NVIDIA's NV17M: The warrior without a name

Early in the morning of this first day of Comdex we were treated to a breakfast and a new product launch by NVIDIA. As you've undoubtedly heard by now the new product being introduced is the NV17M core intended to replace the GeForce2 Go as NVIDIA's solution for the mobile gamer.

The first thing you can tell by the name alone is that the NV17M is a mobile derivative of a desktop NV17 part. You may have also heard the NV17 referred to in the past as the GeForce3 MX meaning that the NV17M is the mobile GeForce3 MX. NVIDIA has yet to announce an official name for this part which is ok since the chip is actually due out in February.

NVIDIA's first MX product was the GeForce2 MX, released almost a year after the original GeForce 256, and offered performance on par with the original GeForce 256. This NV17 core is no different as it is supposed to offer performance greater than the GeForce3 GPU that was released about a year ago. What's even more impressive is that this power is being brought down to the mobile level, very impressive indeed.

Like the original GeForce2 MX, the NV17/NV17M core only features two rendering pipelines compared to the four pipelines on the GeForce3 GPU. The 0.15-micron chip runs at up to 250MHz when in a laptop configuration but depending on the battery saving mode it is placed in, the clock speed can be throttled down considerably.

Since NVIDIA had a demo of the NV17M core running Chameleon and many of the other pixel/vertex shader demos from the GeForce3 launch we can only assume (until our meeting on Wednesday that is) that the NV17M does not lose any of its older brother's pixel/vertex shader functionality. As soon as we get confirmation or hear otherwise we will report it.

As you know, memory bandwidth is key to killer 3D performance. Like its predecessor, the NV17M has a very flexible memory subsystem; the design supports a 32-bit, 64-bit, or 128-bit DDR memory bus with memory running at up to 250MHz. Remember that this 250MHz figure is just a recommendation by NVIDIA and not all OEMs will adhere to that frequency for shipping notebooks.

Index NV17M continued
Comments Locked

0 Comments

View All Comments

Log in

Don't have an account? Sign up now