How do you overclock an A64, I'm used to the P4

[bill gates]

Hello, I recently made the switch to A64 from my 3.4ghz P4. I am a little confused about some of the new settings for the A64. My setup is listed below. Please explain what the following settings should be set to. I will list what I currently have them set at below. I cannot get it to boot with anything over 215 fsb even if I lower the multiplier down to like 7, lower HT to 3x and raise the cpu and memory voltage!!! I hope I don't have a piece of junk motherboard or cpu. By the way I am using the stock cooler for now but the temps are reading 39 celsius. I need to know if my setup is not a good overclocker as soon as possible because if it sucks I will send it back since it is still within my first 30 days.

HT Frequency: 5x

Memory timing: 2T (options are 1T or 2T not sure what this is)


Also the following settings are more obvious but I want to make sure they are set correctly:

AGP Fastwrite: Disabled
AGP Frequency: 66mhz
Spread Spectrum: Disabled
Max Memory Clock: Auto
Cas:3
Ras to Cas:7
Min Ras active time:10
Row Precharge:6
High performance mode:disabled


Athlon 64 3500+ (939)
MSI K8n Neo2 Platinum (latest bios)
1 gig A-data pc-4200 2.85V loose timings
2 36gb Raptors raid 0
OCZ powerstream 520watt
Radeon 9800 pro (not overclocked)
Win XP PRO SP1

[bill gates]

One more thing, now when my system boots up it says that it is running in single channel mode! How do I put it back in dual chanell??? Also it is only running at 333mhz, shouldn't it be 400mhz?????????

[Gascogne]

You should have the HT frequency at x4 up to 250fsb and above that HT frequency x3.

You need to put your Max Memclock at 200 in bios.

Im also new to Amd 64 but learning.

edit* oh and agp frequency set to 67 or 68 then it locks.

[WesM63]

I'm not to sure on the single/dual channel thing, just make sure you have the ram in the right slots.

Anywase, the higher the mem speed the lower you want your HTT multi.

If your using clock-gen, just set your FSB to 201 in bios and then the locks will work.

Edit:karnivore

[bh2k]

All 754 boards are single chan, there is no dual channel ram for them. 939 boards have dual channel on them.

[Super strokey]

Hes on a 939 board

[bh2k]

Oh darn...

[WesM63]

The post I had linked too, althou it violates rule #2 of the forum rules ( ), I have copied and pasted the post I was refering too here. You should note, this was not made by me and credit was given to the original poster.

Posted by Gautam in another forum:

**A HUGE amount of this couldn’t have been written without the insight of D]g[ts and Hitechjb1 from the Overclockers Forums**

My hats off to them
Intro to A64 Architecture

Traditionally, a Northbridge exists between the memory bus and the CPU. The rate at which data is transferred between the memory and CPU is known as the front side bus. However, the Athlon64’s memory controller is on-die, and as such, has no Northbridge, nor a front side bus. The Athlon64’s have two independent buses; one between the memory and the on-die controller, and another bus that communicates with the other system devices- the HyperTransport bus. The CPU’s clock speed is determined by the HyperTransport speed multiplied by a clock multiplier, which is why it’s often suggested to view the HyperTransport bus as if it were the front side bus. However, this is about where the similarities between the two diverge. Traditionally, the memory speed is derived off of the front side bus, and can be manipulated by FSB/memory ratios. In contrast, in the A64, memory speed is derived off of the CPU speed in CPU/memory ratios. This is why it’s rather inaccurate to say that the memory is ever running “synchronously.” The memory is always running asynchronously with respect to the CPU speed, off of which it’s derived. How fast it’s running with respect to the HyperTransport bus does not matter at all. There is no latency hit in running the memory slower than the HyperTransport bus. The HyperTransport bus’ effective speed is determined by an LDT(Lighting data transport) multiplier. While the front side bus could’ve been traditionally double or quad-pumped, the HyperTransport’s effective data rate can be anywhere from 1x to 5x it’s speed on the CPU.

