1 00:00:04,440 --> 00:00:12,360 Speaker 1: Welcome to tech Stuff, a production from iHeartRadio. Hey there, 2 00:00:12,400 --> 00:00:15,560 Speaker 1: and welcome to tech Stuff. I'm your host Jonathan Strickland. 3 00:00:15,640 --> 00:00:18,840 Speaker 1: I'm an executive producer with iHeartRadio. And how the tech 4 00:00:18,920 --> 00:00:21,439 Speaker 1: are you? You know? Over the last few years, there's 5 00:00:21,480 --> 00:00:24,880 Speaker 1: been a lot of conversation around microchips in general, and 6 00:00:25,200 --> 00:00:30,440 Speaker 1: CPUs and GPUs in particular. The pandemic led to bottlenecks 7 00:00:30,520 --> 00:00:33,680 Speaker 1: in the supply chain. Manufacturing facilities had to shut down 8 00:00:33,720 --> 00:00:38,640 Speaker 1: multiple times, particularly in China, and the initial skyrocketing value 9 00:00:38,640 --> 00:00:43,680 Speaker 1: of cryptocurrencies all had an effect on microchip supply. Meanwhile, 10 00:00:44,080 --> 00:00:47,920 Speaker 1: multiple countries, including the United States, started looking into ways 11 00:00:47,920 --> 00:00:51,239 Speaker 1: to shift away from depending so heavily on China for 12 00:00:51,360 --> 00:00:55,160 Speaker 1: chip fabrication. And when we talk about chips like CPUs, 13 00:00:55,720 --> 00:00:59,840 Speaker 1: we often will focus on two major factors. So first 14 00:01:00,320 --> 00:01:04,480 Speaker 1: is the process used to actually fabricate discrete components on 15 00:01:04,520 --> 00:01:07,640 Speaker 1: the chip. We typically reference this in terms of a 16 00:01:07,840 --> 00:01:13,160 Speaker 1: nanometer process, and the fewer nanometers represents more advanced processes, 17 00:01:13,200 --> 00:01:19,080 Speaker 1: so you're working backward in numbers. Secondly, is the chip's architecture, 18 00:01:19,360 --> 00:01:21,200 Speaker 1: and that's really what we're going to focus on in 19 00:01:21,240 --> 00:01:23,039 Speaker 1: this episode. But in order to do that, we also 20 00:01:23,080 --> 00:01:25,560 Speaker 1: have to talk about the other stuff. So a quick 21 00:01:25,600 --> 00:01:29,720 Speaker 1: word on the fabrication process part. You might hear that 22 00:01:29,880 --> 00:01:34,039 Speaker 1: company used a seven nanometer process or a five or 23 00:01:34,080 --> 00:01:38,240 Speaker 1: even a three nanometer process to make the chip. And 24 00:01:38,440 --> 00:01:42,440 Speaker 1: you may know that a nanometer is one billionth of 25 00:01:42,680 --> 00:01:46,880 Speaker 1: a meter. It's one scale up from the atomic scale. 26 00:01:47,319 --> 00:01:51,040 Speaker 1: So a typical human hair measures between eighty thousand and 27 00:01:51,200 --> 00:01:55,800 Speaker 1: one hundred thousand nanometers thick, as in, if you measured 28 00:01:55,800 --> 00:01:59,160 Speaker 1: the diameter of the hair, that's the range you would 29 00:01:59,200 --> 00:02:02,520 Speaker 1: be at. So when you're talking about seven or five 30 00:02:02,760 --> 00:02:08,320 Speaker 1: or even three nanometers, that's super duper small, right, Well, 31 00:02:08,360 --> 00:02:11,480 Speaker 1: it would be if the nanometer designation still referred to 32 00:02:11,720 --> 00:02:15,760 Speaker 1: component size. Now, once upon a time the scale reference 33 00:02:15,800 --> 00:02:19,480 Speaker 1: to a process actually did correspond with at least some 34 00:02:19,639 --> 00:02:23,320 Speaker 1: component size on the chip itself, but that has not 35 00:02:23,480 --> 00:02:26,880 Speaker 1: been the case for several generations. Now. Part of the 36 00:02:26,919 --> 00:02:29,800 Speaker 1: reason for that comes down to the limitations of physics. 