1
00:00:00,160 --> 00:00:06,800
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.

2
00:00:09,320 --> 00:00:13,040
Speaker 2: This is the Bloomberg Daybreak Asia podcast. I'm Brian Curtis

3
00:00:13,119 --> 00:00:15,840
Speaker 2: along with Doug Krisner. Join us each day for the

4
00:00:15,880 --> 00:00:19,240
Speaker 2: stories making news and moving markets in the Asia Pacific.

5
00:00:19,480 --> 00:00:21,920
Speaker 2: You can subscribe to the show anywhere you get your

6
00:00:21,920 --> 00:00:25,880
Speaker 2: podcasts and always on Bloomberg Radio, the Bloomberg Terminal, and

7
00:00:25,920 --> 00:00:27,240
Speaker 2: the Bloomberg Business App.

8
00:00:27,520 --> 00:00:31,160
Speaker 1: Let's bring in Bloomberg Intelligence Senior tech industry analyst man

9
00:00:31,200 --> 00:00:34,640
Speaker 1: Deep Singh joining me here in the Bloomberg Interactive Brokers

10
00:00:34,640 --> 00:00:37,720
Speaker 1: studio at Bloomberg Headquarters in New York. So you heard

11
00:00:37,720 --> 00:00:40,920
Speaker 1: the conversation that Ed Ludlow had with CEO Jensen Wang.

12
00:00:41,479 --> 00:00:42,840
Speaker 1: What did you think? What's your reaction?

13
00:00:43,240 --> 00:00:47,159
Speaker 3: Results were great, it's still a very solid growth story.

14
00:00:47,200 --> 00:00:50,680
Speaker 3: But when it comes to convincing investors that there are

15
00:00:50,720 --> 00:00:56,640
Speaker 3: more catalysts for upside relative to where expectations are, I

16
00:00:56,680 --> 00:01:00,560
Speaker 3: think a lot of it right now is the on

17
00:01:00,840 --> 00:01:06,000
Speaker 3: how the industry actually deploys generative AI. That's what Jensen

18
00:01:06,080 --> 00:01:12,399
Speaker 3: highlighted that. Look, GENAI is transformational. It's changing how compute

19
00:01:12,520 --> 00:01:16,160
Speaker 3: is done and everyone will use it and we are

20
00:01:16,200 --> 00:01:18,880
Speaker 3: seeing that in their numbers. The data center growth is

21
00:01:19,120 --> 00:01:24,880
Speaker 3: just unprecedented. Question for investors is will we see more

22
00:01:25,200 --> 00:01:29,680
Speaker 3: positive revisions? Will estimates go up by five ten percent tomorrow?

23
00:01:30,000 --> 00:01:34,600
Speaker 3: And the answer is no, simply because right now they

24
00:01:34,640 --> 00:01:39,440
Speaker 3: talked about Blackwell getting delayed to four Q. Obviously they

25
00:01:39,440 --> 00:01:43,360
Speaker 3: are very positive on the ramp, but everything is more

26
00:01:43,720 --> 00:01:47,040
Speaker 3: like pushed out by a quarter and they I mean,

27
00:01:47,120 --> 00:01:50,120
Speaker 3: Jensen was very positive with ed you know in the

28
00:01:50,200 --> 00:01:54,600
Speaker 3: interview about how he expects twenty twenty five to shake out,

29
00:01:55,120 --> 00:01:58,880
Speaker 3: So he's his enthusiasm is intact. It's just when it

30
00:01:58,920 --> 00:02:02,080
Speaker 3: comes to estimates going up, probably investors didn't see enough.

31
00:02:02,440 --> 00:02:05,280
Speaker 1: There was a lot of conversation around the hyperscalers. These

32
00:02:05,280 --> 00:02:08,080
Speaker 1: are the big cloud computing companies that have been investing

33
00:02:08,120 --> 00:02:10,680
Speaker 1: in this technology, and I think the big question for

34
00:02:10,760 --> 00:02:13,000
Speaker 1: the street is whether or not these companies are going

35
00:02:13,080 --> 00:02:16,240
Speaker 1: to begin to demonstrate a return on investment the ROA

36
00:02:16,440 --> 00:02:18,720
Speaker 1: on AI. What do we know about that part of

37
00:02:18,760 --> 00:02:19,200
Speaker 1: the story.

38
00:02:19,360 --> 00:02:23,800
Speaker 3: I think with Hyperscalers, specifically, their cloud business will ramp up.

39
00:02:23,880 --> 00:02:27,200
Speaker 3: And that's what he said, like he's not Nvidia is

40
00:02:27,240 --> 00:02:30,639
Speaker 3: not competing with the hyperscalers. They don't intend to become

41
00:02:30,760 --> 00:02:35,160
Speaker 3: a hyperscaler vendor like a Microsoft, Amazon, or Google. They're

42
00:02:35,200 --> 00:02:39,840
Speaker 3: deploying within Microsoft and Google, and I think that's where

43
00:02:40,560 --> 00:02:43,839
Speaker 3: probably you will see a lot of ramp up from

44
00:02:43,840 --> 00:02:46,960
Speaker 3: the hyperscalers. But that doesn't mean they will continue to

45
00:02:47,080 --> 00:02:50,480
Speaker 3: raise Capex for twenty twenty six and beyond. Ultimately, the

46
00:02:50,560 --> 00:02:55,639
Speaker 3: question for Nvidia investors is is this CAPEX wave sustainable?

47
00:02:56,000 --> 00:02:58,600
Speaker 3: How long of a tail it has and is there

48
00:02:58,639 --> 00:03:01,040
Speaker 3: a cliff three years out?

49
00:03:01,400 --> 00:03:04,400
Speaker 1: Well, you mentioned the Blackwell design, the new chip, the

50
00:03:04,840 --> 00:03:07,680
Speaker 1: GPU that in video will roll out. Apparently there's some

51
00:03:07,840 --> 00:03:10,920
Speaker 1: production issues right now. Is this more so on the

52
00:03:10,960 --> 00:03:13,440
Speaker 1: design side, that's the sense that I'm getting, rather than

53
00:03:13,480 --> 00:03:14,760
Speaker 1: production itself, right.

54
00:03:14,880 --> 00:03:18,080
Speaker 3: Yeah, I mean, look, the complexity of the chips is growing.