Core variants

Clawhammer- Come in both 512k and 1024k cache variants. Have a 64-bit wide, single channel DDR SDRAM controller. They come in speeds ranging from 1.6GHz to 2.4GHz, with PR ratings ranging from 2800+ to 3700+. There are two revisions of the A64. The first is the C0, which came in both 512k and 1024k flavors. The newer revision, the CG, comes only in the 1024k variety. The last letter in the OPN code indicates it’s revision; AP for C0’s, and AR for CG’s. Both desktop and mobile parts are available on this core. Desktops and desktop replacement mobiles have a default voltage of 1.5v, and an approximate heat output of 81.9W. Mobiles have a default voltage of 1.4v, and approximate heat output of 62W.

Newcastle (754) - These are all 512k parts, and are about identical to the Clawhammers in every other respect. There are no C0 Newcastles. All Newcastles currently are CG’s, as denoted by their OPN code of AX. These range in speeds from 1.6GHz to 2.4GHz, PR ratings from 2700+ to 3200+. The mobiles have a default core voltage of only 1.2v, and heat output of only 35W.

Sledgehammer (940)- Found the Opterons and the FX51, and some FX53’s. These all have 1024k of L2 cache, and have a 144-bit wide dual channel memory DDR SDRAM controller, supporting only registered memory. Speeds range from 1.4GHz to 2.4GHz. These are all based on a 940-pit socket package. The OPN codes that you will likely see are AG, which are B3’s, AK, which are C0’s, and AT’s, which are CG’s. They have stock voltages of 1.50v, and wattage ranges from 82.9W to 89W. The Sledgehammers have eight HyperTransport links for multi-way processing. These are factory enabled/disabled as appropriate, and cannot be modified.

Clawhammer (939)- Found in the FX53, and soon to be in the FX55. Currently only the FX53 exists, with a clock speed of 2.4GHz, OPN code AS, revision CG. Can you tell I’m getting tired yet? Stock voltage of 1.5v, wattage of 89W. Sports a 128-bit wide, dual channel memory controller, and supports unregistered memory.

Newcastle (939)- Exactly the same as above, except only have 512k of L2 cache, and are found in the 3500+ and 3800+, OPN code AW, revision CG. Unlike the 940, the 939's have only one HyperTransport link, although for the end-user, this doesn't actually make any difference.



Available Chipsets

VIA K8T800- One of the highest performance A64 chipsets. Supports a maximum HyperTransport effective rate of 800MHz, but unfortunately, lacks AGP and PCI locks. AGP and PCI rates are determined by either a 1/6 or 1/7 divider off of the HyperTransport bus.

VIA K8T800 Pro- HyperTransport effective rate of 1000MHz supported, some motherboards have AGP/PCI locks, some don’t. Available for all sockets. Also has native support for SATA RAID 0+1, which is advantageous, as the PCI bus wouldn’t be used.

nVidia nForce3 150- Slightly lower in performance compared to the VIA’s, supports a max effective HyperTransport rate of 600MHz, but sports AGP and PCI locks in most boards, a huge plus. For the boards that don't have PCI locks, some SFF boards, PCI dividers up to HTT/9 are available, so even these shouldn't have trouble overclocking.

nVidia nForce3 250- Same as above, except supports a 1000MHz HyperTransport rate.

nVidia nForce3 250GB- Same as above, but with a richer feature set. Native support for SATA RAID 0+1, Gigabit Ethernet, and has a built-in firewall feature. This is the preferred chipset; with such rich native support, the PCI bus can be kept quite clean.

SiS 755- A very promising chipset, humbling both the nVidia’s and VIA’s handily at the same speed, and sports AGP/PCI locks. However, motherboard support for this chipset is quite lacking, and there isn’t a solid solution for overclocking that utilizes it to date.

Configuring an A64 System

For your everyday overclocker, one of the most cost-effective solutions is to opt for a s754 512k 3000+, a CG Newcastle. CG Clawhammers are better performers, but are difficult to find outside of mobile parts. Most boards to date have certain issues with these.

The 939 Newcastles are rather pricey. Their dual-channel compensates for the lack off cache and adds a bit of a performance as well, but nothing especially significant. However, if mathematical applications, etc. are a must, this may be worth a consideration.

For the best of the best, there is no substitute for the 939 Clawhammer, AMD’s flagship line. The FX53 virtually untouchable by another processor once you get it going.