37 00:02:30,200 --> 00:02:32,720 Speaker 1: As you shrink down to the bottom end of the 38 00:02:32,800 --> 00:02:36,400 Speaker 1: nanoscale and into the atomic scale, you start to have 39 00:02:36,440 --> 00:02:40,799 Speaker 1: to contend with quantum mechanics. Now, we don't encounter quantum 40 00:02:41,520 --> 00:02:45,760 Speaker 1: mechanic effects on our normal scale like in our everyday lives. 41 00:02:46,440 --> 00:02:49,440 Speaker 1: But at that tiny scale, things start to behave in 42 00:02:49,480 --> 00:02:53,880 Speaker 1: a really wonky way, and relying on physical structures to 43 00:02:54,000 --> 00:02:58,079 Speaker 1: rain in quantum silliness becomes a big challenge. I've done 44 00:02:58,080 --> 00:03:00,440 Speaker 1: full episodes kind of about this. We're not going to 45 00:03:00,560 --> 00:03:05,120 Speaker 1: dive too deeply into it, so instead, the scale really 46 00:03:05,160 --> 00:03:08,120 Speaker 1: is more of a marketing strategy. When you hear it's 47 00:03:08,120 --> 00:03:10,760 Speaker 1: a five nanometer process, it doesn't mean that anything on 48 00:03:10,800 --> 00:03:14,040 Speaker 1: that chip actually measures five nanometers in size. It's a 49 00:03:14,040 --> 00:03:18,520 Speaker 1: way of indicating this process is more advanced than the 50 00:03:18,520 --> 00:03:24,400 Speaker 1: previous seven nanometer process. So it really means that when 51 00:03:24,440 --> 00:03:27,040 Speaker 1: you get down to the process and the architecture, you 52 00:03:27,080 --> 00:03:30,600 Speaker 1: start to converge on essentially the same meaning. So let's 53 00:03:30,680 --> 00:03:37,040 Speaker 1: talk about that architecture. What does chip architecture actually mean. Well, 54 00:03:37,120 --> 00:03:40,160 Speaker 1: we're going to stick with CPUs, also known as central 55 00:03:40,200 --> 00:03:43,280 Speaker 1: processing units, and we can think of a CPU as 56 00:03:43,320 --> 00:03:49,600 Speaker 1: having three major components. These are the registers, the arithmetic 57 00:03:49,680 --> 00:03:54,240 Speaker 1: logic unit or ALU, and the control unit. So registers 58 00:03:54,280 --> 00:03:58,000 Speaker 1: act kind of like memory, and that they hold information 59 00:03:58,080 --> 00:04:02,280 Speaker 1: that the CPU needs in order to complete operations. Logic 60 00:04:02,320 --> 00:04:06,200 Speaker 1: gates make up the quote unquote memory of registers, and 61 00:04:06,240 --> 00:04:09,000 Speaker 1: a logic gate follows a specific rule. It creates an 62 00:04:09,000 --> 00:04:12,560 Speaker 1: output based upon the input coming into the logic gate. 63 00:04:13,400 --> 00:04:15,560 Speaker 1: I'll do a full episode just about logic gates in 64 00:04:15,560 --> 00:04:18,240 Speaker 1: the future to kind of expand on that and explain 65 00:04:18,560 --> 00:04:22,000 Speaker 1: how these logic gates work and how by combining logic 66 00:04:22,040 --> 00:04:26,440 Speaker 1: gates you can create different outcomes. So registers operate faster 67 00:04:26,800 --> 00:04:31,240 Speaker 1: than RAM, a random access memory, which I often at 68 00:04:31,320 --> 00:04:36,560 Speaker 1: least will compare to short term memory in humans. RAM, 69 00:04:36,720 --> 00:04:40,520 Speaker 1: in turn operates faster than a solid state drive or 70 00:04:40,520 --> 00:04:43,760 Speaker 1: a hard drive, which I compare to long term memory 71 00:04:43,839 --> 00:04:47,839 Speaker 1: with humans. So you've got registers, which are the fastest 72 00:04:47,920 --> 00:04:51,880 Speaker 1: access of memory total, but it holds very little information. 