1
00:00:02,520 --> 00:00:08,039
Speaker 1: Bloomberg Audio Studios, podcasts, radio news fourth.

2
00:00:07,840 --> 00:00:11,200
Speaker 2: Quarter revenue has beat estimates. Fourth quarter data center revenue

3
00:00:11,200 --> 00:00:14,080
Speaker 2: coming in at thirty five point six billion dollars thirty

4
00:00:14,080 --> 00:00:17,279
Speaker 2: four point oh nine billion dollars. Fourth quarter revenue that

5
00:00:17,360 --> 00:00:21,160
Speaker 2: top line thirty nine point three billion dollars, exceeding estimates

6
00:00:21,160 --> 00:00:24,680
Speaker 2: of thirty eight point twenty five billion dollars. Now coming

7
00:00:24,720 --> 00:00:27,880
Speaker 2: to that revenue estimate for the first quarter. In Video

8
00:00:27,960 --> 00:00:31,960
Speaker 2: C's first quarter revenue plus or minus two percent, uh,

9
00:00:32,040 --> 00:00:35,440
Speaker 2: the estimate was for forty two point twenty six billion dollars.

10
00:00:35,479 --> 00:00:38,040
Speaker 2: Fourth quarter data center revenue coming in at thirty five

11
00:00:38,080 --> 00:00:41,360
Speaker 2: point six percent, the company saying that it ramped up

12
00:00:41,400 --> 00:00:44,560
Speaker 2: massive scale production of Blackwell AI. In the after hour

13
00:00:44,640 --> 00:00:47,160
Speaker 2: shares hire right now by about three point four percent.

14
00:00:47,320 --> 00:00:51,000
Speaker 3: Yeah, actually in Video saying demand for Blackwell is amazing,

15
00:00:51,120 --> 00:00:53,520
Speaker 3: So that's what we talked about in terms of that transition.

16
00:00:54,120 --> 00:00:56,800
Speaker 3: But again the key metrics, as Tim said, stock is

17
00:00:56,880 --> 00:00:59,639
Speaker 3: up almost four percent here in the aftermarket. The first

18
00:00:59,720 --> 00:01:04,559
Speaker 3: quarter revenue, so looking ahead outlook forty three billion dollars

19
00:01:04,600 --> 00:01:07,720
Speaker 3: plus or minus two percent. The estimate was for forty

20
00:01:07,760 --> 00:01:10,520
Speaker 3: two point three billion, and again fourth quarter data center

21
00:01:10,560 --> 00:01:13,319
Speaker 3: so looking at the quarter that they just wrapped up

22
00:01:13,480 --> 00:01:15,759
Speaker 3: data center revenue. That's the bulk of their business, that's

23
00:01:15,840 --> 00:01:19,520
Speaker 3: those hyperscalers and more. Thirty five point six billion. That

24
00:01:19,680 --> 00:01:22,520
Speaker 3: is a beat. Thirty four point zero nine billion was

25
00:01:22,560 --> 00:01:25,679
Speaker 3: what the street was expecting first quarter, justin gross margin

26
00:01:25,760 --> 00:01:29,640
Speaker 3: again looking forward seventy point five percent to seventy one

27
00:01:29,640 --> 00:01:32,200
Speaker 3: point five percent. Street was looking for a little bit more.

28
00:01:32,240 --> 00:01:35,000
Speaker 3: This is a measure of profitability, and the estimate on

29
00:01:35,000 --> 00:01:37,000
Speaker 3: the street was seventy two point one percent. Again, the

30
00:01:37,040 --> 00:01:39,319
Speaker 3: stock now not up as much, In fact, it's down

31
00:01:39,480 --> 00:01:39,880
Speaker 3: just a hair.

32
00:01:40,000 --> 00:01:41,880
Speaker 2: Yeah, shares climbed as much as four percent after this

33
00:01:41,920 --> 00:01:44,440
Speaker 2: first quarter revenue forecast top estimates. We're getting some more

34
00:01:44,480 --> 00:01:47,800
Speaker 2: commentary too, in Vidia saying that Blackwell is achieving billions

35
00:01:47,840 --> 00:01:50,200
Speaker 2: of dollars in sales in the first quarter. The company

36
00:01:50,240 --> 00:01:53,920
Speaker 2: also saying that it's successfully ramped up Blackwell massive production.