As far as chipsets go, the nF3 250GB is the best choice for most. 939 boards based on this chipset are right around the corner, as well. However, the K8T800 Pro has a slight edge in performance. Some VIA boards have PCI/AGP locks, and others don’t. The one’s that do can sometimes be temperamental. For hardcore overclockers, the VIA may be the best route. Currently, good choices for socket 754 are the EPoX 8KDAJ/3+, based on the nF3 250GB chipset, the Gigabyte K8NS Pro, which, although does not have the GB chipset, is still a very solid overclocker, and the VIA-based Abit KV8 Pro. The KV8’s seem to have PCI/AGP locks, and sport a wide range of voltage options. The VIA-based Asus A8V and nVidia-based Gigabyte K8NSNXP are the front-runners. I would opt for the latter personally, because of its no-fuss vdimm mod, and guaranteed PCI/AGP lock. (The Asus appears to be hit or miss).

Memory is where things get tricky. Running at memory voltages of over 3.3v has proven to be rather dangerous, so getting low latency memory, such as BH5/6, up to the extremely quick speeds that the A64 can take is usually not possible. Still, IMHO, it is a much better choice than high latency, high speed memory, even without a voltage mod. The tight latencies tend to be more useful than slighter higher bandwidth. Another complication with the A64 is that it tends to be very, very picky with double-sided modules. When running two at a time, it can be very difficult to get a decent overclock, but even with one, don’t expect to get as far as you may on other platforms. This is another reason why it’s a good idea to use low latency memory. Hitting a relatively low speed wall is quite possible, and if you’re using high latency memory, you’re pretty much trapped with low performance in this situation. Low latency, low speed, on the other hand, is still very competitive. My top picks would be memory based on Micron mT (e.g. OCZ 3200/3500/3700 EB), Winbond BH5/6 (e.g. Mushkin 222 Special), or Samsung TCCD (Corsair 3200XL), in that order. All are solid choices, and can be found in dirt cheap value memory as well, if you look hard enough. If you plan to go the high speed route, Hynix D5 is just about the choice that I can recommend, as it can reach 280+ speeds, and give low latency memory a run for it’s money. Only use single-sided memory in this situation.

If you are planning to use a motherboard without PCI/AGP locks, it is advisable to use a PCI66 compliant SATA/ATA controller card, to allow you to use high PCI speeds without corruption. Also desirable in this case are video cards that can support high AGP speeds. (Especially nVidia-based cards)

Overclocking Technique

You are essentially in control of the speeds of four different…ummm lets call em data paths; the CPU, the memory bus, the HyperTransport bus, and the HyperTransport’s effective data rate. Overclocking the CPU is done pretty much as it always has been, except that the HyperTransport substitutes the front side bus.

The HyperTransport can almost always go just as high as you need it, granted that you don’t exceed the motherboard’s maximum supported data rate by too much. For systems that support a 1000MHz HyperTransport data rate, for example, one could use a 200MHz HyperTransport bus with a 5x LDT multiplier. However, using 5x250 would result in an effective 1250MHz, which would almost certainly lead to instability. The LDT could be dropped to 4x, allowing for a higher HyperTransport bus speed with stability, 250, but resulting in the same 1000MHz effective data rate as default. The HyperTransport bus being increased alone doesn’t accomplish anything; unlike increasing the front side bus does other platforms. Even raising the effective data doesn’t add any noticeable increase in performance, as the bus is already so wide, that saturating it isn’t very likely. For this reason, the nF3 250’s and 150’s perform quite similarly to one another. My suggestion would be to leave the effective rate at as close to stock as possible, and raise the HyperTransport bus speed only as much as necessary. Unless you’re using an 8x multiplier, there shouldn’t be much reason to go far above 300MHz in many cases.