73 00:04:52,000 --> 00:04:55,320 Speaker 1: It's just tiny, tiny bits of information. Then you have RAM. 74 00:04:55,520 --> 00:04:59,600 Speaker 1: Then you've got solid state drive or hard drive in registers. 75 00:04:59,640 --> 00:05:03,080 Speaker 1: We actually we have five basic types, so let's list 76 00:05:03,080 --> 00:05:07,640 Speaker 1: them off, shall we. The instruction register stores the address 77 00:05:07,680 --> 00:05:10,560 Speaker 1: and random access memory of the instruction to be used 78 00:05:10,720 --> 00:05:14,159 Speaker 1: in a given operation. So that instruction could be some 79 00:05:14,480 --> 00:05:19,600 Speaker 1: basic arithmetic function for example like AD. Next, you've got 80 00:05:19,600 --> 00:05:23,320 Speaker 1: the memory address register. This stores the address within RAM 81 00:05:23,720 --> 00:05:26,279 Speaker 1: of the data that is to be processed. So this 82 00:05:26,400 --> 00:05:28,680 Speaker 1: is the data that's going to be transformed by that 83 00:05:28,800 --> 00:05:32,400 Speaker 1: instruction in some way. Your instruction register has the info 84 00:05:32,480 --> 00:05:35,240 Speaker 1: on what operation to use. The memory address register has 85 00:05:35,279 --> 00:05:37,800 Speaker 1: the info on which data is going to undergo that operation. 86 00:05:39,160 --> 00:05:42,560 Speaker 1: Then you've got the memory data register. This stores the 87 00:05:42,640 --> 00:05:46,680 Speaker 1: data that the CPU is actually processing at any given time. 88 00:05:46,800 --> 00:05:48,800 Speaker 1: So while the other two registers are kind of like 89 00:05:48,880 --> 00:05:51,680 Speaker 1: looking into the future like the next step, the memory 90 00:05:51,760 --> 00:05:54,839 Speaker 1: data register is concerned with what's going on right now, 91 00:05:54,880 --> 00:05:58,760 Speaker 1: gosh darn it. Then you've got the program counter. This 92 00:05:58,920 --> 00:06:02,480 Speaker 1: stores the address and RAM of the next instruction coming up, 93 00:06:02,520 --> 00:06:05,960 Speaker 1: so the next one down the line. Finally, you've got 94 00:06:05,960 --> 00:06:10,800 Speaker 1: the accumulator that stores the results of the calculations that 95 00:06:10,839 --> 00:06:14,520 Speaker 1: were just performed. So the registers are one part of 96 00:06:14,680 --> 00:06:19,919 Speaker 1: CPU architecture. Now let's talk about the ALU or arithmetic 97 00:06:20,120 --> 00:06:25,960 Speaker 1: logic unit. The ALU is the brains of the CPU. 98 00:06:26,440 --> 00:06:29,840 Speaker 1: Within the ALU are logic circuits which actually carry out 99 00:06:29,880 --> 00:06:33,839 Speaker 1: the operations on data. These operations span a wide range 100 00:06:33,839 --> 00:06:39,800 Speaker 1: of arithmetic tasks like addition and subtraction, to things like 101 00:06:39,920 --> 00:06:44,080 Speaker 1: incrementation and also comparison. So, for example, you might have 102 00:06:44,240 --> 00:06:47,480 Speaker 1: a pair of operations that each produce a result and 103 00:06:47,520 --> 00:06:50,320 Speaker 1: the ALU has to compare these results with one another 104 00:06:50,600 --> 00:06:54,039 Speaker 1: to determine if they are the same or different. That's 105 00:06:54,080 --> 00:06:57,800 Speaker 1: the kind of basic task the ALU handles, and it 106 00:06:57,800 --> 00:07:02,839 Speaker 1: does this super fast. Finally, you have the control unit, which, 107 00:07:02,880 --> 00:07:07,120 Speaker 1: as the name suggests, controls the process. The