The first generation of Opteron's. The 800 models are able to run in systems with eight (octa) CPU's. I have tested two of them in a MSI MS-9131 0B engineering sample motherboard and work great even despite they are not recognized correctly.

Unfortunatly I have only one DDR266 CL2.5 DDR-SDRAM ECC Registered memory module so the benchmarks of the 844 are not optimal.
> Read more

Just like this Opteron 844. They run great in a dual system together. > Read more

Just like this 2218 and it's brother but a 400MHz slower.

This CPU works great. Interesting note is that if I combine this 2214 with a 2216 or 2218 the clock frequency of both processors is reduced to the lowest clock frequency possible because both CPU's can handle that frequency. In this case they will run at 1GHz. > Read more

My first AMD Opteron engineering sample on socket 940. Does it work? No idea, I have no ways of testing it and assembling a system for only one CPU is a bit expensive for my taste .

This particular engineering sample has no information grooved into the heatspreader but rather like it's stamped on. As seen on the pictures this method is not very durable as the paint falls off. For those who want to know what was on the heatspreader:

AMD
AM8660595035
F11140
D45B800
Engineering Sample

There is not much information on the internet about this CPU. Unfortunatly I don't know the production date which makes it difficult to put it in a time frame. I suspect this is quite an early sample. The introduction of this CPU was in December 2004.

Based on the packaging of this CPU I think this is a 90nm model. The 130nm models used a ceramic packaging. > Read more

The same as the other Santarosa Opterons . > Read more

Almost the same as the other Santarosa Opterons. This model ends with 635 instead 631 (ZSW2400GAA635). I'm not sure what the difference is. > Read more

Unfortunately this CPU does not work. It never did when I got it and according to the heatspreader the CPU didn't work for a long time (or at all). The other socket F 'Santarosa' Opterons in my collection have scratched heatspreaders because they have been used (a lot) for testing hardware. Because this CPU doesn't have those scratches I suspect it was either DOA (Dead On Arrival) or got damaged quickly. > Read more

I obtained this CPU along with one nephew and four little brothers. All these CPU's are my first socket F samples so I didn't have any hardware to test it. Because I was interested whether the CPU's would work or not I bought suitable hardware. The result? A new server with four cores!

See this CPU-z screenshot while I was running in a test environment:


Currently I'm waiting for heatsinks so I can actually use two CPU's. For the moment I have only one CPU running with an AMD Athlon64 X2 6000+ heatsink/fan attached to it.

Little history
In the end of 2004 it was already known that the new AMD Opteron dual-core processors would fit in a socket with 1207 pins. Halfway 2005 this was confirmed and not much later the first pictures of socket F showed up.

The AMD Opteron 2218 you are looking at has been made in September 2005. This is almost one year before the official launch of the CPU on 15 August 2006. AMD had it's technology prepared long before it was actually used. According to some research some motherboard companies were already finished in June 2006. AMD was also willing to wait for Intel to ship the Xeon 'Woodcrest' in the end of June. In a summary: getting a CPU to the market takes some time . > Read more

The brother of this AMD Opteron 2218. They are made in the same week and this CPU is 27 CPU's older then it's brother.

Today both CPU's run in my server and present you this webpage. The server' host OS is Microsoft Windows Server 2008 Enterprise, the webserver runs on Ubuntu inside a virtual server. Down below you'll find a picture of the server running Windows Server 2008:


As seen on the photo's on the right the laser etched inscription is difficult to see. When seeing the CPU with the naked eye the inscription is also not easy to read. > Read more