Just to clarify, Tek had the first real-time oscilloscope up to 20GHz (DPO72004) which came out in Feb 2007.
Looks like you're right, Tek had indeed the first 20GHz real-time scope:
https://web.archive.org/web/20070222220943/http://www.tek.com:80/products/oscilloscopes/dpo70000_dsa70000/At the time LeCroy had 11GHz
Not true, at that time LeCroy already had the SDA18000 with 18GHz bandwidth and 60GSa/s sample rate, which actually came out 8 months earlier than the DPO72004 and at a time when the DPO70000 was still limited to 8Ghz and 25GSa/s:
https://web.archive.org/web/20060615065930/http://www.lecroy.com:80/https://web.archive.org/web/20070111030525/http://www.tek.com:80/products/oscilloscopes/dpo70000_dsa70000/index.htmlThe practical difference between 18Ghz BW and 20Ghz BW are pretty marginal (for the tasks these scopes are generally used for, the 20Ghz scope won't allow you to do more than a 18Ghz scope), and of course the individual BW steps vary between individual manufacturers (i.e. not all scope manufacturers will have a 20Ghz scope but may have a 22Ghz scope).
Interesting is however how long Tek got stuck at that level while others marched on:
LeCroy achieved 30GHz and 80GSa/s (WaveMaster 830zi) in Feb 2009:
https://web.archive.org/web/20090224035412/http://lecroy.com:80/tm/products/default.aspAnd 45Ghz and 120GSa/s in Dec 2010 (WaveMaster 845Zi-A and LabMaster 945Zi-A):
https://web.archive.org/web/20101225135320/http://www.lecroy.com:80/At the same time, Tek was still stuck with the DPO70000 at 20GHz and 50GSa/s:
https://web.archive.org/web/20101211163642/http://www.tek.com:80/products/oscilloscopes/And shortly before LeCroy went up to 60Ghz and 160GSa/s (LabMaster 10Zi) in Jan 2012:
https://web.archive.org/web/20120118135329/http://www.lecroy.com/Tek only made the jump from 20Ghz & 50GSa/s to 33Ghz and 100GSa/s:
https://web.archive.org/web/20120126030535/http://www.tek.com:80/oscilloscopeJuly 24, 2013: LeCroy demonstrates a lab version of its 100Ghz 240GSa/s scope
https://web.archive.org/web/20131219143511/http://teledynelecroy.com:80/100ghz/Which then became officially available (LabMaster 10zi) in Dec 2014
https://web.archive.org/web/20141021083059/http://teledynelecroy.com:80/oscilloscope/while Tek was still stuck with the old DPO70k platform at 33Ghz and 100GSa/s:
https://web.archive.org/web/20141225082629/http://www.tek.com:80/oscilloscope#alland Agilent had 13GHz
Correct, in 2007 Agilent were still stuck with the old DSO80k Series with up to 13Ghz (which also suffered from various hardware limitations):
https://web.archive.org/web/20070207230130/http://www.home.agilent.com:80/USeng/nav/-536902447.0/pc.htmlHowever this DSO80k was actually introduced in Dec 2004:
https://web.archive.org/web/20041204090657/http://we.home.agilent.com:80/USeng/nav/-536895900.0/pc.htmlThe successor DSO90k came out in Feb 2008, still limited to 13GHz and 40GSa/s but now without the limitations of the old platform and better overall performance:
https://d3fdwrtpsinh7j.cloudfront.net/Docs/datasheet/agilent_dso90000a.pdf(no archive.org link as by then Agilent has started to refuse them scanning their website)
In 2010, while Tek was still sitting at 20Ghz/50GSa/s, Agilent introduced the DSO90k X-Series with BW up to 33GHz and 80GSa/s, and in Nov 2012 the DSO90k Q-Series with up to 63GHz and 160GSa/s (Tek only made the jump to 33GHz and 100GSa/s a year before).
Today Keysight offers a 63GHz real-time oscilloscope, Tektronix offers a 70GHz and LeCroy offers a 100GHz.
Keysight has stuck at 63Ghz and invested in improving its scope architecture instead, probably because tasks requiring more than 60Ghz are truly niche applications, and I can't say that this was a bad decision.
However, what the above list shows is that, for high BW scopes, stating that manufacturer A had the first X Ghz scope is pretty useless when everyone of them uses different BW steppings. What's important is the overall package, i.e. the BW, and even more so the sample rate (or better the sample rate to BW ratio), the processing backend, and last but not least, the available analysis and measurement options. And unfortunately the whole package is something Tek has been lacking for a very long time.
The MSO5x seems to be a solid scope (although in search of a target market) and gives hope that things in Tek improve, but it's by no far something that "changes everything". God knows we can use more serious competition, would be great if Tek could step up and become one again. Using its new ADC in a entry level scope with strong processing, decent maths and analysis tools and competitive pricing, as suggested in the other MSO5 thread, might be an easier start.