control unit 108 00:07:07,200 --> 00:07:10,520 Speaker 1: receives instructions, decodes those to get to the meaning of 109 00:07:10,560 --> 00:07:13,920 Speaker 1: the instructions, sends commands to the other components to carry 110 00:07:13,960 --> 00:07:17,360 Speaker 1: out those instructions, et cetera. The control unit is kind 111 00:07:17,360 --> 00:07:19,880 Speaker 1: of like a floor manager. It makes sure all the 112 00:07:19,920 --> 00:07:23,560 Speaker 1: departments are responding appropriately given the program that's running at 113 00:07:23,600 --> 00:07:27,480 Speaker 1: any given time. The control unit also has a clock, 114 00:07:27,840 --> 00:07:30,520 Speaker 1: but that clock isn't meant to keep your computer's time 115 00:07:30,640 --> 00:07:35,040 Speaker 1: accurate to local time. This clock oscillates a certain number 116 00:07:35,040 --> 00:07:37,920 Speaker 1: of times per second, and we measure this in hurts, 117 00:07:38,040 --> 00:07:42,640 Speaker 1: So one oscillation per second would be one hurts. Typically, 118 00:07:43,080 --> 00:07:46,960 Speaker 1: with processors today, we're talking about the gigahertz range. A 119 00:07:47,000 --> 00:07:50,720 Speaker 1: gigahertz would be a billion oscillations per second. So a 120 00:07:50,800 --> 00:07:54,480 Speaker 1: three point two gigahertz CPU has a clock that in 121 00:07:54,520 --> 00:07:58,360 Speaker 1: the control unit that oscillates three point two billion times 122 00:07:59,000 --> 00:08:02,880 Speaker 1: every single So now the clock speed relates to how 123 00:08:03,000 --> 00:08:07,920 Speaker 1: quickly the processor can actually complete these operations. Some operations 124 00:08:07,960 --> 00:08:11,560 Speaker 1: require multiple oscillations, but that clock speed or frequency, if 125 00:08:11,600 --> 00:08:14,280 Speaker 1: you prefer, gives you an idea of how fast or 126 00:08:14,320 --> 00:08:17,560 Speaker 1: powerful your computer is. Now. Other factors also play into 127 00:08:17,640 --> 00:08:20,080 Speaker 1: this too. It's not just clock speed, but that is 128 00:08:20,160 --> 00:08:23,000 Speaker 1: one big component in it. If you're familiar with the 129 00:08:23,080 --> 00:08:26,280 Speaker 1: term overclocking, then all of the stuff I'm talking to 130 00:08:26,320 --> 00:08:29,760 Speaker 1: you about is old news to you, right. Overclocking is 131 00:08:29,800 --> 00:08:34,040 Speaker 1: the practice of increasing that clock oscillation speed in the 132 00:08:34,080 --> 00:08:40,600 Speaker 1: control unit beyond its default settings, which typically the manufacturer creates. 133 00:08:40,640 --> 00:08:44,079 Speaker 1: Like they create default settings, they say this processor is 134 00:08:44,240 --> 00:08:48,000 Speaker 1: rated at this particular clock speed, and going beyond that 135 00:08:48,120 --> 00:08:52,640 Speaker 1: could potentially reduce the useful lifespan of the processor or 136 00:08:53,120 --> 00:08:57,600 Speaker 1: cause it to overheat, et cetera. So elite gamers typically 137 00:08:57,679 --> 00:09:02,160 Speaker 1: will use programs to boost the clock speed on CPUs 138 00:09:02,160 --> 00:09:05,840 Speaker 1: to get past these limitations and to push it faster 139 00:09:05,960 --> 00:09:08,800 Speaker 1: than what it was rated as in order to milk 140 00:09:08,840 --> 00:09:12,360 Speaker 1: out higher performance in their gaming ricks. Doing this does 141 00:09:12,400 --> 00:09:14,520 Speaker 1: come with some trade offs. I mean, it does mean 142 00:09:14,559 --> 00:09:16,800 Speaker 1: that you might be burning through your CPU faster than 143 00:09:16,800 --> 00:09:19,840 Speaker 1: you usually would. It also typically means that the computer's 144 00:09:19,880 --> 00:09:21,920 Speaker 1: going to generate a lot more heat, so you need 145 00:09:21,960 --> 00:09:25,040 Speaker 1: to have a good heat dispersal system in place to 146 00:09:25,120 --> 00:09:27,880 Speaker 1: carry that heat away from the processor because, as we know, 147 00:09:29,080 --> 00:09:33,440 Speaker 1: heat and electronics are not super friendly with one another. 