55
00:03:18,160 --> 00:03:21,560
Speaker 3: He talked about power, how Blackwell will use the same

56
00:03:21,600 --> 00:03:25,040
Speaker 3: amount of power but be able to compute more than

57
00:03:25,080 --> 00:03:29,200
Speaker 3: the current Hopper architecture. So clearly they're doing a lot

58
00:03:29,320 --> 00:03:32,080
Speaker 3: more on the content side of the chip and it

59
00:03:32,160 --> 00:03:36,920
Speaker 3: requires sophisticated production, which TSMC is doing. But again he

60
00:03:37,080 --> 00:03:40,480
Speaker 3: came across very confident that they will ramp up four

61
00:03:40,600 --> 00:03:44,520
Speaker 3: Q and then if going into twenty twenty five, Blackwell

62
00:03:44,560 --> 00:03:49,440
Speaker 3: will be multi billions in revenue, so that conviction was

63
00:03:49,480 --> 00:03:53,440
Speaker 3: there in what he said during the interview. It's more about,

64
00:03:53,680 --> 00:03:59,040
Speaker 3: you know, will it be enough to drive more upside

65
00:03:59,080 --> 00:04:03,560
Speaker 3: in terms of the incremental buyers CAPEX going up. I mean,

66
00:04:03,600 --> 00:04:07,720
Speaker 3: he talked about there are more companies that are trading models.

67
00:04:08,200 --> 00:04:12,000
Speaker 3: Is just not clear enough, like who are these beyond

68
00:04:12,040 --> 00:04:16,479
Speaker 3: the hyperscaler sovereigns he mentioned sovereigns he did, and those

69
00:04:16,480 --> 00:04:20,839
Speaker 3: are low double digit and billion dollars. But again when

70
00:04:20,880 --> 00:04:24,400
Speaker 3: you compare it to the hyperscalers, it's not very clear

71
00:04:24,839 --> 00:04:28,560
Speaker 3: if the magnitude of those buyers will be the same

72
00:04:28,600 --> 00:04:29,760
Speaker 3: as the hyperscalers.

73
00:04:29,839 --> 00:04:30,080
Speaker 4: End.

74
00:04:30,240 --> 00:04:33,520
Speaker 1: So you mentioned the production side, that's TSMC, the company

75
00:04:33,560 --> 00:04:36,279
Speaker 1: that is the foundry for these Nvidia chips. Do we

76
00:04:36,320 --> 00:04:38,520
Speaker 1: have an indication that TSMC is going to be able

77
00:04:38,560 --> 00:04:39,880
Speaker 1: to keep up with demand as well?

78
00:04:40,080 --> 00:04:40,359
Speaker 4: Yes.

79
00:04:40,440 --> 00:04:43,679
Speaker 3: In fact, he was very clear that supply is growing,

80
00:04:43,839 --> 00:04:46,880
Speaker 3: and it's growing in a way where TSMC is able

81
00:04:46,920 --> 00:04:52,960
Speaker 3: to accommodate in Vidia becoming a larger buyer of the

82
00:04:52,960 --> 00:04:56,880
Speaker 3: manufacturing capacity that TSMC has. So all the things that

83
00:04:56,960 --> 00:04:59,599
Speaker 3: he mentioned of the on the call pointed to a

84
00:04:59,600 --> 00:05:04,120
Speaker 3: steady ramp up more supply of chips but what it

85
00:05:04,160 --> 00:05:08,800
Speaker 3: didn't suggest is where are more upside catalysts. I mean,

86
00:05:08,839 --> 00:05:12,440
Speaker 3: everyone knows about the new architecture, everyone knows about the

87
00:05:12,560 --> 00:05:15,440
Speaker 3: ramp up and supply, So it's more a question of

88
00:05:15,720 --> 00:05:19,240
Speaker 3: is the upside going to come from networking or software

89
00:05:19,920 --> 00:05:23,080
Speaker 3: or cloud. He said they don't intend to become a

90
00:05:23,080 --> 00:05:26,039
Speaker 3: cloud player. So the software business, they said they're going

91
00:05:26,120 --> 00:05:28,120
Speaker 3: to end at a two billion dollar rund rate, So

92
00:05:28,520 --> 00:05:33,840
Speaker 3: that upside for consensus revisions probably that is what is missing.

93
00:05:33,960 --> 00:05:36,000
Speaker 1: One of the things that I didn't hear any discussion

94
00:05:36,040 --> 00:05:39,359
Speaker 1: around was the high bandwidth memory chips. And you need

95
00:05:39,400 --> 00:05:41,880
Speaker 1: those to kind of work with the GPUs, don't you

96
00:05:41,960 --> 00:05:45,560
Speaker 1: when you're creating these cloud infrastructure units. Yes, and we're

97
00:05:45,560 --> 00:05:48,919
Speaker 1: talking about companies like sk Heinez, like Samsung. Do we

98
00:05:49,040 --> 00:05:51,039
Speaker 1: know that these companies are going to be able to

99
00:05:51,120 --> 00:05:54,400
Speaker 1: keep up at the same level that Nvidia is producing chips?

100
00:05:54,560 --> 00:05:57,000
Speaker 3: Yeah, And actually that's a great point. So when you

101
00:05:57,160 --> 00:06:00,919
Speaker 3: look at the gross margins for Nvidia, they seem to

102
00:06:01,000 --> 00:06:04,280
Speaker 3: have peaked last quarter. So last quarter gross margins were

103
00:06:04,320 --> 00:06:07,960
Speaker 3: seventy eight percent. And remember this business right now has

104
00:06:08,000 --> 00:06:11,679
Speaker 3: no seasonality. I mean, it's growing, you know, triple digits

105
00:06:11,680 --> 00:06:14,960
Speaker 3: so obviously you can't factor in any seasonality. This quarter,

106
00:06:15,240 --> 00:06:18,400
Speaker 3: the margins were seventy five percent. They said they can

107
00:06:18,480 --> 00:06:21,920
Speaker 3: maintain that mid seventy percent gross margin. But the reason

108
00:06:21,960 --> 00:06:25,400
Speaker 3: why gross margin may trend down going forward is because

109
00:06:25,400 --> 00:06:27,359
Speaker 3: they have to buy more HBM chips.

110
00:06:27,600 --> 00:06:28,120
Speaker 4: They have to.

111
00:06:28,120 --> 00:06:32,680
Speaker 3: Buy more advance packaging capacity from TSMC. So that's where

112
00:06:32,720 --> 00:06:36,040
Speaker 3: the bill of materials does add up for Nvidia and

113
00:06:36,120 --> 00:06:38,320
Speaker 3: that will start to impact their gross margins.