37
00:01:54,400 --> 00:01:57,520
Speaker 2: In Vidia also coming out and saying, as Carol mentioned,

38
00:01:57,560 --> 00:02:01,600
Speaker 2: demand for Blackwell is amazing. We're continuing to get more

39
00:02:01,680 --> 00:02:04,720
Speaker 2: numbers here too. Fourth quarter compute revenue coming in above

40
00:02:04,800 --> 00:02:07,680
Speaker 2: estimates at thirty two point five six billion dollars. Fourth

41
00:02:07,760 --> 00:02:10,720
Speaker 2: quarter networking revenue coming in just shy of estimates a

42
00:02:10,800 --> 00:02:12,320
Speaker 2: three point oh two billion dollars.

43
00:02:12,440 --> 00:02:15,079
Speaker 3: Ye, now the stock is down about one percent here,

44
00:02:15,120 --> 00:02:17,200
Speaker 3: it's bouncing around a lot, not surprising a lot of

45
00:02:17,240 --> 00:02:19,320
Speaker 3: investors in this name. So you're going to see a

46
00:02:19,360 --> 00:02:22,399
Speaker 3: lot of movement as we get tick by tick, headline

47
00:02:22,400 --> 00:02:25,400
Speaker 3: by headline. But again, the key ones right now, it

48
00:02:25,440 --> 00:02:28,800
Speaker 3: has to do with the fourth quarter data center revenue.

49
00:02:28,800 --> 00:02:31,760
Speaker 3: Excuse me, thirty five point six billion. Again, that's above

50
00:02:31,800 --> 00:02:34,440
Speaker 3: what the Street was expecting. Street was looking for thirty

51
00:02:34,440 --> 00:02:37,240
Speaker 3: four point zero nine billion. And again key is the

52
00:02:37,280 --> 00:02:40,840
Speaker 3: revenue outlook forty three point zero billion plus or minus

53
00:02:40,840 --> 00:02:44,280
Speaker 3: two percent. Street was at forty two point three billion.

54
00:02:44,360 --> 00:02:44,880
Speaker 4: And again we.

55
00:02:44,919 --> 00:02:47,760
Speaker 3: Got some context context, if you will, in terms of

56
00:02:47,800 --> 00:02:51,520
Speaker 3: demand for Blackwell being amazing, successfully ramping up the Blackwell

57
00:02:51,520 --> 00:02:54,560
Speaker 3: massive production, Blackwell achieving billions of dollars in sales in

58
00:02:54,560 --> 00:02:57,080
Speaker 3: the first quarter and ramping up the massive scale production

59
00:02:57,200 --> 00:02:58,040
Speaker 3: of Blackwell AI.

60
00:02:58,080 --> 00:02:59,520
Speaker 2: If the company else is saying it to lilivered eleven

61
00:02:59,520 --> 00:03:02,959
Speaker 2: billion dollars of Blackwell architecture revenue in the fourth quarter.

62
00:03:03,400 --> 00:03:05,960
Speaker 3: Joining us right now, Red Brown is here in our

63
00:03:06,000 --> 00:03:09,160
Speaker 3: New York studio, Bloomberg News Earnings reporter Caroline Hyde, co

64
00:03:09,200 --> 00:03:14,280
Speaker 3: host of Bloomberg Technology, also here in studio. You know, Red,

65
00:03:14,280 --> 00:03:16,120
Speaker 3: I want to bring you into this conversation. You keep

66
00:03:16,120 --> 00:03:18,480
Speaker 3: a watch on earnings, just you know, your initial thoughts

67
00:03:18,600 --> 00:03:20,000
Speaker 3: and what we're seeing in the trade because the stock

68
00:03:20,080 --> 00:03:21,040
Speaker 3: is bouncing around a lot.