148 00:09:34,520 --> 00:09:39,160 Speaker 1: Connecting all these different components are wires called buses, So 149 00:09:39,280 --> 00:09:43,480 Speaker 1: a bus might carry instructions, another bus might carry data. 150 00:09:44,120 --> 00:09:46,520 Speaker 1: The capacity of buses also plays a part in how 151 00:09:46,559 --> 00:09:49,800 Speaker 1: powerful a computer is. I'll have to do another episode 152 00:09:49,840 --> 00:09:52,960 Speaker 1: to explain things like what is a thirty two bit 153 00:09:53,120 --> 00:09:56,920 Speaker 1: machine versus a sixty four bit machine, or even with 154 00:09:57,000 --> 00:10:00,400 Speaker 1: the old game consoles, an eight bit machine. We'll talk 155 00:10:00,440 --> 00:10:03,240 Speaker 1: about bitwidth and that kind of stuff, but that kind 156 00:10:03,240 --> 00:10:06,040 Speaker 1: of plays into things like buses. You can think of 157 00:10:06,040 --> 00:10:08,720 Speaker 1: it sort of like roads. How wide is the road 158 00:10:09,040 --> 00:10:12,280 Speaker 1: so how many vehicles can pass side by side at 159 00:10:12,280 --> 00:10:15,440 Speaker 1: the same time. And one other thing that we will 160 00:10:15,440 --> 00:10:18,880 Speaker 1: mention will be cores, and I'm going to get to 161 00:10:18,920 --> 00:10:32,640 Speaker 1: that after we take this quick break. Okay, before the break, 162 00:10:32,760 --> 00:10:35,360 Speaker 1: I teased that we're going to talk about cores. A 163 00:10:35,360 --> 00:10:38,760 Speaker 1: CPU core is the smallest unit that can carry out 164 00:10:38,800 --> 00:10:42,560 Speaker 1: all the jobs that a CPU does. So if you 165 00:10:42,600 --> 00:10:44,959 Speaker 1: hear of a multicore CPU, that means each of those 166 00:10:45,000 --> 00:10:48,360 Speaker 1: cores can do the job of a CPU, and they 167 00:10:48,400 --> 00:10:51,040 Speaker 1: can have multiple cores. You'll hear things like dual core, 168 00:10:51,200 --> 00:10:53,920 Speaker 1: which means there's two of them, or quad core, meaning 169 00:10:53,920 --> 00:10:57,880 Speaker 1: there's four of them, and beyond, each core can carry 170 00:10:57,880 --> 00:11:00,680 Speaker 1: out the duties of a CPU. So does that mean 171 00:11:00,720 --> 00:11:04,360 Speaker 1: a dual core or quad core processor is automatically better 172 00:11:04,440 --> 00:11:10,120 Speaker 1: than a single core processor. Not necessarily so. For some 173 00:11:10,520 --> 00:11:13,960 Speaker 1: types of computational problems, you can actually divide up the 174 00:11:14,000 --> 00:11:18,640 Speaker 1: problem into smaller tasks that could be completed simultaneously. So 175 00:11:18,800 --> 00:11:22,400 Speaker 1: these are the types of problems that multicore processors are 176 00:11:22,600 --> 00:11:25,760 Speaker 1: great at tackling because each core can tackle a different 177 00:11:25,800 --> 00:11:29,480 Speaker 1: set of tasks and thus collectively they'll get to the 178 00:11:29,480 --> 00:11:33,760 Speaker 1: answer faster. But if the problem cannot be broken down 179 00:11:33,800 --> 00:11:38,480 Speaker 1: into