What complicates what multiplier and HyperTransport speed you should use are the CPU/memory dividers. No motherboard allows you to manipulate them directly. Instead, they provide “maximum memory clocks,” or supposed HTT/memory ratios. Make no mistake, no such thing exists. The memory is derived off of the CPU speed, but it’s never made clear, and the dividers need to be manipulated indirectly. Also, CPU/mem dividers are integral only; there are no half dividers, so it’s advisable not to use half multipliers. Ok, what I just said probably doesn’t make much sense, so here are some examples of how to get certain get CPU/mem dividers:

CPU/5-8 - Set memory to 200 and multiplier to desired divider

CPU/9
Memory to 200, multi to 9x(if available)
Memory to 183*, multi to 8x

CPU/10
Memory to 200, multi to 10x(if available)
Memory to 183*, multi to 9x(if available)
Memory to 166, multi to 8x

CPU/11
Memory to 200, multi to 11x(if available)
Memory to 183*, multi to 10x(if available)
Memory to 166, multi to 9x(if available)
Memory to 150, multi to 8x

CPU/12
Memory to 200, multi to 12x(if available)
Memory to 183, multi to 11x(if available)
Memory to 166, multi to 10x(if available)
Memory to 150, multi to 9x(if available)
Memory to 133, multi to 8x

Doesn’t make much sense? Don’t worry, it shouldn’t. You will probably need to experiment with different multipliers and max mem clocks to find the CPU/mem divider that you desire. Using half multipliers complicates things further, as the memory is divided integrally. Just to eliminate variables, drop the LDT multiplier down to 3x if your board supports a 1000MHz HyperTransport speed, or down to 2x if 800MHz or 600MHz is it’s maximum. You can also increase in the HTT/LDT voltage on some motherboards, which can give you an extra 20-50MHz extra MHz on your effective rate in some cases.

As I've always stressed, overclocking needs to be done carefully and systematically. This is especially important with the A64. Focus on one area that you wish to overclock, and overclock it alone. For example, if you wish to overclock your memory, drop your LDT and CPU multipliers as low as they can go, and see how far your memory can go with everything else clocked low enough to not hinder stability. For overclocking the CPU, drop the LDT as low as it can go, and set the max memory clock as low as it can go as well. Once you find your maximum memory speed and CPU clock, play around with the max memory and CPU multiplier to find the suitable CPU/mem ratio. Once you've already got in mind how far the CPU and memory each can go, this isn't too difficult. I cannot stress enough how important it is to isolate variables. It's all too common that people try to max everything out at once, fail, and then give up out of frustration. Take your time, be patient, and have fun. Dividing and conquering can make the task of overclocking the A64 a lot less daunting.

This excellent tool by Cpjk can be very helpful in removing some of the confusion in figuring out how to run things.

If you're lucky enough to have an FX, though, you don't need to bother with finding the right CPU/memory ratio. Simply find your maximum memory clock, and then increase the multiplier as necessary to max out the processor.

On a related note, the absolute core voltage for A64’s rated by AMD is 1.65v, opposed to 2.25v(I believe) for Bartons. The heat output of A64’s at the same speeds as AXP’s is roughly equivalent to what they’d put out with 0.2v less. For this reason, it usually is not too beneficial to exceed a core voltage of 1.7v or so on air cooling. On your everyday R404A setup, 1.8-1.85v usually appears to be all that’s needed for an optimal overclock.

One rather important memory-related setting is the command rate, a.k.a CPU Interface on many other boards. The default for C0 processors is 1t, and the default for the CG’s is 2t. 1t is quicker, but makes overclocking the memory with double-sided sticks especially difficult in many cases. Running at 2t, however, takes off about 1 sec in SuperPI and PIFast, and makes a couple hundred point difference in 3DMark01. The one benchmark where it takes a significant toll is the Sandra Memory Bandwidth Benchmark, where it takes 10%, or 300-400 MB/sec off. I don’t see having to run at 2t as the end of the world, unless you’re a Sandra fanatic. The difference between 1t and 2t is actually less than that between tRCD2 and tRCD3 in my experience. Again, there is no one size fits all solution. It may take some experimentation to see what combination of command rate, latencies and memory speeds are optimally for you. Low tRCD is highly recommended, but CAS doesn't matter very much. tRAS at 10, and nothing else, seems to deliver the best performance, while backing it down much lower begins to hurt.

