Download the Tom's Hardware App from the App Store
The reference for current tech news
Yes No

What's A Stepping?

by

As already described, a processor stepping is simply a new hardware revision, but it does not involve significant changes or even a redesign of the micro architecture. There can be feature enhancements or other improvements, but microchip manufacturers mainly look at their products to fix technical bugs - these are referred to as errata and most are typically published by AMD and Intel. A so-called die shrink - e.g. when transitioning an existing product from a 90-nm process to a smaller 65-nm process - isn’t referred to as a stepping. However, shrinking the device can be also be used to apply modifications, such as adding larger cache memory capacities or enhanced instruction sets.

While a new stepping has been sold using the same product and model numbers by AMD and Intel in the past, a die shrink can involve sufficient modifications to justify creating new model numbers for the devices. Intel’s Core 2 Duo is a good example. The device’s 45-nm versions with 6 and 3 MB L2 cache are called the E8000 series, while the E6000 with 4 or 2 MB L2 cache is based on the 65-nm process. In contrast, AMD has kept the model number 5000 for all 90-nm and 65-nm versions.

Steppings are not only a good opportunity to fix known issues and to introduce improvements on the feature side, but they are used to deploy what the chipmaker has learned over time in its fabs. The production process can often be improved by applying minimal changes, or it can include changing an entire part of the manufacturing process. Adjustments in manufacturing can lead to improved yield rates. They can also eliminate - or at least diffuse - so-called hot spots or allow for the manufacturer to make a given process more environmentally friendly. Clearly, new steppings do not always have to lead to better features or noticeable improvements, while business reasons and other considerations are always part of the equation.

CPU-Z is a useful tool to learn more about processor revisions and stepping histories. Windows typically doesn’t disclose this information, as it is of little relevance for the majority of PC users.

AMD Socket AM2 Steppings And Model Numbers

AMD has had two major steppings for both the 90-nm and the 65-nm versions of the Athlon 64 X2 on socket AM2. When it was first introduced, there was the 90-nm F2 stepping, which was followed by the 90-nm F1. After the launch of AMD’s 65-nm Athlon 64 X2 processors AMD followed up with its G2 stepping a few months later.

Processor AMD AthlonTM 64 X2 Dual-Core
Model 5000
OPN Tray ADO5000IAA5DO ADO5000IAA5DD ADA5000IAA5CU ADO5000IAA6CZ
OPN Pib ADO5000DOBOX ADO5000DDBOX ADA5000CUBOX ADO5000CZBOX
Operating Mode 32 Bit Yes
Operating Mode 64 Bit Yes
Revision G2 G1 F2 F3
Core Speed (MHz) 2600
System Bus Speed (MHz) 2000
Voltages 1.325/1.35/1.375V 1.25V/1.35V 1.30V/1.35V 1.20V/1.25V
Max Temps (C) 55-68’C 55-72 55-70 55-72
Wattage 65 W 89 W 65 W
L1 Cache Size (kB) 128
L1 Cache Count 2
L2 Cache Size (kB) 512
L2 Cache Count 2
L3 Cache Size (kB) 0
CMOS 65nm SOI 90nm SOI
Share:
2
Comments
X
Submit

Comments
Read the comments on the forums
Flakes 24/03/2008 13:28
Hide
-0+

you could of chucked in a core2duo so that we could see the difference between intel and the new stepping.

sdack 30/03/2008 14:42
Hide
-0+

You are reading the bottom of a coffee cup.

You say PCMark05 results show a significant difference and believe it to explain something to you. However, just beneath the PCMark05 result you show the results of another synthetic benchmark - SiSoft Sandra - and you ignore it. Can you explain why the G2 sticks out from all the others?

You then use the single F3 stepping to claim that the 65nm process makes no difference in power consumption compared to the 90nm process, and again ignore a detail. Is the F3 not using much more power than the other steppings and was it not manufactured in a 90nm process?

Last but not least, you are not giving out any information about the variances of the results, which however would have made them conclusive. When some results are inconclusive because they are all the same it does not mean that it makes difference in another benchmark automatically conclusive. To show that results are conclusive you need to look at the variances, too, when you see that the results are close together.

In short, you have made a large effort for saying nothing conclusive. If you are sure about the difference between the steppings (i.e. with the increased L2 latency) you should be able to make it much more visible since you know what to look for. If it turns out to be impossible then you should search for the reason why it is impossible instead of ignoring important details.

Best offers

Newsletters


OK