69
00:03:21,200 --> 00:03:22,560
Speaker 5: Yeah, I mean I do feel like we got to

70
00:03:22,600 --> 00:03:25,120
Speaker 5: like kind of take a cyber relief here because they

71
00:03:25,160 --> 00:03:27,800
Speaker 5: did check off all of the boxes kind of what

72
00:03:27,880 --> 00:03:31,440
Speaker 5: Kun John answer the question marks Kun John has kind

73
00:03:31,440 --> 00:03:34,440
Speaker 5: of touched on there. The outlook is solid, the fourth

74
00:03:34,560 --> 00:03:36,520
Speaker 5: quarter was solid, and there seems to be real kind

75
00:03:36,560 --> 00:03:40,000
Speaker 5: of optimism around the progress with Blackwell. So yeah, I

76
00:03:40,040 --> 00:03:42,320
Speaker 5: think overall this kind of was in line with what

77
00:03:42,360 --> 00:03:45,320
Speaker 5: analysts were expecting, and the markets maybe been a little

78
00:03:45,320 --> 00:03:48,720
Speaker 5: bit down going into this court into this report, but yeah,

79
00:03:48,800 --> 00:03:50,520
Speaker 5: kind of a cyber relief I think for everybody.

80
00:03:50,640 --> 00:03:52,520
Speaker 2: The company did give an up eat forecast. It is

81
00:03:52,560 --> 00:03:55,080
Speaker 2: a sign that the AI build out is strong. Carolin Heid,

82
00:03:55,120 --> 00:03:56,720
Speaker 2: I want to bring you in, co host of Bloomberg

83
00:03:56,760 --> 00:04:01,920
Speaker 2: Technology on Bloomberg TV Weekday at eleven am Wall Street time.

84
00:04:03,160 --> 00:04:05,400
Speaker 2: I do want to ask about is there is the

85
00:04:05,440 --> 00:04:08,320
Speaker 2: sigh of relief sort of coming too early? Did we

86
00:04:08,400 --> 00:04:11,160
Speaker 2: get any answers about the impact of deep seek here.

87
00:04:11,280 --> 00:04:12,880
Speaker 4: Well, I think there's flies in the ointment.

88
00:04:13,320 --> 00:04:14,720
Speaker 1: I don't know how much you've talked about them, but I

89
00:04:14,720 --> 00:04:17,880
Speaker 1: think the gross margin is clearly a sign that we're

90
00:04:17,920 --> 00:04:20,440
Speaker 1: having perhaps a slight air pocket that many had talked

91
00:04:20,440 --> 00:04:23,680
Speaker 1: about is where you flip from Hopper to Blackwell. And indeed,

92
00:04:23,720 --> 00:04:26,359
Speaker 1: how much expense is spent on making sure that Blackwell

93
00:04:26,440 --> 00:04:27,960
Speaker 1: is easy to use and get up and running as

94
00:04:28,000 --> 00:04:28,760
Speaker 1: quickly as possible.

95
00:04:28,760 --> 00:04:30,919
Speaker 4: That could be an issue. And you have also got.

96
00:04:30,720 --> 00:04:33,280
Speaker 1: This potential guide of slightly weaker on the revenue for

97
00:04:33,320 --> 00:04:36,760
Speaker 1: the fiscal first quarter plus or minus forty three billion.

98
00:04:37,160 --> 00:04:38,920
Speaker 1: If they come into the low end, that does indeed

99
00:04:39,080 --> 00:04:42,160
Speaker 1: just slightly lose out. But I think we are trying

100
00:04:42,200 --> 00:04:46,800
Speaker 1: to hear ultimately whether that insane demand is now an

101
00:04:46,839 --> 00:04:49,800
Speaker 1: amazing amount of demand and does that basically put off

102
00:04:49,800 --> 00:04:52,479
Speaker 1: any concerns that we had about whether or not more

103
00:04:52,520 --> 00:04:54,760
Speaker 1: can be done with less in the future. And I

104
00:04:54,760 --> 00:04:57,440
Speaker 1: think we'll hear on the call, particularly from Jensen, as

105
00:04:57,480 --> 00:05:01,039
Speaker 1: to ultimately whether inferenced demand is really there for his

106
00:05:01,200 --> 00:05:02,400
Speaker 1: very expensive chick tyrol.

107
00:05:02,400 --> 00:05:04,159
Speaker 2: It is like reading a statement from the Federal Reserve

108
00:05:04,279 --> 00:05:05,479
Speaker 2: insane versus amazing.

