WEBVTT NOTE This file was exported by MacCaption version 7.0.06 to comply with the WebVTT specification dated March 27, 2017. 00:00:00.167 --> 00:00:07.174 align:center line:-1 position:50% size:59% What are some other gotchas that people run into with new USB4 testing? 00:00:07.174 --> 00:00:11.278 align:center line:-1 position:50% size:39% We talked about the LTSSM, which was one clear gotcha. 00:00:11.278 --> 00:00:14.381 align:center line:-1 position:50% size:43% The second was around the fact that you need a controller. 00:00:14.381 --> 00:00:16.283 align:center line:-1 position:50% size:31% That's the next gotcha. 00:00:16.283 --> 00:00:26.159 align:center line:-1 position:50% size:67% I think the one that comes up most often is around the notion of how much testing needs to be done. 00:00:26.159 --> 00:00:38.538 align:center line:-1 position:50% size:67% Even though we are talking about USB4, a typical USB4 host has to transmit not just USB4. 00:00:38.538 --> 00:00:44.478 align:center line:-1 position:50% size:58% It has to transmit USB4 at 10 and 20 Gbps. 00:00:44.478 --> 00:00:49.917 align:center line:-1 position:50% size:40% Potentially, on that same port, it also has to transmit 00:00:49.917 --> 00:00:54.454 align:center line:-1 position:50% size:64% Thunderbolt rates at 10.3125 and 20.625 Gbps. 00:00:54.454 --> 00:01:01.828 align:center line:-1 position:50% size:60% It also must send USB 3.2 at 5 and 10 Gbps, 00:01:01.828 --> 00:01:06.266 align:center line:-1 position:50% size:41% it must do USB 2 at 480 Mbps, 00:01:06.266 --> 00:01:14.241 align:center line:-1 position:50% size:57% and it will need to do DP (DisplayPort) Alternate Mode potentially up to 8.1 Gbps, 00:01:14.241 --> 00:01:19.913 align:center line:-1 position:50% size:41% all over the same Type-C port. 00:01:19.913 --> 00:01:22.316 align:center line:-1 position:50% size:17% It's so much. 00:01:22.316 --> 00:01:23.951 align:center line:-1 position:50% size:18% It is, it's a lot. 00:01:23.951 --> 00:01:30.524 align:center line:-1 position:50% size:39% "Jit," they said, "if I can show that this port in my silicon 00:01:30.524 --> 00:01:36.096 align:center line:-1 position:50% size:48% can transmit 20 Gbps or 20.6 Gbps, 00:01:36.096 --> 00:01:42.235 align:center line:-1 position:50% size:52% why would I possibly need to test all the other standards I just reeled off? 00:01:42.235 --> 00:01:47.674 align:center line:-1 position:50% size:61% Because if I can put the fastest thing through, we're done." 00:01:47.674 --> 00:01:49.776 align:center line:-1 position:50% size:35% It's definitely a valid input. 00:01:49.776 --> 00:01:52.045 align:center line:-1 position:50% size:43% Customers ask that all the time. 00:01:52.045 --> 00:01:58.852 align:center line:-1 position:50% size:57% The answer is you can't, and the reason is because the use cases are different. 00:01:58.852 --> 00:02:03.457 align:center line:-1 position:50% size:61% For example, using just USB4 as an example, 00:02:03.457 --> 00:02:11.932 align:center line:-1 position:50% size:51% in the 20 Gbps use case, the customer could use a 0.8 m cable, 00:02:11.932 --> 00:02:19.373 align:center line:-1 position:50% size:49% but at 10 Gbps, the customer could use a 2 m cable. 00:02:19.373 --> 00:02:23.944 align:center line:-1 position:50% size:54% It's not that the bit rates are slow or fast. 00:02:23.944 --> 00:02:26.813 align:center line:-1 position:50% size:37% It's just that every use case is a little bit different. 00:02:26.813 --> 00:02:28.682 align:center line:-1 position:50% size:22% I mentioned DP. 00:02:28.682 --> 00:02:36.089 align:center line:-1 position:50% size:49% In most use cases, a 0.8 m DP cable would just be too short to connect 00:02:36.089 --> 00:02:40.160 align:center line:-1 position:50% size:36% a customer's laptop to a large movable display. 00:02:40.160 --> 00:02:47.067 align:center line:-1 position:50% size:55% I'm working from home, and my DP cable is probably a good 4 or 5 feet 00:02:47.067 --> 00:02:52.406 align:center line:-1 position:50% size:41% just to be able to reach around to my other monitor back here. 00:02:52.406 --> 00:02:55.976 align:center line:-1 position:50% size:41% The net result is, unfortunately, 00:02:55.976 --> 00:03:01.014 align:center line:-1 position:50% size:59% we have to test all these standards, which you mentioned there are lots of them, 00:03:01.014 --> 00:03:05.419 align:center line:-1 position:50% size:44% because the use case of the channel model is different. 