Athlon 64 X2 5000+ Socket AM2 Review

[epicstruggle]

http://www.hardwaresecrets.com/article/328
first review

epic

 

[arjan de lumens]

Anandtech has a review up as well.

It's a bit sobering to see just how little the extra bandwidth is able to help the processor (mostly about ~0-3%, even though the bandwidth is doubled and the latency appears to be the same as before) . Also, the fact that the memory controller runs at 742Mhz instead of a clean 800, and appears to be unable to take advantage of more than ~65% of theoretical bandwidth indicate some design immaturity still.

 

[epicstruggle]

Anandtech has a review up as well.

AMD does have one last trick up its sleeve before the end of the year, and you will hear about it in June. It's not K8L and it's not going to affect the majority of people, but it is an interesting stop gap solution for the high end in 2006...

I wonder what they are refering to here. Anyone have a clue?

 

[epicstruggle]

Another review:
http://enthusiast.hardocp.com/article.html?art=MTA2NSwxLCxoZW50aHVzaWFzdA==

 

[Gubbi]

It's a bit sobering to see just how little the extra bandwidth is able to help the processor (mostly about ~0-3%, even though the bandwidth is doubled and the latency appears to be the same as before) . Also, the fact that the memory controller runs at 742Mhz instead of a clean 800, and appears to be unable to take advantage of more than ~65% of theoretical bandwidth indicate some design immaturity still.

Or maybe the majority of applications don't need the bandwidth.

Edit: HL2 water hazard level has a D$ (L1) hitrate of 98.6%, L2 serves 2/3 of L1 misses -> very little bandwidth is actually used.

I see the release of the AM2 socket now as a mean to get the system infrastructure in place for quad core late 06/early 07

Cheers

 

[Geo]

NV generally seems to have a winner on their hands, tho chipset cooling seems to be ramping up.

But "if you can't hide it, then decorate it" is working for some, it looks like. VR-Zone, writing about the new ASUS M2N32-SLI Deluxe said:

VR-Zone brings you the exclusive review of the ASUS M2N32-SLI Deluxe with that beautiful heatpipe chipset cooler, based on the new and orgasmic Nvidia Nforce 590 SLI chipset...

Not sure I'd take a hand me down from the VR-Zone boys!

 

[Skrying]

I'm rather impressed. Leading up to the launch I thought that performance was actually going to be worse because of latency increases. But, no fear, seems like general boost all around, if it rather small.

 

[hoom]

Hmm, I think I'd rather have a DDR1 X2 5000+.
Would have liked them to have managed a 3ghz FX chip though.

 

[epicstruggle]

I'm rather impressed. Leading up to the launch I thought that performance was actually going to be worse because of latency increases. But, no fear, seems like general boost all around, if it rather small.

I was quite disappointed, I was mistaken in expecting double digit improvement across the board.

epic

 

[John Reynolds]

NV generally seems to have a winner on their hands, tho chipset cooling seems to be ramping up.

But "if you can't hide it, then decorate it" is working for some, it looks like. VR-Zone, writing about the new ASUS M2N32-SLI Deluxe said:



Not sure I'd take a hand me down from the VR-Zone boys!

Having owned previous nForce-based ASUS mobos, I love the M2N32. I pulled that beast out of its shipping box a few weeks ago, took one look at the heatpipes that zip-zag across its surface, and wet myself in glee. Well, ok, I didn't get that excited, but it beats some of the noisy-ass fans ASUS has used in the past on their NBs.

 

[pcchen]

I'm rather impressed. Leading up to the launch I thought that performance was actually going to be worse because of latency increases. But, no fear, seems like general boost all around, if it rather small.

Actually, DDR-2 800 has roughly the same latency as DDR 400. Of course, it's anticipated that DDR-2 800 won't help K8 much.

 

[hoom]

Yah, because the core memory chips in DDR2 800 are running the same speed as DDR 400.

 

[hoom]

Hmm, I think I'd rather have a DDR1 X2 5000+.

Well waddayaknow...http://www.theinquirer.net/?article=32014

 

[Moloch]

Wonder how long till they "fix" the memory controller? (see xbit labs...)

 

[hoom]

Awww poop http://www.theinquirer.net/?article=32063

 

[KimB]

Well waddayaknow...http://www.theinquirer.net/?article=32014

From what I've heard, there will be no new Socket 939 processors from AMD.

 

[Moloch]

here's techreports review..
http://www.techreport.com/reviews/2006q2/socket-am2/index.x?pg=1

 

[Slides]

Wonder how long till they "fix" the memory controller? (see xbit labs...)

Can you link to this please?

 

[Moloch]

Can you link to this please?

It was actually digit-life.
http://www.digit-life.com/articles2/mainboard/ddr2-800-am2.html
Scroll about half way down for ddr1 vs ddr2.

 

[hoom]

So, can someone explain this memory divider thing for me?

The bit where the 5000+ can't actually run DDR2 800 at 800 because of the int memory divider.
Does the same issue exist for the socket 939 CPUs with DDR?
ie does an FX60 run DDR 400 at 371mhz (2600/7)?!