smaller pieces, a very powerful single core processor might 180 00:11:38,559 --> 00:11:43,360 Speaker 1: be better than a decently powerful multicore processor. And I 181 00:11:43,480 --> 00:11:46,760 Speaker 1: use this analogy all the time. Longtime listeners are probably 182 00:11:47,360 --> 00:11:49,880 Speaker 1: tired of it, and they've anticipated it, and yes, it's 183 00:11:49,920 --> 00:11:52,319 Speaker 1: okay to skip ahead a little bit if you are 184 00:11:52,320 --> 00:11:55,080 Speaker 1: one of those people. But I like to describe multicore 185 00:11:55,200 --> 00:11:58,800 Speaker 1: processors versus a single core processor by talking about an 186 00:11:58,840 --> 00:12:01,839 Speaker 1: advanced math class, And in this version of it, I'm 187 00:12:01,840 --> 00:12:04,400 Speaker 1: going to say there are five students in this advanced 188 00:12:04,640 --> 00:12:08,320 Speaker 1: math class. Now imagine four of those five students are 189 00:12:08,360 --> 00:12:12,160 Speaker 1: all really good at math, right, they're gifted students. However, 190 00:12:12,200 --> 00:12:16,200 Speaker 1: the fifth student is a genuine math genius. And the 191 00:12:16,280 --> 00:12:20,920 Speaker 1: genius always completes any given problem faster than the other 192 00:12:21,000 --> 00:12:24,080 Speaker 1: four students can. And one day the teacher presents a 193 00:12:24,200 --> 00:12:27,360 Speaker 1: challenge to the class. It's a pop quiz that has 194 00:12:27,600 --> 00:12:31,319 Speaker 1: four questions on the quiz. The genius has to try 195 00:12:31,360 --> 00:12:35,079 Speaker 1: and complete all four problems, but the other four students 196 00:12:35,280 --> 00:12:38,040 Speaker 1: can actually divide up the quiz and each student can 197 00:12:38,080 --> 00:12:41,560 Speaker 1: tackle a single problem on there, so collectively they can 198 00:12:41,600 --> 00:12:46,400 Speaker 1: solve the quiz together. So who is going to finish first? Well, 199 00:12:46,440 --> 00:12:50,560 Speaker 1: if we assume that each problem is discreete and independent 200 00:12:50,920 --> 00:12:53,960 Speaker 1: of the outcomes of the other problems, the four students 201 00:12:54,000 --> 00:12:57,280 Speaker 1: are likely to finish their quiz collectively before the genius 202 00:12:57,280 --> 00:13:00,840 Speaker 1: because each one's just doing one question, and the genius 203 00:13:00,920 --> 00:13:04,120 Speaker 1: is still faster than all the individuals. But they have 204 00:13:04,200 --> 00:13:07,960 Speaker 1: to do all four questions, whereas each smart student just 205 00:13:08,000 --> 00:13:11,120 Speaker 1: has to do one. The multi core processor wins in 206 00:13:11,160 --> 00:13:16,800 Speaker 1: that scenario, but Let's say you find out that problem 207 00:13:16,880 --> 00:13:19,679 Speaker 1: two on the quiz actually depends upon the outcome of 208 00:13:19,720 --> 00:13:23,640 Speaker 1: problem one, and you find out the problem three depends 209 00:13:23,720 --> 00:13:27,520 Speaker 1: upon the outcome of problem two, and the problem four 210 00:13:27,600 --> 00:13:30,120 Speaker 1: depends on the outcome of problem three. Well, now you 211 00:13:30,160 --> 00:13:32,560 Speaker 1: can't just divide up the problems between the four students 212 00:13:33,040 --> 00:13:35,720 Speaker 1: because the student working on problem two has to wait 213 00:13:36,080 --> 00:13:37,960 Speaker 1: to find out what the answer to problem one is 214 00:13:38,000 --> 00:13:41,120 Speaker 1: before they can get started. The genius in that case 215 00:13:41,280 --> 00:13:43,640 