109
00:05:06,520 --> 00:05:08,000
Speaker 3: This is what I was going to say ask you,

110
00:05:08,160 --> 00:05:10,320
Speaker 3: is that you know you follow this company so closely,

111
00:05:10,480 --> 00:05:12,800
Speaker 3: and it's how do you read the tea leaves because

112
00:05:12,800 --> 00:05:16,320
Speaker 3: he does tend to be very enthusiastic and maybe changes

113
00:05:16,360 --> 00:05:18,520
Speaker 3: the word and maybe there's a nuance to that change,

114
00:05:18,520 --> 00:05:21,320
Speaker 3: But how do you kind of read through what he says,

115
00:05:21,320 --> 00:05:22,760
Speaker 3: because we know he's going to be a beat.

116
00:05:23,000 --> 00:05:24,960
Speaker 1: I think he's going to be saying, look, trust us

117
00:05:25,000 --> 00:05:26,919
Speaker 1: with this R and D, trust us on a slightly

118
00:05:27,080 --> 00:05:30,240
Speaker 1: lower margin because we have got quote the fastest product

119
00:05:30,320 --> 00:05:31,279
Speaker 1: ramp in our countries.

120
00:05:31,320 --> 00:05:32,480
Speaker 4: His company's history.

121
00:05:32,600 --> 00:05:34,800
Speaker 1: This is a company that suddenly become like Apple and

122
00:05:34,839 --> 00:05:36,440
Speaker 1: giving us a new chip.

123
00:05:36,240 --> 00:05:38,599
Speaker 4: Every year that ever used to happen before.

124
00:05:38,760 --> 00:05:42,760
Speaker 1: There's an extraordinary pace which they're trying to innovate that

125
00:05:42,839 --> 00:05:45,599
Speaker 1: costs money. And indeed, of course you have to be

126
00:05:45,680 --> 00:05:48,239
Speaker 1: patient that sometimes there's going to be a slight friction

127
00:05:48,320 --> 00:05:51,200
Speaker 1: when you come to actually putting these Blackwell supercomputers into

128
00:05:51,240 --> 00:05:51,520
Speaker 1: the mix.

129
00:05:51,839 --> 00:05:54,640
Speaker 2: She shares, we're as high as four percent higher in

130
00:05:54,680 --> 00:05:58,640
Speaker 2: the after hours, now down about four tenths of one percent. Yeah,

131
00:05:58,720 --> 00:06:01,320
Speaker 2: not so much, like though investors are really trying to

132
00:06:01,320 --> 00:06:04,479
Speaker 2: make sense of these numbers. Yeah, it's not going and

133
00:06:04,520 --> 00:06:05,960
Speaker 2: it's not moving in a clear direction.

134
00:06:06,440 --> 00:06:09,120
Speaker 1: I think the clear direction is still always growth, isn't it?

135
00:06:09,200 --> 00:06:12,400
Speaker 1: And the idea that sure and one hundred percent increase

136
00:06:12,440 --> 00:06:14,640
Speaker 1: in revenue that got for their fiscal twenty twenty five,

137
00:06:14,680 --> 00:06:17,240
Speaker 1: which is the one that's just gone, and the one

138
00:06:17,600 --> 00:06:20,400
Speaker 1: prior to that, for two straight years we've had a

139
00:06:20,440 --> 00:06:21,560
Speaker 1: doubling of revenue.

140
00:06:21,680 --> 00:06:23,080
Speaker 4: We're going to have to start to get used.

141
00:06:23,000 --> 00:06:25,200
Speaker 1: To a fifty percent increase in revenue and then a

142
00:06:25,240 --> 00:06:28,440
Speaker 1: twenty three percent growth in revenue. The key question I

143
00:06:28,480 --> 00:06:31,000
Speaker 1: think there is still going to be what about the

144
00:06:31,080 --> 00:06:34,840
Speaker 1: cheap alternatives? What about Trainium too that Amazon has that

145
00:06:34,960 --> 00:06:37,240
Speaker 1: is much cheaper, What about some of these startups that.

146
00:06:37,240 --> 00:06:40,920
Speaker 2: You're seeing, does Tridium too, does that compare with Hopper?