114
00:06:38,360 --> 00:06:41,560
Speaker 1: Great conversation, Man, Deep, Thank you so much. Bloomberg Intelligence

115
00:06:41,600 --> 00:06:45,320
Speaker 1: senior tech analyst Mandeep Singh joining us here as we

116
00:06:45,360 --> 00:06:47,960
Speaker 1: look at the price action in shares of Nvidia in

117
00:06:48,080 --> 00:06:52,520
Speaker 1: late trading down nearly seven percent. I'm Doug Krisner, and

118
00:06:52,640 --> 00:07:04,279
Speaker 1: this is Bloomberg. We welcome our guest, Mario Morales, Group

119
00:07:04,440 --> 00:07:10,679
Speaker 1: VP of Enabling Technologies and Semiconductors at IDC. He's joining

120
00:07:10,760 --> 00:07:14,760
Speaker 1: us from Silicon Valley, right down the road from San Jose. Mario,

121
00:07:14,800 --> 00:07:16,800
Speaker 1: good of you to make time. What was your reaction

122
00:07:16,880 --> 00:07:18,800
Speaker 1: to what you heard tonight from in Nvidia.

123
00:07:20,320 --> 00:07:23,080
Speaker 5: Well, I think I'm a lot more positive than what

124
00:07:23,120 --> 00:07:26,480
Speaker 5: I've been seeing after the market closed for in video.

125
00:07:27,000 --> 00:07:30,120
Speaker 5: I think when you look at what they basically shared

126
00:07:30,120 --> 00:07:33,600
Speaker 5: with us is that they're still growing pretty phenomenally in

127
00:07:33,640 --> 00:07:36,480
Speaker 5: the data center. I think the other markets that they

128
00:07:36,560 --> 00:07:40,760
Speaker 5: serve are now returning back to more normalized growth. So

129
00:07:40,800 --> 00:07:42,840
Speaker 5: I think those are all positives. I think that the

130
00:07:42,880 --> 00:07:46,320
Speaker 5: market sentiment was a little bit down just because the

131
00:07:46,360 --> 00:07:48,920
Speaker 5: expectations have been so high for the company.

132
00:07:49,400 --> 00:07:51,200
Speaker 1: What do you make of the fact that the company

133
00:07:51,280 --> 00:07:54,880
Speaker 1: is having problems with this new Blackwell chip design? Is

134
00:07:54,880 --> 00:07:55,800
Speaker 1: that a concern to you?

135
00:07:57,520 --> 00:08:00,200
Speaker 5: Not really, because I look at the position and they

136
00:08:00,200 --> 00:08:03,120
Speaker 5: have such a dominant position right now. Even if they're

137
00:08:03,160 --> 00:08:05,680
Speaker 5: late by a quarter or two, it still will not

138
00:08:05,800 --> 00:08:09,080
Speaker 5: change the pecking order in terms of the incumbency that

139
00:08:09,120 --> 00:08:12,280
Speaker 5: they continue to enjoy within the data center market. I

140
00:08:12,320 --> 00:08:15,200
Speaker 5: think what you're looking at nowadays is that these chips,

141
00:08:15,240 --> 00:08:19,520
Speaker 5: whether it's the Black Blackwell platform or any other that's

142
00:08:19,560 --> 00:08:22,720
Speaker 5: in the market, they're very complex chips, and they're basically

143
00:08:22,760 --> 00:08:26,280
Speaker 5: now putting all of these chiplets together in one very

144
00:08:26,280 --> 00:08:29,960
Speaker 5: intricate package, and so you're going to have some yield

145
00:08:29,960 --> 00:08:33,360
Speaker 5: issues initially when you ramp some of these new technologies.

146
00:08:34,160 --> 00:08:37,480
Speaker 1: I know that about forty five percent of Nvidia's total

147
00:08:37,520 --> 00:08:41,480
Speaker 1: business comes from the so called hyperscaleers, the companies that

148
00:08:41,559 --> 00:08:45,719
Speaker 1: invest in cloud computing names like Alphabet. Amazon is in there,

149
00:08:45,760 --> 00:08:48,600
Speaker 1: along with Microsoft, IBM, you could even add Oracle. I

150
00:08:48,640 --> 00:08:52,480
Speaker 1: think if it's forty five percent of the business that

151
00:08:52,559 --> 00:08:55,720
Speaker 1: has been really restricted to companies that have already built

152
00:08:55,760 --> 00:08:59,840
Speaker 1: out their AI infrastructure, isn't it reasonable to think that

153
00:09:00,080 --> 00:09:03,520
Speaker 1: growth is going to begin tapering off because those companies

154
00:09:03,559 --> 00:09:06,440
Speaker 1: really at this point aren't in a position to spend more.

155
00:09:06,480 --> 00:09:07,199
Speaker 1: They have enough.

156
00:09:09,120 --> 00:09:11,440
Speaker 5: You know, when you look at the cloud service provider

157
00:09:11,480 --> 00:09:16,040
Speaker 5: space and how they procure silicon, it tends to be lumpy.

158
00:09:16,160 --> 00:09:20,400
Speaker 5: But the investment in AI has been so strong that

159
00:09:20,480 --> 00:09:23,880
Speaker 5: I think it can continue on. Maybe not necessarily only

160
00:09:23,920 --> 00:09:26,480
Speaker 5: by the cloud service providers, but I think the next

161
00:09:26,520 --> 00:09:30,040
Speaker 5: wave that's going to come will be from telcos and enterprise.

162
00:09:30,160 --> 00:09:33,400
Speaker 5: These are all industries that are very large that are

163
00:09:33,440 --> 00:09:37,839
Speaker 5: just beginning to experiment with AI and want to really

164
00:09:38,200 --> 00:09:41,880
Speaker 5: turn it on in a big way across their enterprises.

165
00:09:41,920 --> 00:09:44,600
Speaker 5: So I see more growth ahead. I think the cloud

166
00:09:44,640 --> 00:09:48,120
Speaker 5: service corriders might take a pause maybe over the coming quarters,

167
00:09:48,160 --> 00:09:51,200
Speaker 5: but I still believe that fundamentally on a longer term basis,

168
00:09:51,200 --> 00:09:53,760
Speaker 5: we're just still you know, maybe in the second inning

169
00:09:54,000 --> 00:09:56,360
Speaker 5: of what we're seeing today in terms of the amount

170
00:09:56,400 --> 00:09:58,000
Speaker 5: of investment that we're seeing in AI.