00:03:05.419 --> 00:03:11.091 align:center line:-1 position:50% size:55% That's one of the traps, which is because the Type-C is universal, 00:03:11.091 --> 00:03:16.797 align:center line:-1 position:50% size:60% truly now, with all these standards that all have different cables and connectors, 00:03:16.797 --> 00:03:21.501 align:center line:-1 position:50% size:57% we have to test for all these potentially different use cases. 00:03:21.501 --> 00:03:28.909 align:center line:-1 position:50% size:71% That's one of the most common pitfalls that we run into for customers who are doing USB4 work. 00:03:28.909 --> 00:03:32.179 align:center line:-1 position:50% size:67% Are customers only testing for the highest speeds, 00:03:32.179 --> 00:03:37.084 align:center line:-1 position:50% size:55% and then they go to show up at a plugfest and are failing other routes, 00:03:37.084 --> 00:03:40.854 align:center line:-1 position:50% size:47% or is this just something people need to know to do from the get-go 00:03:40.854 --> 00:03:44.958 align:center line:-1 position:50% size:37% when they're testing USB4, to test everything? 00:03:44.958 --> 00:03:46.359 align:center line:-1 position:50% size:28% It's a great comment. 00:03:46.359 --> 00:03:53.567 align:center line:-1 position:50% size:55% I think, in general, customers understand they need to do all these rates, 00:03:53.567 --> 00:03:58.238 align:center line:-1 position:50% size:39% but I think, like all of us, there's a lot of time pressure. 00:03:58.238 --> 00:04:04.244 align:center line:-1 position:50% size:54% Occasionally, we see a customer say-- let's just use USB4 only as an example-- 00:04:04.244 --> 00:04:10.283 align:center line:-1 position:50% size:56% "I've tested USB4 at 20 Gbps, so I think it should work fine for 10 Gbps," 00:04:10.283 --> 00:04:14.988 align:center line:-1 position:50% size:59% but the problem is the use case for 10 Gbps sometimes fails more than 20 Gbps, 00:04:14.988 --> 00:04:20.460 align:center line:-1 position:50% size:51% and the reason is because it goes from a 0.8 m cable which is very lossy 00:04:20.460 --> 00:04:25.932 align:center line:-1 position:50% size:54% to a 2 m cable which is even more lossy, so sometimes we see customers 00:04:25.932 --> 00:04:33.473 align:center line:-1 position:50% size:50% who will pass the 20 Gbps USB4 rate but will fail the USB4 10 Gbps rate 00:04:33.473 --> 00:04:36.143 align:center line:-1 position:50% size:60% because they didn't take into account the fact 00:04:36.143 --> 00:04:41.515 align:center line:-1 position:50% size:40% that you may have to put in a 2 m cable into the use case. 00:04:41.515 --> 00:04:43.917 align:center line:-1 position:50% size:37% That happens occasionally. 00:04:43.917 --> 00:04:44.951 align:center line:-1 position:50% size:17% Fascinating. 00:04:44.951 --> 00:04:47.988 align:center line:-1 position:50% size:46% That would be really disappointing to have that happen to me. 00:04:47.988 --> 00:04:50.390 align:center line:-1 position:50% size:40% Putting myself in those shoes. 00:04:50.390 --> 00:04:53.760 align:center line:-1 position:50% size:41% You bring a good point though. 00:04:53.760 --> 00:04:59.032 align:center line:-1 position:50% size:52% When customers come in to be tested, they're very proud of this product. 00:04:59.032 --> 00:05:02.936 align:center line:-1 position:50% size:40% They worked for months on it, they believe the silicon works, 00:05:02.936 --> 00:05:05.405 align:center line:-1 position:50% size:50% their layout works, the system works, 00:05:05.405 --> 00:05:10.944 align:center line:-1 position:50% size:66% and to show up and have some of these tests fail 00:05:10.944 --> 00:05:14.047 align:center line:-1 position:50% size:62% is very disappointing because now they have to go back and retool. 00:05:14.047 --> 00:05:20.921 align:center line:-1 position:50% size:50% Sometimes it's simple, maybe it's just a simple capacitor that was too large 00:05:20.921 --> 00:05:25.392 align:center line:-1 position:50% size:44% or maybe a simple board change or a software change, 00:05:25.392 --> 00:05:28.161 align:center line:-1 position:50% size:48% but, for example, if it's in the silicon, 00:05:28.161 --> 00:05:31.932 align:center line:-1 position:50% size:54% that becomes much more complicated because now they have to do a big turn. 00:05:31.932 --> 00:05:38.939 align:center line:-1 position:50% size:45% It can be fairly simple or it could be massively complex. 00:05:38.939 --> 00:05:45.845 align:center line:-1 position:50% size:51% Testing early and testing often sometimes can be a good thing to do.