147
00:06:40,960 --> 00:06:41,960
Speaker 2: Does it compare with black?

148
00:06:42,000 --> 00:06:44,599
Speaker 4: Well? About inference much more as well?

149
00:06:44,680 --> 00:06:46,600
Speaker 1: And so this is when they're trying to say, hey,

150
00:06:46,960 --> 00:06:49,599
Speaker 1: you've got your beautiful large language model, come and now

151
00:06:49,839 --> 00:06:51,800
Speaker 1: use it with us, and we will save you so

152
00:06:51,960 --> 00:06:53,440
Speaker 1: much more money about thirty percent.

153
00:06:53,640 --> 00:06:56,600
Speaker 2: So after you've used Nvidia to train it, yes, we

154
00:06:56,640 --> 00:06:59,120
Speaker 2: will process those requests for you exactly.

155
00:06:59,200 --> 00:07:00,600
Speaker 4: And I think that's where you're to start to get

156
00:07:00,600 --> 00:07:00,880
Speaker 4: the edge.

157
00:07:00,880 --> 00:07:02,839
Speaker 1: And in fact, there was an interesting interview done with

158
00:07:02,880 --> 00:07:06,080
Speaker 1: a rival with our CEO of course knows data centers

159
00:07:06,080 --> 00:07:09,559
Speaker 1: and mobiles deeply, and his own design work of course

160
00:07:09,640 --> 00:07:11,920
Speaker 1: helping a lot of those out there, including in Video,

161
00:07:11,960 --> 00:07:14,280
Speaker 1: but he himself saying, look, this is where the opening

162
00:07:14,520 --> 00:07:15,640
Speaker 1: is for competitors.

163
00:07:15,680 --> 00:07:18,400
Speaker 4: It's the inference stage. No one can really compete with the.

164
00:07:18,360 --> 00:07:20,400
Speaker 1: GPU power that in Video has when it comes to

165
00:07:20,400 --> 00:07:22,760
Speaker 1: training these models. But the next layer, when we actually

166
00:07:22,800 --> 00:07:25,480
Speaker 1: get it in and use these incredibly powerful things, that's

167
00:07:25,520 --> 00:07:26,400
Speaker 1: where the edge can come.

168
00:07:27,040 --> 00:07:30,240
Speaker 3: I mean, I don't know. In terms of market expectations,

169
00:07:30,720 --> 00:07:31,520
Speaker 3: not bad.

170
00:07:31,520 --> 00:07:34,440
Speaker 1: Yeah, signaling life kind of everyone was on the money,

171
00:07:34,920 --> 00:07:37,560
Speaker 1: alas did their job well. Jensen's done his job well

172
00:07:37,600 --> 00:07:39,760
Speaker 1: in terms of guiding. But had we got used to

173
00:07:40,240 --> 00:07:42,360
Speaker 1: beaten raises. But look look back to when we were

174
00:07:42,360 --> 00:07:44,440
Speaker 1: sit here together with Meta and Meta's numbers were a

175
00:07:44,560 --> 00:07:46,960
Speaker 1: really late be in miss the forecast, and then Zuck

176
00:07:47,000 --> 00:07:49,440
Speaker 1: gone on and just talked really optimistically and we all

177
00:07:49,440 --> 00:07:49,800
Speaker 1: bought it.

178
00:07:50,000 --> 00:07:52,400
Speaker 2: Yeah that's a good point. Yeah, wait to see what

179
00:07:52,400 --> 00:07:54,680
Speaker 2: happens in the cash session. Wait to hear from Jensen

180
00:07:54,720 --> 00:07:57,040
Speaker 2: a little later. We can only take the moves as

181
00:07:57,040 --> 00:07:58,480
Speaker 2: soon as the numbers come out with a grain of

182
00:07:58,480 --> 00:07:59,200
Speaker 2: solid red brown.

183
00:07:59,400 --> 00:08:02,440
Speaker 3: I put deep seek in the search and nothing, nothing's

184
00:08:02,440 --> 00:08:02,800
Speaker 3: coming up.