171
00:09:58,559 --> 00:10:02,199
Speaker 1: So Nvidia designs that they are manufactured at a foundry

172
00:10:02,240 --> 00:10:06,600
Speaker 1: that's operated by TSMC. That's essentially what TSMC does, they're

173
00:10:06,679 --> 00:10:10,680
Speaker 1: the manufacturer. Are you concerned that TSMC may not be

174
00:10:10,760 --> 00:10:12,120
Speaker 1: able to keep up with demand.

175
00:10:13,679 --> 00:10:16,280
Speaker 5: Yeah, I mean that's continued to be the theme right

176
00:10:16,320 --> 00:10:18,600
Speaker 5: for over a year now. I think last year we

177
00:10:18,640 --> 00:10:21,640
Speaker 5: saw a correction in the semic metor space, but when

178
00:10:21,679 --> 00:10:24,160
Speaker 5: you look at TSMC, they tend to buck the trend.

179
00:10:24,360 --> 00:10:25,160
Speaker 4: They're a very.

180
00:10:26,480 --> 00:10:29,120
Speaker 5: Strong player in the founder space. The leader they control

181
00:10:29,160 --> 00:10:32,560
Speaker 5: over sixty percent of the market. Where they're investing heavily

182
00:10:32,640 --> 00:10:35,959
Speaker 5: now is in advanced packaging. So when you hear about costs,

183
00:10:36,679 --> 00:10:39,080
Speaker 5: all of that means is that that's where they're seeing

184
00:10:39,400 --> 00:10:41,680
Speaker 5: the big investments that they need to make in order

185
00:10:41,720 --> 00:10:45,040
Speaker 5: to serve companies like Nvidia and Broadcom and AMD and

186
00:10:45,120 --> 00:10:49,000
Speaker 5: others that are now beginning to use more advanced packaging

187
00:10:49,000 --> 00:10:50,880
Speaker 5: in order to continue to drive integration.

188
00:10:51,559 --> 00:10:56,599
Speaker 1: So the GPUs that are designed by Nvidio manufactured by TSMC,

189
00:10:57,200 --> 00:10:59,959
Speaker 1: they have to be paired with the high bandwidth memory

190
00:11:00,160 --> 00:11:04,280
Speaker 1: chips do they not, and those devices or chips are

191
00:11:04,679 --> 00:11:08,079
Speaker 1: created by companies like Samsung and s k Heinex in

192
00:11:08,120 --> 00:11:12,440
Speaker 1: South Korea. How are those companies How have they been

193
00:11:12,480 --> 00:11:14,760
Speaker 1: able to keep up with the demand. Should we be

194
00:11:14,840 --> 00:11:18,400
Speaker 1: concerned about maybe a little weakness on that side.

195
00:11:19,200 --> 00:11:22,640
Speaker 5: Well, we're not seeing the weakness. In fact, since probably

196
00:11:22,640 --> 00:11:27,280
Speaker 5: September of last year, the memory companies began to change

197
00:11:27,320 --> 00:11:29,960
Speaker 5: their operations in order to get back to profitability, but

198
00:11:30,400 --> 00:11:34,000
Speaker 5: demand returned, especially from the data center, and today we're

199
00:11:34,000 --> 00:11:37,920
Speaker 5: expecting that both DRAM and NAND will grow significantly from

200
00:11:37,960 --> 00:11:41,040
Speaker 5: a revenue basis, close to seventy three percent in terms

201
00:11:41,040 --> 00:11:43,880
Speaker 5: of revenue growth, so very phenomenal growth. A lot of

202
00:11:43,920 --> 00:11:46,959
Speaker 5: that led not just by high bandwidth memory, but also

203
00:11:47,320 --> 00:11:50,719
Speaker 5: DDR five that's being used in the flagship smartphones and

204
00:11:51,240 --> 00:11:55,280
Speaker 5: the upcoming aipcs that we'll see later during the holiday season.

205
00:11:55,720 --> 00:11:58,000
Speaker 5: So I think in general that the demand for memory

206
00:11:58,040 --> 00:12:02,120
Speaker 5: will still remain robut this year and likely into twenty

207
00:12:02,160 --> 00:12:02,720
Speaker 5: twenty five.

208
00:12:03,200 --> 00:12:05,480
Speaker 1: One of the things that is simply stunning when you

209
00:12:05,480 --> 00:12:08,240
Speaker 1: look at any area of high technology is the rate

210
00:12:08,320 --> 00:12:13,320
Speaker 1: of change, how quickly technology evolves. Give me a sense

211
00:12:13,320 --> 00:12:15,880
Speaker 1: of what we're seeing now with not only a company

212
00:12:16,000 --> 00:12:19,400
Speaker 1: like in Nvidia, but broadly when it comes to companies

213
00:12:19,440 --> 00:12:22,960
Speaker 1: that are involved in artificial intelligence. Is some of this

214
00:12:24,640 --> 00:12:27,240
Speaker 1: equipment going to be obsolete a year from now? Do

215
00:12:27,280 --> 00:12:27,680
Speaker 1: you think?

216
00:12:29,000 --> 00:12:29,240
Speaker 4: Well?

217
00:12:29,760 --> 00:12:32,520
Speaker 5: You definitely. When you look at the SEMAC, nintor space,

218
00:12:32,559 --> 00:12:36,000
Speaker 5: it's always followed Moore's law, which means that that performance

219
00:12:36,080 --> 00:12:39,760
Speaker 5: or power changes every eighteen to twenty four months. Now

220
00:12:39,840 --> 00:12:43,440
Speaker 5: you add AI to this, and when you start looking

221
00:12:43,440 --> 00:12:46,559
Speaker 5: more at the large language models, they're actually changing every

222
00:12:46,600 --> 00:12:49,880
Speaker 5: three to four months. So we've seen an acceleration of

223
00:12:50,080 --> 00:12:53,240
Speaker 5: change that's happening. And you look at companies like in

224
00:12:53,320 --> 00:12:57,320
Speaker 5: Nvidia or AMD, they're trying to go and keep up

225
00:12:57,320 --> 00:12:59,440
Speaker 5: with that change. And I think that's going to be

226
00:12:59,640 --> 00:13:03,119
Speaker 5: the big challenge for these companies because they've been operating

227
00:13:03,920 --> 00:13:06,960
Speaker 5: with More's law in mind. But these but the market

228
00:13:07,040 --> 00:13:10,080
Speaker 5: for AI is moving so much faster. So now AMD

229
00:13:10,360 --> 00:13:14,520
Speaker 5: and Nvidia and Intel now have to basically introduce products

230
00:13:14,679 --> 00:13:17,160
Speaker 5: every single year now, and you're going to have some

231
00:13:17,240 --> 00:13:20,199
Speaker 5: hiccups because of that, because the cadence is now changing

232
00:13:20,240 --> 00:13:22,360
Speaker 5: in order to keep up with this demand and this

233
00:13:22,440 --> 00:13:26,520
Speaker 5: appetite that the world has for AI, especially around infrastructure.

