<div dir="ltr"><div dir="ltr"><div dir="ltr"><div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><span style="font-family:Arial,Helvetica,sans-serif">On Thu, Jan 1, 2026 at 11:03 AM <</span><a href="mailto:spike@rainier66.com" target="_blank" style="font-family:Arial,Helvetica,sans-serif">spike@rainier66.com</a><span style="font-family:Arial,Helvetica,sans-serif">> wrote:</span></div></div><div class="gmail_quote"><div dir="ltr" class="gmail_attr"><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div><div lang="EN-US"><div><div><p class="MsoNormal"><i><font face="georgia, serif" size="4"><span class="gmail_default" style="font-family:arial,helvetica,sans-serif">> </span>We tend to imagine away heat control and radiation effects as something that is a low level engineering problem, but we have had microprocessors in space for a long time and neither of these problems have been solved for what we think of as modern processors. If we imagine a high end modern GPU we can estimate a power consumption of about a kW. That’s nearly a square meter of solar cells to power it and we still haven’t gotten to how to extract the heat from that processor. </font></i></p></div></div></div></div></blockquote><div><br></div><font size="4" face="tahoma, sans-serif"><b>If a one square meter solar cell powers a GPU in space then a one square meter heat radiator angled at right angles to the solar cell panel should be <u style="">more than enough</u> to cool th<span class="gmail_default" style="">at</span> GPU down because <span class="gmail_default" style="">the </span>radiator has access to a very good heat sink that is at only 2.7° kelvin, empty space. </b></font><div><br></div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div><div lang="EN-US"><div><div><p class="MsoNormal"><font size="4" style="" face="georgia, serif"><i style=""><span class="gmail_default" style="font-family:arial,helvetica,sans-serif">> </span>The radiation is a big deal too: It causes SEUs or Single Event Upsets from cosmic rays hitting individual components. This usually necessitates larger components, limiting compute power in space.</i></font></p></div></div></div></div></blockquote><div><br></div><font size="4" face="tahoma, sans-serif"><b>Communication satellites are in geostationary orbits 36,000 km high and that puts them right in the middle of the outer van Allen radiation belt, and yet the average lifetime of a communication satellite is about 15 years, and some of them last for 25. Because of atmospheric drag a data center satellite couldn't be in low earth orbit, but it would have no need to be in a geostationary orbit either; if it was in a polar orbit of 8000 km atmospheric drag wouldn't be a problem, its solar cells would be in constant sunshine, and it would receive considerably less radiation than communication satellites do because it would be above the inner van Allen radiation belt and below the outer van Allen radiation belt. </b></font></div><div class="gmail_quote"><font size="4" face="tahoma, sans-serif"><b><br></b></font></div><div class="gmail_quote"><font size="4" face="tahoma, sans-serif"><b>And a data satellite wouldn't need to last for 15 years because the GPUs in it would become obsolescent in about 3 years and after 5 years or less they would be downright obsolete, so they would need to be replaced anyway. But the solar cells and the satellite's cooling and communication systems would still be OK.</b></font></div><div class="gmail_quote"><font size="4" face="tahoma, sans-serif"><b><br></b></font></div><div class="gmail_quote"><font size="4" face="tahoma, sans-serif"><b><span class="gmail_default" style="">John K Clark</span><br></b></font><div><br></div><div><br></div></div></div>
</div>