Some notes on Windows Tweaking/Overclocking
Overclocking A64s within Windows was originally done when high HTT’s caused BIOS corruption, however this doesn’t appear to be an issue today. It still can be very convenient, and for some boards like mine that don’t allow overclocking in the BIOS with mobiles, can be a godsend. ClockGen is a Windows-based utility for overclocking. It allows multiplier, voltage, HyperTransport speed, and PCI/AGP bus speed manipulation. Changing the voltage doesn’t work on all boards, and the CPU/mem ratio and LDT multiplier cannot be changed using the utility, so some settings must be set in the BIOS. It also allows for profiles, so you can quickly change speeds on the fly. To make a profile, put the signature of your board, as found on the website in brackets on the first line, e.g. [CG-NVNF3] for nForce3’s, and then the values you want to change in the succeeding lines, e.g. FID=9.0, HTT=250. For nForce3 boards, if you set your AGP rate or HTT rate anywhere above spec in the BIOS, the AGP/PCI lock is enabled, so you can increase the HTT easy in Windows. Setting the HTT to 201 in the BIOS is the most common technique. The nVidia System utility is a nice tool to have for nVidia-based boards. It allows manipulation of the tRAS, tRCD and tRP within Windows, and also the changing of the HyperTransport and AGP/PCI speeds. A64 Tweaker is an excellent utility written by CodeRed. It allows manipulation of just about every memory-related setting on the fly in Windows. It’s made my life dozens of times easier when trying to test things out. It also has much more functionality than you’ll find in most BIOSes.

[bill gates]

Thanks for the great info. It is time to go back to the drawing board and start overclocking things little by little. The A64 is definitely a lot more complicated to overclock than the P4. I will let you know what my results are.

[WesM63]

It sure is a lil more complicated but once you get the hang of it, its a peice of cake.

[bill gates]

Well, so far I have got the fsb up to 225mhz with the HT set at 4 and the multiplier set at 10. I finally figured out that my memory had to go into a different slot and that made a big difference. It isn't to great for the synthetic benchmarks but in 3Dmark 03 it is better than my P4 was. In Sandra I got 6137/6065 for a memory bandwidth score which I think is pretty good for an A64 isn't it? Well I have just begun to figure out how to overclock this thing but I am pretty happy so far. Also my temps are staying around 35-40 celsius with the stock cooler although I haven't reaised the voltage yet.

[The Mofo]

No C0 steppings on 754 Newcastles eh? I guess i have an extremely rare chip.

[WesM63]

I didn't write the article, but I'll let the arthor know.

[saaya]

ras to cas7???? 0_o

timings are usually listed like this:

cas (cas latency)
trcd (ras to cas delay)
trp (row precharge)
tras (min ras active time)

and are

2.0
2
2
5

or

2.0
2
2
6

for the lowest latency memor and

3.0
4
4
8

for the highest latency DDR memory.

i never heard of memory that uses

3
7
6
10

timings

[bill gates]

Right now I am up to 250mhz x9 with 4HT and the memory running at 1T. It is running smooth but I really want to push it to 260 x9. I was able to boot up at 295mhz x7 but that only gave me 2065mhz. (just wanted to see how high the memory would go). I think I am finally learning how to overclock the A64 platform and now that I understand it I realize it isn't very hard at all. Basically make sure the HT doesn't exceed 1000mhz by too much and it runs pretty stable.

[WesM63]

Originally posted by bill gates
Right now I am up to 250mhz x9 with 4HT and the memory running at 1T. It is running smooth but I really want to push it to 260 x9. I was able to boot up at 295mhz x7 but that only gave me 2065mhz. (just wanted to see how high the memory would go). I think I am finally learning how to overclock the A64 platform and now that I understand it I realize it isn't very hard at all. Basically make sure the HT doesn't exceed 1000mhz by too much and it runs pretty stable.

[boshi]

Originally posted by bill gates
Right now I am up to 250mhz x9 with 4HT and the memory running at 1T. It is running smooth but I really want to push it to 260 x9. I was able to boot up at 295mhz x7 but that only gave me 2065mhz. (just wanted to see how high the memory would go). I think I am finally learning how to overclock the A64 platform and now that I understand it I realize it isn't very hard at all. Basically make sure the HT doesn't exceed 1000mhz by too much and it runs pretty stable.