234
00:13:26,720 --> 00:13:29,560
Speaker 1: What is your sense of the competition that in Vidia

235
00:13:29,720 --> 00:13:32,199
Speaker 1: may face in the future. Right now, there is clearly

236
00:13:32,320 --> 00:13:35,679
Speaker 1: very little. The company is essentially a monopolist when it

237
00:13:35,679 --> 00:13:38,760
Speaker 1: comes to these GPUs. But can you imagine a world

238
00:13:38,800 --> 00:13:41,839
Speaker 1: where Nvidia begins to look over at shoulder a bit?

239
00:13:43,640 --> 00:13:46,040
Speaker 5: I think so. I think over the coming year, you're

240
00:13:46,040 --> 00:13:49,640
Speaker 5: definitely going to see AAMD stepping up their game. They

241
00:13:49,760 --> 00:13:52,559
Speaker 5: you know, back in computext in June they announced an

242
00:13:52,559 --> 00:13:56,040
Speaker 5: annual cadence for their own accelerators with the M three

243
00:13:56,120 --> 00:13:58,120
Speaker 5: hundred beginning to fray, and then you got the three

244
00:13:58,200 --> 00:14:00,680
Speaker 5: twenty five and three fifty. So you're going to see

245
00:14:00,720 --> 00:14:04,320
Speaker 5: AMD now introducing an annual cadence of accelerators in order

246
00:14:04,679 --> 00:14:07,520
Speaker 5: to try to catch up to Nvidia. But I think

247
00:14:07,840 --> 00:14:12,280
Speaker 5: in Nvidia's own customers continue to chip away little by

248
00:14:12,320 --> 00:14:15,480
Speaker 5: little in terms of the instances and services that they're

249
00:14:15,520 --> 00:14:18,080
Speaker 5: turning on with their own silicon, And so when I

250
00:14:18,120 --> 00:14:21,240
Speaker 5: look at the competition, I think it's going to come

251
00:14:21,320 --> 00:14:24,920
Speaker 5: more so from the customers. Initially, so a company like

252
00:14:25,000 --> 00:14:31,840
Speaker 5: Google and Microsoft and AWS all developing their own chips

253
00:14:32,240 --> 00:14:34,840
Speaker 5: and using some of those ships with their infrastructure, and

254
00:14:34,880 --> 00:14:37,880
Speaker 5: then on a longer term basis. Broadcom is still a

255
00:14:37,920 --> 00:14:40,720
Speaker 5: significant player here that not a lot of people talk about.

256
00:14:40,720 --> 00:14:42,960
Speaker 5: But when you think about a lot of the designs

257
00:14:42,960 --> 00:14:46,480
Speaker 5: of the APIs that are today serving the cloud service providers,

258
00:14:46,640 --> 00:14:49,080
Speaker 5: a lot of those acs are designed in collaboration with

259
00:14:49,160 --> 00:14:51,760
Speaker 5: Brock them, so I would call them the number two player.

260
00:14:51,960 --> 00:14:55,680
Speaker 5: And then AMD coming in being able to leverage their

261
00:14:55,720 --> 00:15:00,040
Speaker 5: CPU entrenchmen, especially in the data center, and hopefully that

262
00:15:00,080 --> 00:15:05,480
Speaker 5: will help accelerate their own entrance into the GPU market

263
00:15:05,520 --> 00:15:07,040
Speaker 5: for training and infancy.

264
00:15:07,600 --> 00:15:10,960
Speaker 1: Mario very quickly in about thirty seconds. A big semiconductor

265
00:15:10,960 --> 00:15:14,080
Speaker 1: industry a conference next week in Taiwan. What do you

266
00:15:14,120 --> 00:15:18,280
Speaker 1: think the big issues are for the semiconductors supply chain

267
00:15:18,520 --> 00:15:19,640
Speaker 1: in Asia right now?

268
00:15:20,840 --> 00:15:23,640
Speaker 5: I think that we're still seeing a slower recovery in

269
00:15:24,040 --> 00:15:27,600
Speaker 5: the spending and I'm talking about capital spending of the

270
00:15:27,760 --> 00:15:31,320
Speaker 5: major companies. Even though TSMC has increased their numbers, the

271
00:15:31,720 --> 00:15:34,360
Speaker 5: rest of the industry is still a lot more cautious

272
00:15:34,640 --> 00:15:36,360
Speaker 5: around spending. So that's going to be one of the

273
00:15:36,440 --> 00:15:39,760
Speaker 5: key things. How much of the tools are still making

274
00:15:39,800 --> 00:15:42,440
Speaker 5: their way into China. That is also becoming something that

275
00:15:43,040 --> 00:15:47,280
Speaker 5: governments are watching more closely. And then how long can

276
00:15:47,320 --> 00:15:50,600
Speaker 5: we continue to sustain the growth that is needed to

277
00:15:50,600 --> 00:15:52,080
Speaker 5: believe it there billion dollars.

278
00:15:52,200 --> 00:16:01,640
Speaker 1: Mario Morales, this is Bloomberg. We bring in our guest,

279
00:16:01,760 --> 00:16:05,040
Speaker 1: Dan Newman. He is the CEO of the Futuram Group,

280
00:16:05,160 --> 00:16:08,320
Speaker 1: joining us from Austin, Texas. Dan, pleasure to have you

281
00:16:08,360 --> 00:16:10,440
Speaker 1: on the program. I hope you had some time to

282
00:16:10,480 --> 00:16:13,400
Speaker 1: go through what we heard from Nvidia tonight. The stock,

283
00:16:13,440 --> 00:16:15,360
Speaker 1: as I mentioned, was down more than seven percent in

284
00:16:15,360 --> 00:16:17,920
Speaker 1: the late US trading. Do you think this is an overreaction?