185
00:08:03,320 --> 00:08:05,240
Speaker 2: Let's just say it'll likely be asked about by at

186
00:08:05,280 --> 00:08:06,800
Speaker 2: least one analyst tonight.

187
00:08:07,160 --> 00:08:07,560
Speaker 3: Red round.

188
00:08:07,560 --> 00:08:08,680
Speaker 2: I want to bring you in here on sort of

189
00:08:08,720 --> 00:08:12,240
Speaker 2: the broader market reaction to this. There was this idea

190
00:08:12,360 --> 00:08:14,680
Speaker 2: that given the segiment we've seen in the last five

191
00:08:14,800 --> 00:08:17,600
Speaker 2: six days, uh in the broader market, the cell off

192
00:08:17,600 --> 00:08:20,360
Speaker 2: that we've seen in the megacap tech companies this year

193
00:08:20,400 --> 00:08:22,360
Speaker 2: and in the last few days, this would sort of

194
00:08:22,440 --> 00:08:23,840
Speaker 2: change the tone, does it.

195
00:08:24,200 --> 00:08:25,840
Speaker 5: I mean, it doesn't really seem like the tone has

196
00:08:25,880 --> 00:08:28,160
Speaker 5: changed at all. We're looking at the numbers is flat right,

197
00:08:28,800 --> 00:08:31,200
Speaker 5: kind of across the board, with everybody not really reacting.

198
00:08:32,320 --> 00:08:34,600
Speaker 5: You know, I think a lot of the narrative kind

199
00:08:34,640 --> 00:08:36,720
Speaker 5: of has been said. You know, the spending is still there.

200
00:08:36,800 --> 00:08:38,600
Speaker 5: I think if we read all the analysts reports heading

201
00:08:38,640 --> 00:08:41,200
Speaker 5: into this, that was the one data point that everyone

202
00:08:41,240 --> 00:08:43,040
Speaker 5: was hanging on to that there hasn't been anyone coming

203
00:08:43,040 --> 00:08:44,720
Speaker 5: out and saying, you know, we're going to be adjusting

204
00:08:44,720 --> 00:08:47,680
Speaker 5: our AI spending. They've stuck to you know, those massive

205
00:08:47,679 --> 00:08:51,000
Speaker 5: capex numbers that we saw throughout the quarter, so I think, yeah,

206
00:08:51,040 --> 00:08:53,160
Speaker 5: maybe a little bit of a wait and see still unfortunately,

207
00:08:53,920 --> 00:08:55,160
Speaker 5: you know, for the time being.

208
00:08:55,280 --> 00:08:57,199
Speaker 3: Well, and they did note I'm looking at our live

209
00:08:57,240 --> 00:08:59,760
Speaker 3: blog noted margins are being squeezed by the cost of

210
00:08:59,760 --> 00:09:02,480
Speaker 3: bringing new, more complicated products to market, and it sounds

211
00:09:02,520 --> 00:09:02,880
Speaker 3: like what.

212
00:09:03,120 --> 00:09:06,160
Speaker 2: Well, Carolyn explained that means because you know Kunjohn also

213
00:09:06,200 --> 00:09:08,160
Speaker 2: mentioned this idea of an air pocket or the transition

214
00:09:08,320 --> 00:09:12,040
Speaker 2: from Harper to Blackwell. What exactly happens behind the scenes

215
00:09:12,040 --> 00:09:13,240
Speaker 2: here and why is it so expensive?

216
00:09:13,400 --> 00:09:15,400
Speaker 1: Well, I think you've just ultimately got to make sure

217
00:09:15,440 --> 00:09:18,600
Speaker 1: that it works hardware seamlessly with the software that you've

218
00:09:18,640 --> 00:09:22,400
Speaker 1: got going, and ultimately that becomes more a powerful piece

219
00:09:22,400 --> 00:09:25,480
Speaker 1: of equipment that is able to use your data and

220
00:09:25,520 --> 00:09:27,160
Speaker 1: do what you want with it. I think when it

221
00:09:27,160 --> 00:09:30,400
Speaker 1: came to Hopper, there was that was working beautifully for many,

222
00:09:30,400 --> 00:09:32,360
Speaker 1: but everyone got deeply excited about the fact that there

