NORMALLY I DO not bother with preliminary numbers, but sometimes you have to go with what you have. This time, the die size of the NVidia GF110/GTX580 is such a hot topic we will sacrifice a little precision to get you the info.
Short story, the die size of the GF110 is about 550mm^2, about 4% larger than the GF100/GTX480. These were hand done caliper measurements, not instrumented, so they will change once the thing is put under the TSA screener set to ‘dark tan’. We are confident enough in them to say that the GF110 is a larger chip.
.1mm is hard to measure by hand, and only made more unpleasant by the coating of underfill that hasn’t managed to glassify yet. Since the chip is around 23mm on a side, each .1mm of inaccuracy puts the die size off by over 2mm, hence our caution.
For all those who say it is a massively reworked GF100, or it can be much cheaper to produce, this die measurement proves otherwise. The GF110 is bug fixed GF100 with some low hanging fruit added on. Nvidia is calling this both a new architecture and a new series, which is a big step up from G92. *LE SIGH* S|A
Latest posts by Charlie Demerjian (see all)
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016
- Everspin and Globalfoundries team up for embedded ST-MRAM - Sep 15, 2016
- Intel’s Xpoint is pretty much broken - Sep 12, 2016
- ARM adds 2048-bit vectors to v8A with SVE - Sep 7, 2016