285
00:16:19,560 --> 00:16:21,520
Speaker 4: Yeah, thanks for having me, and I have spent some

286
00:16:21,560 --> 00:16:24,960
Speaker 4: time and actually listened to a really compelling interview that

287
00:16:25,040 --> 00:16:29,280
Speaker 4: was done Bloomberg with Ed Ludlow. I think the overall

288
00:16:29,600 --> 00:16:32,240
Speaker 4: reaction is a little bit of what I would have

289
00:16:32,320 --> 00:16:36,040
Speaker 4: expected based on them just doing enough. And as crazy

290
00:16:36,080 --> 00:16:40,080
Speaker 4: as this sounds, just doing enough was beating raising and

291
00:16:40,120 --> 00:16:44,000
Speaker 4: being really solid and consistent in what the company is delivering.

292
00:16:44,120 --> 00:16:46,400
Speaker 4: The problem is, over the last several quarters, we've come

293
00:16:46,440 --> 00:16:50,680
Speaker 4: to expect these massive beats, huge revenue upside, these guides

294
00:16:50,720 --> 00:16:53,760
Speaker 4: that look phenomenal, and this sort of endless upgrade cycle

295
00:16:54,160 --> 00:16:57,000
Speaker 4: and this just was good. And you know what I

296
00:16:57,080 --> 00:17:00,160
Speaker 4: said coming into this, it has to be great or

297
00:17:00,160 --> 00:17:00,800
Speaker 4: there's only.

298
00:17:00,640 --> 00:17:04,080
Speaker 1: Downside the price to perfection story. I think we can

299
00:17:04,119 --> 00:17:06,680
Speaker 1: agree on that, right. But are you troubled at all

300
00:17:06,720 --> 00:17:09,880
Speaker 1: by the word that in Nvidia is having problems with

301
00:17:09,920 --> 00:17:11,560
Speaker 1: that new Blackwell chip design.

302
00:17:13,560 --> 00:17:16,800
Speaker 4: I'm not really troubled because it seems he has the

303
00:17:16,880 --> 00:17:20,320
Speaker 4: situation under control. He talked about the update to the mask,

304
00:17:20,400 --> 00:17:23,000
Speaker 4: which was going to improve the yield, which in the

305
00:17:23,720 --> 00:17:27,000
Speaker 4: in the near term means probably an uplift from Hopper.

306
00:17:27,520 --> 00:17:31,440
Speaker 4: These Hopper chips are still very significant, and they're still

307
00:17:31,440 --> 00:17:35,160
Speaker 4: selling in volume. Amazon and Microsoft are still building out

308
00:17:35,160 --> 00:17:38,320
Speaker 4: significant infrastructure with them, so is x and Tesla and others.

309
00:17:38,560 --> 00:17:40,359
Speaker 4: I think where this is going to head, though, is

310
00:17:40,600 --> 00:17:44,240
Speaker 4: when does that ramp, that one year cycle continue to

311
00:17:44,320 --> 00:17:46,760
Speaker 4: drive these triple digit growth rates. And we're starting to

312
00:17:46,760 --> 00:17:49,400
Speaker 4: see that slow down. And I think whether we see

313
00:17:49,400 --> 00:17:51,480
Speaker 4: this ship into the fourth quarter or even if it

314
00:17:51,520 --> 00:17:54,040
Speaker 4: falls into the first quarter, what we end up having

315
00:17:54,080 --> 00:17:56,720
Speaker 4: happen is we've got an entire world that's kind of

316
00:17:56,760 --> 00:17:59,680
Speaker 4: watching this stock and saying triple digit revenue growth, triple

317
00:17:59,680 --> 00:18:01,600
Speaker 4: digit earnings growth, and now we're starting to see that

318
00:18:01,680 --> 00:18:05,000
Speaker 4: quarter on quarter you're on youar it's decelerating and then

319
00:18:05,080 --> 00:18:09,040
Speaker 4: bigger questions about AI digestion and when does this start

320
00:18:09,080 --> 00:18:11,600
Speaker 4: to really find its way into companies like Salesforce, as

321
00:18:11,680 --> 00:18:14,240
Speaker 4: Numbers and others that reported today that didn't get as

322
00:18:14,280 --> 00:18:14,720
Speaker 4: much attention.

323
00:18:14,920 --> 00:18:17,000
Speaker 1: Yeah, you're right about that. It's a good point too.

324
00:18:17,040 --> 00:18:20,560
Speaker 1: But let's get back to Nvidia and the approved fifty

325
00:18:20,600 --> 00:18:24,160
Speaker 1: billion dollars in additional stock buyback. Is that a good

326
00:18:24,200 --> 00:18:27,960
Speaker 1: way to kind of repurpose or redeploy capital.

327
00:18:29,359 --> 00:18:31,760
Speaker 4: I think if the company is looking at the situation

328
00:18:31,840 --> 00:18:35,000
Speaker 4: as hey, there may be some some hesitancy, there may

329
00:18:35,040 --> 00:18:37,399
Speaker 4: be some sellers. We saw some selling pressure into the

330
00:18:37,480 --> 00:18:40,240
Speaker 4: numbers today. The company wants to be armed, to be

331
00:18:40,280 --> 00:18:43,080
Speaker 4: able to lower the flow and to be able to

332
00:18:43,200 --> 00:18:46,000
Speaker 4: take advantage of a lower share price as it expects

333
00:18:46,040 --> 00:18:49,400
Speaker 4: that this Blackwell and then followed by Rubin will continue

334
00:18:49,400 --> 00:18:52,240
Speaker 4: to drive this acceleration. You know, we did hear from

335
00:18:52,359 --> 00:18:56,200
Speaker 4: Mark Luckerberg center of a Chai, you know, Satia Nadella.

336
00:18:56,320 --> 00:18:58,840
Speaker 4: There was really no indication that they plan to slow

337
00:18:58,880 --> 00:19:03,159
Speaker 4: their capex spending. And that means good things for Nvidia

338
00:19:03,280 --> 00:19:05,280
Speaker 4: because despite the fact that all of them are making

339
00:19:05,320 --> 00:19:09,440
Speaker 4: investments or have been making investments in somewhat competitive offerings.