Speaker 1: is going to win that race, right, because they're still 216 00:13:43,679 --> 00:13:47,160 Speaker 1: faster than any individual is. So for certain types of 217 00:13:47,240 --> 00:13:52,000 Speaker 1: computational problems and processes, multicore is the way to go, 218 00:13:52,600 --> 00:13:55,400 Speaker 1: but not in every case, just in a lot of them. 219 00:13:55,679 --> 00:13:59,320 Speaker 1: For a lot of computer users, it's more important to 220 00:13:59,320 --> 00:14:02,880 Speaker 1: go multi corep because the typical uses that they rely 221 00:14:03,000 --> 00:14:06,680 Speaker 1: upon with computers falls into that multi core set of problems. 222 00:14:06,679 --> 00:14:10,760 Speaker 1: This includes gamers, So a multi core processor matched with 223 00:14:10,840 --> 00:14:15,360 Speaker 1: a really good graphics processing unit that's more important than 224 00:14:15,400 --> 00:14:19,200 Speaker 1: having just a single core super fast processor. But again, 225 00:14:19,280 --> 00:14:22,920 Speaker 1: it all depends on how you can thread the computational problems. 226 00:14:23,560 --> 00:14:27,480 Speaker 1: And that's the general description of what computer architecture means. 227 00:14:27,520 --> 00:14:30,760 Speaker 1: The actual design and layout of these components is what 228 00:14:30,840 --> 00:14:34,600 Speaker 1: sets one chip apart from another chip. Since it is 229 00:14:34,720 --> 00:14:39,000 Speaker 1: increasingly challenging to shrink components down without getting into quantum 230 00:14:39,040 --> 00:14:42,360 Speaker 1: effects or generating too much heat in a very small space, 231 00:14:42,760 --> 00:14:47,120 Speaker 1: finding the best possible layout and orientation of components is critical. 232 00:14:48,200 --> 00:14:49,600 Speaker 1: You know you're not going to be able to cram 233 00:14:49,720 --> 00:14:52,120 Speaker 1: a whole lot more on, but you might be able 234 00:14:52,160 --> 00:14:55,800 Speaker 1: to find an orientation that gets a little better performance 235 00:14:55,880 --> 00:14:58,920 Speaker 1: out of the components you have now. Back in the day, 236 00:14:59,320 --> 00:15:02,080 Speaker 1: Intel which which is one of two major companies behind 237 00:15:02,120 --> 00:15:05,800 Speaker 1: the processors used in most computers these days, used a 238 00:15:05,880 --> 00:15:08,680 Speaker 1: development approach and chip designed that the company referred to 239 00:15:08,760 --> 00:15:12,160 Speaker 1: as the tick talk method. So you can think of 240 00:15:12,200 --> 00:15:16,080 Speaker 1: the tick part of TikTok as taking the same chip 241 00:15:16,160 --> 00:15:20,200 Speaker 1: layout design from the previous generation, but then shrinking everything 242 00:15:20,240 --> 00:15:22,320 Speaker 1: down a little bit, which allows you to cram more 243 00:15:22,320 --> 00:15:26,040 Speaker 1: components on the chip. So you're following the same architectural 244 00:15:26,160 --> 00:15:29,440 Speaker 1: plan as the previous generation, but now all the components 245 00:15:29,440 --> 00:15:31,560 Speaker 1: are slightly smaller so you can have more of them there. 246 00:15:32,080 --> 00:15:36,600 Speaker 1: The talk sequence would involve creating a new architecture that 247 00:15:36,880 --> 00:15:40,800 Speaker 1: better takes advantage of these smaller components, and then it 248 00:15:40,800 --> 00:15:44,440 Speaker 1: would repeat tech talk Tick talk. So with tick you 249 00:15:44,480 --> 00:15:46,680 Speaker 1: shrink stuff down, but you follow the same game plan 250 00:15:46,760 --> 00:15:49,240 Speaker 1: as before. With talk, you create a new game plan, 251 00:15:49,720 --> 00:15:52,480 Speaker 1: and then you do tick again. And so each generation 252 00:15:52,520 --> 00:15:59,040 Speaker 1: of Intel fell into one of those two design principles, 253 00:15:59,560 --> 00:16:02,040 Speaker 1: and in this way Intel would iterate its chip designs. 