Tom's Hardware Forums » CPU & Components » CPUs » AMD's Athlon Stepping Improvements
 

AMD's Athlon Stepping Improvements

Add a reply



 Word :   Username :  
 
Bottom
Author
 Thread : AMD's Athlon Stepping Improvements
 
muk
Profile: member
More Information

While processor architectures are typically available for years, AMD and Intel typically offer improved steppings throughout a product's lifecycle. We looked at the history of the Athlon 64 X2 5000.
 
http://www.tomshardware.com/2008/0 [...] index.html

Related Pr oduct
Register or log in to remove.

Profile: old hand
More Information

Interesting article.  What I'd really like to see though is the difference between the old chips with 2x1MB cache and the new ones with 2x512KB cache.  Also found it interesting that the power consumption ddin't improve going from 90nm to 65nm.  I guess AMD had really mastered that 90nm process.


---------------
Gaming: FX-60 @ 2.81GHz (x14, 1.375, 90nm) > A8N-SLI Deluxe > Asus 4850 - 625/1986 > 2GB Corsair XMS 400MHz  
2-3-3-6-1T
HTPC/Light Gaming: X2 5400+ 2.8GHz Brisbane > Gigabyte 780G MATX - 900MHz Core > 2GB Corsair XMS2 800MHz  
4-4-4-12-2T
Profile: addict
More Information

or their 65nm doesnt work 'properly' :/

I have nothing witty to say.
Profile: old hand
More Information

No overclocking or temperature comparison?  I was interested in that, and it's the only reason I started reading it in the first place.  I only hope you include these things in the Intel article - since they are so crucial to the enthusiast community (which makes up the majority of the daily readers of this site).
 
Edit: also no mention of the difference in operating voltage between 90 & 65nm parts (1.184/1.184 & 1.328/1.344, respectively from the CPU-Z shots).  Why the .14-.16 voltage bump when you're going to a smaller manufacturing tech?  How did the 90nm stepping cut so much power when it's running at the same voltage as the original?


Message edited by KyleSTL on 03-20-2008 at 11:11:47 PM

---------------
Lian-Li PC-7B | XClio Greatpower 550W | P4 3.2 Prescott SL7E5 | Scythe Ninja
2GB DDR400 Corsair VS (4*512) | eVGA nVidia GF 7600GS AGP vmod 1.46/1.91 OCd 740/910
WD 120GB & 250GB PATA & WD 640GB SATA (on PCI SATA card LOL)
WinXP MCE 2004
Profile: Ancient Poster
More Information

Well power consumption pretty much tells you about temperature.
 
I was quite surprised that the 65nm process did not help with power consumption at all.


---------------
If its good in theory but not in practice,
its not good theory.
Factboy
Profile: Ancient Poster
More Information

So basically the 65nm saves some power at idle.
 
What I think would have made this article worthwhile would be to overclock them.  I hope they do so, because I have a feeling the 90nm products would clock far higher.


---------------
http://promotions.newegg.com/Intel/2Over3/478x88.jpg
Profile: addict
More Information

I wouldnt hold my breathe on overclock giving a huge preformance gain until they get away from there current architecture, and bring something out that is really worth buying.

Profile: member
More Information

Why on earth would you not talk about overclocking in an article about stepping? I didn't even know what steppings were until I started overclocking.

Warranty Offender
Profile: enthusiast
More Information

The only upside I can see is the use of 1066 RAM. No power changes, 65nm, and a loss of Cache.
If THG has these procs, OC em till they beg for mercy! Then tell me about temps, and voltage.

Profile: member
More Information

Hmm are they faster than Intels models? Dont think so. Thanks for playing. When is amd gonna get their act together? Geez Athlon 64's used to rock...


---------------
Laptop Gateway P6831 FX-- Intel Core 2 1.67 GHZ, 3 GIGS DDR2 Ram, 250 Gig SATA, 17 Inch LCD 1440x900, Nvidia 8800GTS 512 Meg Video, Vista Home Prem. 3Dmark06- 7032 marks.
Profile: member
More Information

amdfangirl wrote :

I reckon that an article on Athlon K8 vs Athlon K10 would have been sooo much more interesting, but I digress


 
http://www.tomshardware.com/2007/1 [...] age25.html

Profile: old hand
More Information

Yeah, I'm trying to decide whether to wait for K10 dual core or not.  I wish they would just release them already.


---------------
Gaming: FX-60 @ 2.81GHz (x14, 1.375, 90nm) > A8N-SLI Deluxe > Asus 4850 - 625/1986 > 2GB Corsair XMS 400MHz  
2-3-3-6-1T
HTPC/Light Gaming: X2 5400+ 2.8GHz Brisbane > Gigabyte 780G MATX - 900MHz Core > 2GB Corsair XMS2 800MHz  
4-4-4-12-2T
Profile: old hand
More Information

I plan on building in May so I don't think I can get the 45nm parts.  If none of the 65nm Phenoms look good, I'll just get a 45W 4850e (2.5GHz).


---------------
Gaming: FX-60 @ 2.81GHz (x14, 1.375, 90nm) > A8N-SLI Deluxe > Asus 4850 - 625/1986 > 2GB Corsair XMS 400MHz  
2-3-3-6-1T
HTPC/Light Gaming: X2 5400+ 2.8GHz Brisbane > Gigabyte 780G MATX - 900MHz Core > 2GB Corsair XMS2 800MHz  
4-4-4-12-2T
Profile: enthusiast
More Information

Is it just me or each 65nm stepping requires higher voltage for the same speed? That is just weird.

Profile: stranger
More Information


 
This article misses on the most obvious question - voltage!!!!  Why are the 65nm parts not lower voltage - and what are the procs capable of - undervolt / overclock????
 
Any idiot can tell you that an x2-5000 will all be the same!!!  Thats why they have the same model number!!  My windsor 90nm F2 stepping dual core runs stable with a slight overclock at 1.10 volts core.  Is 65 any better??  No idea after reading this article.

Profile: addict
More Information

Then do some investigating on your own, dont be lazy and try to get other people to do it for you, that way your not a parrot, instead you will be informed.


---------------
Q6600@3.2g, 4 gigs 2x2 ADATA Pc 6400, XFX 8800 GTS 512 G92, GIGABITE GA-P35-DS3L, ARCTIC COOLING FREEZER PRO 7, Antec 900 case, Antec 500 earthwatt p.s. 27.5" Hannspreee monitor (oh yeah!!)

Go to:
Add a reply
  Tom's Hardware Forums » CPU & Components » CPUs » AMD's Athlon Stepping Improvements
 

Google Ads
Ad