Print 68 comment(s) - last by ajalbetjr.. on Oct 2 at 8:19 PM

No, it still isn't a special edition Mercedes

With Barcelona finally launched, the enthusiast community eagerly awaits the launch of AMD's Phenom desktop processors. A large gap, however, still remains between now and then, and what better way to fill that gap than with a new AMD Athlon 64 product?

Today AMD silent-launched its Athlon 64 X2 5000+ "Black Edition." The Athlon 64 X2 5000+ is AMD’s second processor offered under the Black Edition moniker; the first being the AMD Athlon 64 X2 6400+.

In reality, the Athlon 64 X2 5000+ can hardly be considered a completely new product. It still features the same 2.6GHz clock speed, 1 MB of L2 cache, and 65 Watt TDP. In fact, all technical aspects of the processor are identical to its non-Black Edition counterpart. The only new feature that the Athlon 64 X2 5000+ Black Edition brings to the table is an unlocked multiplier -- allowing overclockers to increase the frequency of the processor without the need to adjust the HyperTransport clock.

It is unmistakable that AMD is attempting to entice enthusiasts and DIY PC builders with the unlocked multiplier. Thus, success of the new product centers on its overclocking performance.

According sources familiar with the new CPUs launch schedule, retail availability of the processors should come later this month, but as of right now no major supplier is selling the new processor.   

AMD guidance sets the MSRP of the Athlon 64 X2 5000+ "Black Edition" at $136 in quantities of 1,000. 

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

RE: Awesome
By johnsonx on 9/27/2007 4:25:34 PM , Rating: 1
The memory ratio option is the same thing as Memory Limit. Please read my response to omnicrox, there's no point in me repeating the same info.

I know what I'm doing - I spent 3 months as an Anandtech forum junkie before even ordering my parts.

If the AT forum people told you about running your memory at 1:1 on K8, did they also tell you about Santa Claus, the Easter Bunny and the Tooth Fairy?

RE: Awesome
By clovell on 9/27/2007 4:37:12 PM , Rating: 3
I read your post. You can replace '1:1' with 'stock divider' in my post and still have the same information.

You've completely missed the point of my problem and have focused on how Overclocking my CPU doesn't require Overclocking my memory. I know that. I've never even tried to Overclock my memory - I have 0.05v of voltage headroom on my vDIMM - it's just not going to happen. I've always used a divider - I know how they work.

If you just want to be a troll and argue over symantics, that's fine, but let's call a spade a spade - you're not here to address my problem - you're here to try to impress us with your knowledge of overclocking and the K8 architecture - and to be completely honest, the only thing I'm impressed with is how much of a natural you are at being a jerk and ignoring the point.

RE: Awesome
By johnsonx on 9/27/2007 7:03:24 PM , Rating: 1
Fair enough criticism... I may have been feeling a bit boisterous and cocky. Honestly I was trying to help you, but perhaps I started amusing myself at the end there.

However, I still maintain there is no 'stock divider' for a K8 CPU. This isn't semantics. No one divider is more 'right' than another; the default is calculated by the BIOS at boot time by considering the memory limit setting, SPD values, and CPU multiplier. If you encounter stability problems using other dividers than the default, I suspect you have other settings incorrect. One thing even I, in my infinite mastery of all things overclocking, sometimes forget is setting a memory limit ratio also tells the BIOS to use the SPD timings for the lower memory speed. While I usually get away with using the DDR-333 timings at or a little above DDR-400 speeds, it's something to be mindful of.

This whole thread started out by you saying you couldn't get above 2.7Ghz with your multi-locked CPU because you can't overvolt your DIMMs enough; that clearly implied you were overclocking your memory. Now you're saying you never tried overclocking your memory (so then why do you need more vDIMM?). In an response post, you said you didn't want to use a divider because 1:1 is best, now you say you've always used a divider. It's hard to be of help when the story keeps changing.

I made my original comment only because your original post strongly suggested you didn't know about using a memory divider, and my follow-up comments were due to what I took as your false belief in a magic '1:1 memory ratio' fairy that bestows extra memory performance like so much pixie dust. You were threatening to drop $160 on a CPU upgrade just because it had an unlocked multiplier, and I attempted to save you from wasting that coin. I shall now martyr myself upon a cross and await your sword in my side...

(damn, I'm doing it again aren't I? bad johnsonx, BAD!)

RE: Awesome
By clovell on 9/28/2007 3:47:15 PM , Rating: 2
Well, locking in the stock settings of my memory at its rated 5-5-5-15 @ 2T with the HT multiplier at 4x and the HT bus at 240 MHz and my PCIe bus locked at 100MHz while leaving my CPU multiplier at 10x and running a memory setting of 667 MHz (effectively putting the memory at 800 MHz) while giving the memory 1.95v (an extra 0.05v - for good measure) still gives errors on Orthos Dual Core Blend Stress test in under 10 minutes. I guess I could have got some bad memory when I last upgraded, but once I set everything to auto, I'm 24 hours stable on Orthos DC Blend, so it doesn't seem like bad memory to me...

I'm not sure about magic fairies and performance, but the stock settings are the only thing that seem to give me stability right now. I know that contradicts what you're telling me about the architecture, but that's just my experience with the thing.

I may not have made a lot of sense earlier. Right now, the only way I can get stability is at stock, so I'm using the default/stock/1:1 divider and leaving the HT bus at 200 MHz. If you've got any ideas, though, I can give em a shot.

RE: Awesome
By johnsonx on 9/29/2007 7:32:50 PM , Rating: 2
While it's not all that much of an overclock, is it possible your CPU just isn't stable at 2.4Ghz? Certainly there must be at least a few AMD CPU's out there that are speed binned low because they truly couldn't cut it at any higher clock speed. Most people get much higher with their AM2's, but that doesn't mean everyone does.

Or are you saying you can run stable at 2.4Ghz with the memory overclocked by leaving the memory ratio setting at Auto, but unstable at 2.4Ghz with the memory at stock speed using a memory ratio?

If the second one is what you mean, then I'm stumped. Without having the system to look at personally, I guess I have little else to offer.

RE: Awesome
By clovell on 10/1/2007 11:02:36 AM , Rating: 2
When I first built, my CPU would max out at 2745 MHz stable. That's dropping the memory speed as low as I could using a divider with HT bus set to 305 and a 9x multiplier. I scaled it back to 2700 MHz for good measure, and it was stable for awhile. I can't remember exactly what did it, but I started getting some instability, so I backed off the OC a bit and came down to 2400MHz.

And that was fine, stable - great. But then I add another 2x1 GB of memory (same kind as the 2x512 MB I already had), and I start getting some stability issues that I can only get past by setting everything to auto. Which is weird, because if I lock in the stock settings, I still get some problems.

I may experiment with different BIOS revisions; I think that's what caused me to have to scale back from 2.7 to 2.4 GHz.

RE: Awesome
By Socaddict on 10/1/2007 4:39:10 AM , Rating: 2
hold on, your RAM is rated at 800mhz, and your trying to shove 1.95V through it? I thought almost all 800mhz RAM was rated at 2.0V? That would most likely be the problem.

RE: Awesome
By clovell on 10/1/2007 10:55:43 AM , Rating: 2
Corsair XMS2 PC6400 is rated at 1.90v

"And boy have we patented it!" -- Steve Jobs, Macworld 2007
Related Articles

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki