Fastest Cyrix MediaGX you've seen?

[razor1000]

I'm talking actual Made-By-Cyrix, not NS or AMD, CPGA not BGA.

Supposedly they made a 300Mhz, but the highest I've ever seen is 266 (I have one!)

In case anyone is interested, I found a place that actually sells the boards for these processors. No, they DO NOT work on standard Socket 7 boards (I think most of you knew that, tho.)

[gshv]

The fastest I have is 266 MHz. The fastest I've seen is 366 MHz:



And now you've seen it too I took this picture from http://www.yjfy.com/ website. It looks like the owner of that site has a collection of CPUs and other computer hardware, take a look at the links on the right side of the page.

Genna

[razor1000]

Ohhhhhhhhh, 366! *drool*!!!

Isn't it strange that the 366 thru 180 parts are all the same core voltage?
That isn't very common, I don't think...

[gshv]

[CPUShack]

I dont think they had any die shrinks with the GX.
2.9Volts should have been plenty at the time for that kind of scalling.
similar chips were running on 2V

[Guest]

Do you think the black slab on the bottom of the cpu is a cover for the core? (I wonder how to get it off )

I've been reading the threads on de-soldering the covers of 486s and Pentiums... and it seems like those covers vary in size depending on the actual core size. Considering die-size will decrease with a process reduction, I would guess if the covers are smaller as the speed increases there was probably a shrink.

Since I don't think the covers actually decrease in size, I would guess they just binned the CPUs depending on clock-speed. I'm thinking with all the other junk that's included on the core, (pci controller, memory controller, audio, video) certain batches of silicon were more finicky than others.

It's just really strange that they would ALL be the same core voltage. That would mean that Cyrix picked a single voltage for all the CPUs to run at and then binned them by stability at a certain clock.

[CPUShack]

For something like the GX what they will do is determine a voltage that gives them the bins they want. 2.9Volts is what it took get x amount of 200MHz chips or Y amount of 233s. Anything above that is extra. One of the reasons you dont see many 366s. There wasnt much demand for them. If there was a large demand for them Cyrix would have either increased the Voltage to increase the yields of that speed grade, or shrunk the process to increase the yields.

The GX is an embedded chip mainly so lower voltage is very nice.

The black cover on the bottom is glued on ceramic. It breaks very easy and will expose the core.

[razor1000]

Oooo! I want to see the core of one of these chips! I bet it looks really cool!

Has anyone ever (personally) removed the cover on a MediaGX? Seen a picture of one with the cover removed?

[razor1000]

I just ran across this post:
http://www.cpu-world.com/forum/viewtopic.php?t=443

133Mhz 3.6v! I bet there was a die-shrink after this model!

[Simmayor]

Hi,

I found some new pictures here :
http://www.yjfy.com/yjlt/dispbbs.asp?boardID=1&ID=639






How about that!

Simmayor

[razor1000]

The post says that they are remarked Pentium 100s!

They show the back of the two top ones and they clearly have an Intel S-Spec.

Ack! Makes me wonder about the authenticity of that 366... hence the reason why I was wondering about the core voltage not changing. They might take the time to change the Mhz rating, but not bother with the voltage.

(Take a look at my avatar, I changed the Mhz and the DX-4 into a 9, but I didn't bother changing the core voltage or the little text at the bottom with the rating info....)

[razor1000]

I just noticed that the ebay auction that Genna mentioned that got me started thinking about this:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3463309010&category=3675

He notes his CPU has a 2.2v core voltage!

[gshv]

[razor1000]