[WeaponX]

I'm new to AMD64 overclocking as well. I'm quite at this point. Isn't the AMD64 3500+ clocked at 2.2mhz stock. I plan on buying one plus some OCZ 3500. My goal is to hit 3800+ Speeds and beyond if possible (watercooling). So my question is how do I get a 2.2mhz AMD64 cpu to 2.4mhz and beyond? Let me rephrased that, I'm not worried about the HT speeds, what key points do I have to worry about to get the clock to increase from 2.2mhz to 2.4, 2.5, etc?

Sorry for such noobie questions...

[The Mofo]

Quality ram, higher voltages then motherboards allow stock and of course cooling.

Spring for some OCZ Pc3700EB, get yourself a DDR Booster and go for some watercooling.

OCZ PC3700EB maxed my motherboard out. My mb is only good for 250 fsb.. Before that, i was straining at 230 with other memory. I have no volt mods on this Aopen 89MAX but i was able to push the memory to 265 in windows before she crashed.

What does this say? I should see 270-280 easy with my new DFI with its stock max voltage of 3.1v on the dimm, 2.0v on the CPU and 1.9 for the chipset. I also wont hit the 250 max fsb barrier.

Ive decided to upgrade a lot of stuff. New waterblock, new pump, dual radiator, new case and new PSU for my new DFI that should be here no later then fri afternoon. It may not allow my chip to hit 2.5ghz but ive been using the same case for 5 years, same pump for 3 and the new PSU will give me a better platform to work with down the road. Ive put off certain upgrades for to long and now its caught up with me. Its a to catch up! LOL

Goodluck man

[bill gates]

Well since it is not an FX processor you cannot raise the multiplier any higher which would be the easiest way. (you can lower it though). You will have to experiment with raising the FSB speed (if you can even call it that anymore) to try to get to a higher speed. So far I am having some trouble, I think it is because my Nvidia SATA contoller does not like high FSB speeds. I may have to buy an add-on SATA raid contoller for my raptors. I only wanted to raise my fsb speeds so I could run my memory at a higher speed but I also want to get as close to 2.4ghz as possible. As soon a s I get my new cooler in the mail I will be able to start adding voltage and hopefully will at least be able to do 9 x 260 which would be 2340mhz. One thing I found out with my motherboard is to set the memory to 200 and everything else on the memory to AUTO and that made it work a lot better. Funny thing is my memory does better at 1T then it does at 2t...can't explain it but I will take it! So far the Athlon kicks ass in games, but general windows apps do feel quite a bit slower to me. I kind of miss the Hyperthreading a little bit because I like to burn dvd's and surf the internet without any slowdowns. Oh well Gaming is the main use so I am happy. When I get it overclocked where I want it I think it will blow away my P4.

[WeaponX]

You mean, it doesn't blow away your P4 already? Another question bro, since we both come from P4 overclocking, any disadvantages you can tell me about the AMD64? I just heard it's multitasking abilities just aren't up to par with P4s

[The Mofo]

Originally posted by WeaponX
You mean, it doesn't blow away your P4 already? Another question bro, since we both come from P4 overclocking, any disadvantages you can tell me about the AMD64? I just heard it's multitasking abilities just aren't up to par with P4s

If that was the case, then the opteron would get raped by a Xeon. HT is a bandaid for the 20-26 stage pipeline intel has these days.

[bill gates]

Well yeah after using both I would say that the P4 is faster at multitasking. It feels faster to me anyway. But as far as games go well all you have to do is look at the benchmarks. And no it does not blow my p4 away in it's current state, in fact when using just general windows applications it feels noticeably slower but once I get this thing overclocked then I will see what it's true potentional is. Either way I think both processors have their good and bad points it really comes down to what is more important to you. I didn't have to pay anything to upgrade to my 3500+ so I am happy but if I had to pay to upgrade to it I would be a little dissapointed that it is slower for multitasking and general windows apps. I am just glad that I went with a 939 so that I can upgrade the processor later down the road. I will let you know as soon as I get my heatsink and can up the voltage and get a decent overclock. It should be really fast then. But just for an example of the difference between the two platforms my PCMARK04 score dropped 1000 points but my 3DMARK03 score went up about 100 points. Now remember that my P4 was heavily overclocked almost 50% and my A64 was stock. Sorry I can't give you a better recomendation right now but so far I am not getting the kind of results I was expecting.