340
00:19:09,480 --> 00:19:13,040
Speaker 4: In the current moment, these companies are building the bridges

341
00:19:13,119 --> 00:19:16,160
Speaker 4: and roads and tunnels with Nvidio, and if they want

342
00:19:16,200 --> 00:19:18,520
Speaker 4: to be competitive, they don't want their Blockbuster moment, they

343
00:19:18,520 --> 00:19:20,960
Speaker 4: don't want to have their BlackBerry moments. So I think

344
00:19:21,000 --> 00:19:22,919
Speaker 4: they are going to keep investing. And I think that

345
00:19:22,960 --> 00:19:25,160
Speaker 4: bude really well for Justin Wong in video.

346
00:19:25,400 --> 00:19:27,040
Speaker 1: So I hear what you're saying when it comes to

347
00:19:27,040 --> 00:19:30,880
Speaker 1: the cloud computing companies, those so called hyperscalers talking about

348
00:19:31,320 --> 00:19:34,600
Speaker 1: their intention to continue with the current level of cap x.

349
00:19:34,920 --> 00:19:38,840
Speaker 1: But won't that largely be determined by whether or they

350
00:19:39,440 --> 00:19:42,320
Speaker 1: or not they get a significant return on their investment.

351
00:19:42,359 --> 00:19:46,880
Speaker 1: The so called ROI on AHI has to be the key, right.

352
00:19:48,160 --> 00:19:50,520
Speaker 4: You would think so. But if you listen to Mark

353
00:19:50,600 --> 00:19:54,000
Speaker 4: Zuckerberg's comments, he was pretty clear that he was willing

354
00:19:54,080 --> 00:19:57,160
Speaker 4: to bet even if that ingestion period, even if that

355
00:19:57,600 --> 00:20:00,760
Speaker 4: consumption period is longer than we expect. And we know

356
00:20:00,840 --> 00:20:02,960
Speaker 4: that he's done that before with Reality Labs, and the

357
00:20:03,000 --> 00:20:05,280
Speaker 4: market sort of pushed back at him until he had

358
00:20:05,320 --> 00:20:08,119
Speaker 4: to pivot on his as capex investments on that side.

359
00:20:08,200 --> 00:20:11,600
Speaker 4: But AI is not virtual reality. It is very clear

360
00:20:11,640 --> 00:20:15,040
Speaker 4: that this is going to change every industry from autonomous

361
00:20:15,119 --> 00:20:19,119
Speaker 4: driving to drug discovery to these large language models that

362
00:20:19,119 --> 00:20:21,760
Speaker 4: are going to change software and the way we sell,

363
00:20:21,800 --> 00:20:24,080
Speaker 4: the way we buy, the way we watch. And so

364
00:20:24,160 --> 00:20:26,640
Speaker 4: when you put all these things together, I think these

365
00:20:26,640 --> 00:20:29,280
Speaker 4: companies are willing. They have great balance sheets, most of

366
00:20:29,320 --> 00:20:32,240
Speaker 4: them are strong cash positions. They're willing to make the bet.

367
00:20:32,520 --> 00:20:35,119
Speaker 4: But I agree with you, I think we need to

368
00:20:35,160 --> 00:20:39,840
Speaker 4: see more measurable returns downstream from these software companies. Who

369
00:20:39,880 --> 00:20:43,360
Speaker 4: else is making money besides Nvidia and these cloud companies

370
00:20:43,359 --> 00:20:45,639
Speaker 4: have not been really clear. I think Azure amy Hood

371
00:20:45,640 --> 00:20:47,639
Speaker 4: gave the best numbers so far, but you're talking to

372
00:20:47,680 --> 00:20:49,840
Speaker 4: single digit percent about Azure revenue.

373
00:20:50,160 --> 00:20:54,360
Speaker 1: TSMC is obviously the foundry that produces these chips. I mean,

374
00:20:54,760 --> 00:20:57,280
Speaker 1: are you confident that TSMC can keep up with a

375
00:20:57,359 --> 00:20:58,760
Speaker 1: level of production required.

376
00:21:00,119 --> 00:21:03,880
Speaker 4: Well, TSMC is continuing to talk to how it will

377
00:21:03,880 --> 00:21:07,480
Speaker 4: expand its capabilities. We've heard about some of the you know,

378
00:21:07,560 --> 00:21:11,320
Speaker 4: the availability of its three danimeter being really sold out

379
00:21:11,359 --> 00:21:14,719
Speaker 4: through twenty twenty five. We of course they're hearing about coosts,

380
00:21:14,760 --> 00:21:17,840
Speaker 4: you know, the packaging technology, and that was part of

381
00:21:17,920 --> 00:21:20,480
Speaker 4: what was going on with some of the potential delays.

382
00:21:20,800 --> 00:21:24,359
Speaker 4: This new cost l technology and TSMC has the market.

383
00:21:24,400 --> 00:21:27,919
Speaker 4: They have a huge percentage of this AI movement, and

384
00:21:27,960 --> 00:21:30,439
Speaker 4: of course a lot of Nvidia's business, and they're going

385
00:21:30,520 --> 00:21:33,880
Speaker 4: to have to invest substantially. TSMC is a clear winner here,

386
00:21:34,359 --> 00:21:36,040
Speaker 4: and you know, while we are waiting on some of

387
00:21:36,080 --> 00:21:38,800
Speaker 4: these others, the sks and microns, and they're building the

388
00:21:38,920 --> 00:21:42,080
Speaker 4: HBM side this partner ecosystem, and Jenson said this on

389
00:21:42,119 --> 00:21:44,040
Speaker 4: the interview with that ed it was very clear that

390
00:21:44,119 --> 00:21:47,720
Speaker 4: this supply partnership is monumental in terms of the company's

391
00:21:47,720 --> 00:21:50,720
Speaker 4: ability to continue to hit its growth. TSMC has given

392
00:21:50,760 --> 00:21:54,160
Speaker 4: us no reason to doubt its ability. But this technology

393
00:21:54,200 --> 00:21:56,199
Speaker 4: keeps getting more complex and that puts a lot of

394
00:21:56,240 --> 00:21:58,760
Speaker 4: stress on TSMC to deliver well.