254 00:16:02,040 --> 00:16:05,160 Speaker 1: Each generation would improve upon the last. At least that 255 00:16:05,280 --> 00:16:08,520 Speaker 1: was the idea, either by adding more capability in the 256 00:16:08,560 --> 00:16:11,840 Speaker 1: form of more components added to the chip, or finding 257 00:16:11,920 --> 00:16:15,359 Speaker 1: a new way to arrange those components that improve performance. 258 00:16:15,640 --> 00:16:19,320 Speaker 1: And by improved performance, I mean not just being faster 259 00:16:19,840 --> 00:16:24,320 Speaker 1: or more capable, but also more power efficient or creating 260 00:16:24,400 --> 00:16:28,320 Speaker 1: less heat, because these things do matter quite a bit. 261 00:16:29,000 --> 00:16:33,720 Speaker 1: And that's our overview of chip architecture. I'll do more 262 00:16:33,760 --> 00:16:36,640 Speaker 1: episodes about the basics of CPUs soon. Maybe i'll talk 263 00:16:36,680 --> 00:16:40,120 Speaker 1: a bit about what makes an Intel chip different from say, 264 00:16:40,760 --> 00:16:44,280 Speaker 1: an AMD chip. And you may know, if you've ever 265 00:16:44,360 --> 00:16:49,320 Speaker 1: built a computer, the type of processor you want ends 266 00:16:49,400 --> 00:16:52,080 Speaker 1: up mattering a big deal, because it will tell you 267 00:16:52,160 --> 00:16:54,920 Speaker 1: what kind of motherboard you can use, for example, because 268 00:16:54,960 --> 00:16:57,160 Speaker 1: a motherboard designed to work with an Intel chip is 269 00:16:57,200 --> 00:17:00,840 Speaker 1: not going to work with an AMD chip. Thing. So 270 00:17:00,920 --> 00:17:03,840 Speaker 1: we'll do another episode to talk a bit more about 271 00:17:03,840 --> 00:17:06,480 Speaker 1: this in the future, and keep it nice and short 272 00:17:06,520 --> 00:17:10,040 Speaker 1: and simple so that folks can listen, get a good understanding, 273 00:17:10,080 --> 00:17:11,800 Speaker 1: and then know what to look for when they move 274 00:17:11,800 --> 00:17:14,040 Speaker 1: forward if they ever decide to build their own computer. 275 00:17:14,440 --> 00:17:16,679 Speaker 1: And I think we'll also, like I said, to an 276 00:17:16,720 --> 00:17:20,000 Speaker 1: episode about things like logic gates to kind of understand 277 00:17:20,080 --> 00:17:23,119 Speaker 1: at a very very very basic level, what is going 278 00:17:23,160 --> 00:17:27,040 Speaker 1: on when a computer is processing information. That's it for 279 00:17:27,080 --> 00:17:30,520 Speaker 1: this Tech Stuff Tidbits episode. I hope you are all well. 280 00:17:30,600 --> 00:17:34,600 Speaker 1: Just a reminder next week, I am on vacation and 281 00:17:34,680 --> 00:17:36,680 Speaker 1: I will be back the following week, so we will 282 00:17:36,720 --> 00:17:40,200 Speaker 1: likely have some reruns playing next week, but I will 283 00:17:40,240 --> 00:17:43,520 Speaker 1: be back and I'll talk to you again really soon. 284 00:17:49,720 --> 00:17:54,359 Speaker 1: Tech Stuff is an iHeartRadio production. For more podcasts from iHeartRadio, 285 00:17:54,680 --> 00:17:58,400 Speaker 1: visit the iHeartRadio app, Apple Podcasts, or wherever you listen 286 00:17:58,440 --> 00:18:03,040 Speaker 1: to your favorite shows.