core to northbridge, they all go at 2ghz.

There is no bottleneck with HTT anymore. The bus, and all data on it, runs at 2ghz (unless of course you're OCing).

 

[Bustaballs]

Man.. you guys have left me.

I was a computer tech when I was back in Georgia but I've been away from it for so long that I've forgotten quite a few things and I've been left behind with this 64 bit era. I haven't touched a 64 bit board yet. I'm going to have to catch up when I visit Georgia.

 

[CmdrFly]

I'm pretty sure the majority of the CPU architecture changes came with dual core rather than 64-bit processors. It doesn't take much reading to wrap your head around all the new stuff - pretty much everything I now know about x64 and x2 processors, I've learned within the past week while making my purchases.

Hit up Anandtech every few weeks and spend an hour or two there. Maik u fele geenyus yo