395
00:21:58,800 --> 00:22:02,920
Speaker 1: You mentioned the high band with memory chip manufacturers sk Heinix,

396
00:22:03,119 --> 00:22:07,080
Speaker 1: Samsung in South Korea. Are they kind of the weak

397
00:22:07,160 --> 00:22:09,119
Speaker 1: link in this chain, which is to say, if they

398
00:22:09,160 --> 00:22:11,600
Speaker 1: don't keep up with the level of output that in

399
00:22:12,119 --> 00:22:15,560
Speaker 1: Nvidia is striving for that the overall move in the

400
00:22:15,560 --> 00:22:20,119
Speaker 1: build out of some of these AI infrastructure models, I

401
00:22:20,119 --> 00:22:21,480
Speaker 1: mean that it's held back a bit.

402
00:22:22,760 --> 00:22:25,639
Speaker 4: I think this continues to be the reason why whether

403
00:22:25,960 --> 00:22:28,320
Speaker 4: Blackwell's a little bit delayed or not, they're going to

404
00:22:28,359 --> 00:22:32,439
Speaker 4: produce as much of the hopper as possible. The overall

405
00:22:33,040 --> 00:22:36,720
Speaker 4: situation as it pertains to memory is they're also constrained there.

406
00:22:36,720 --> 00:22:40,680
Speaker 4: So you heard me talk about three nanometer constrained, coosts constrained,

407
00:22:40,720 --> 00:22:44,960
Speaker 4: and of course HBM constrained. So right now you are

408
00:22:45,000 --> 00:22:48,880
Speaker 4: literally seeing these five or so major hyperscalers plus global

409
00:22:48,960 --> 00:22:53,199
Speaker 4: hyperscalers and Tier two's taking up the vast majority of

410
00:22:53,200 --> 00:22:55,879
Speaker 4: the demand for all of these major parts of the supply.

411
00:22:56,560 --> 00:22:59,680
Speaker 4: And you know, Jensen talked about it being around forty

412
00:22:59,720 --> 00:23:03,320
Speaker 4: five percent coming from the three or four biggest, but

413
00:23:03,520 --> 00:23:05,800
Speaker 4: they are taking most of the demand and until we

414
00:23:05,840 --> 00:23:09,160
Speaker 4: can expand capacity. So it's on Micron, which is investing,

415
00:23:09,240 --> 00:23:12,160
Speaker 4: it's on sk and Samsung, all of which are investing.

416
00:23:12,440 --> 00:23:15,160
Speaker 4: But it can't happen fast enough. This does take.

417
00:23:15,000 --> 00:23:18,720
Speaker 1: Time, given the export controls that the Biden administration has

418
00:23:18,800 --> 00:23:22,240
Speaker 1: put on some of this advanced technology from Nvidia to

419
00:23:22,320 --> 00:23:25,919
Speaker 1: the Chinese market. Is it fair to say that the

420
00:23:26,040 --> 00:23:28,720
Speaker 1: US is far out in front of China when it

421
00:23:28,760 --> 00:23:33,879
Speaker 1: comes to AI or are you concerned possibly about Chinese

422
00:23:33,920 --> 00:23:36,480
Speaker 1: advancement in this area and maybe at some point it

423
00:23:36,600 --> 00:23:38,200
Speaker 1: becoming a greater rivalry.

424
00:23:40,760 --> 00:23:46,439
Speaker 4: I am always concerned about China's willingness, will power, capacity

425
00:23:46,480 --> 00:23:49,760
Speaker 4: and capabilities to build and deliver even in the most

426
00:23:49,760 --> 00:23:54,000
Speaker 4: difficult situation. Of course, the US has been incredibly innovative,

427
00:23:54,040 --> 00:24:00,000
Speaker 4: are fabulous leaders, not just in Nvidia, AMD, of course, historically, Intel, Qualcom,

428
00:24:00,280 --> 00:24:05,040
Speaker 4: these companies have all been tremendously important to this innovation wave.

429
00:24:05,400 --> 00:24:07,880
Speaker 4: China is doing what it can to follow suit. It's

430
00:24:07,880 --> 00:24:11,040
Speaker 4: been somewhat successful in mobile devices, in PCs. We've seen

431
00:24:11,080 --> 00:24:13,480
Speaker 4: it with Honor and Huawei. We are hearing that it

432
00:24:13,560 --> 00:24:16,520
Speaker 4: is having some level of success building competitive products to

433
00:24:16,600 --> 00:24:19,960
Speaker 4: the tuned down in Vidio chips, the B twenties h twenties,

434
00:24:20,880 --> 00:24:24,159
Speaker 4: that they could be able to compete in this particular space.

435
00:24:24,440 --> 00:24:26,280
Speaker 4: I never would rule them out. I think that would

436
00:24:26,320 --> 00:24:31,119
Speaker 4: be a tremendously large mistake. Having said that, in Vidia

437
00:24:31,200 --> 00:24:34,320
Speaker 4: has a substantial lead beyond everyone else right now, and

438
00:24:34,359 --> 00:24:37,399
Speaker 4: then others like AMD and like what the hyperscalers are

439
00:24:37,400 --> 00:24:39,879
Speaker 4: building is very advanced, and I do think China has

440
00:24:39,880 --> 00:24:43,360
Speaker 4: significant challenges, but I wouldn't doubt their ability or resolve

441
00:24:43,720 --> 00:24:44,440
Speaker 4: to find a way.

442
00:24:44,680 --> 00:24:46,760
Speaker 1: Well, leave it there, Dan, thank you so much. Always

443
00:24:46,760 --> 00:24:49,080
Speaker 1: a pleasure Dan Newman, he is the CEO of the

444
00:24:49,200 --> 00:24:54,200
Speaker 1: Futurum Group, joining us from Austin, Texas here on Daybreak Asia.

445
00:24:54,280 --> 00:24:57,680
Speaker 2: This is the Bloomberg Daybreak Asia podcast, bringing you the

446
00:24:57,760 --> 00:25:01,240
Speaker 2: stories making news and moving markets in the Asia Pacific.

447
00:25:01,480 --> 00:25:04,600
Speaker 2: Visit the Bloomberg Podcast channel on YouTube to get more

448
00:25:04,640 --> 00:25:08,480
Speaker 2: episodes of this and other shows from Bloomberg. Subscribe to

449
00:25:08,520 --> 00:25:12,320
Speaker 2: the podcast on Apple, Spotify, or anywhere else you listen,

450
00:25:12,560 --> 00:25:15,800
Speaker 2: and always on Bloomberg Radio, the Bloomberg Terminal, and the

451
00:25:15,800 --> 00:25:17,040
Speaker 2: Bloomberg Business App.