223
00:09:32,400 --> 00:09:35,120
Speaker 1: was the latest greatest innovation suddenly coming out and sort

224
00:09:35,120 --> 00:09:38,400
Speaker 1: of decided to go long on Blackwell instead. We actually

225
00:09:38,440 --> 00:09:41,720
Speaker 1: saw a repositioning of that by many customers when they realized, oh,

226
00:09:41,720 --> 00:09:44,440
Speaker 1: actually this is going to take work for everyone to

227
00:09:44,559 --> 00:09:46,160
Speaker 1: ensure to pull this in, to get this into my

228
00:09:46,240 --> 00:09:47,960
Speaker 1: data centers, and to ultimately make sure that.

229
00:09:48,000 --> 00:09:50,280
Speaker 4: It's doing what it wants to do with the most efficiency.

230
00:09:50,480 --> 00:09:53,600
Speaker 1: Heating problems for example, just some of the technicalities that

231
00:09:53,640 --> 00:09:55,880
Speaker 1: they hadn't been able to stress test at this rapid

232
00:09:55,920 --> 00:09:56,600
Speaker 1: rate of innovation.

233
00:09:57,160 --> 00:10:00,000
Speaker 3: All right, just to rehash, Nvidier giving a bullish revenue

234
00:10:00,000 --> 00:10:02,960
Speaker 3: forecast for the current quarter, reassuring investors that spending on

235
00:10:03,000 --> 00:10:06,240
Speaker 3: AI computing remains strung. Just reading some of the reporting

236
00:10:06,280 --> 00:10:08,840
Speaker 3: out by our Oni and King, who follows this sector,

237
00:10:08,920 --> 00:10:11,200
Speaker 3: sales will be about forty three billion in the fiscal

238
00:10:11,240 --> 00:10:14,320
Speaker 3: first quarter, which runs through April. Analystaid An estimated about

239
00:10:14,360 --> 00:10:17,040
Speaker 3: forty two point three billion on average, with some projections

240
00:10:17,440 --> 00:10:19,680
Speaker 3: ranging as high as forty eight billion. So there were

241
00:10:19,760 --> 00:10:24,360
Speaker 3: some outliers and some very high expectations. Carolyn, how are

242
00:10:24,400 --> 00:10:26,600
Speaker 3: you thinking about this story? Obviously we want to see

243
00:10:26,920 --> 00:10:29,400
Speaker 3: what more commentary we get on the call the follow through.

244
00:10:29,440 --> 00:10:31,880
Speaker 3: In terms of the trade, like I'm looking at NASDAQ

245
00:10:31,880 --> 00:10:36,480
Speaker 3: one hundred mini futures, they're relatively little change, maybe up

246
00:10:36,480 --> 00:10:40,240
Speaker 3: about one tenth of one percent. S and P five

247
00:10:40,280 --> 00:10:43,240
Speaker 3: hundred mini futures, they are just a hair lower, pretty flat,

248
00:10:43,280 --> 00:10:46,719
Speaker 3: So it's really very calm. It feels like you're in

249
00:10:46,760 --> 00:10:47,520
Speaker 3: the aftermarket.

250
00:10:47,640 --> 00:10:49,480
Speaker 1: And maybe we just sort of braced ourselves because the

251
00:10:49,559 --> 00:10:52,839
Speaker 1: volatility they experienced yesterday, and then say we got used

252
00:10:52,880 --> 00:10:55,560
Speaker 1: to the fact that the market is trying to find

253
00:10:55,559 --> 00:10:58,400
Speaker 1: its direction when it comes to what is the question

254
00:10:58,480 --> 00:11:00,280
Speaker 1: of capacity, then we've got to go back to what

255
00:11:00,320 --> 00:11:01,679
Speaker 1: happened in the beginning of the week, and we're all

256
00:11:01,720 --> 00:11:05,119
Speaker 1: asking questions about TD Cowan's note over the weekend about Microsoft.

257
00:11:05,640 --> 00:11:08,280
Speaker 1: I'm sure that it will be questions put to Jensen

258
00:11:08,360 --> 00:11:12,360
Speaker 1: about ultimately is there a capacity over stretch or underreach

259
00:11:12,400 --> 00:11:14,360
Speaker 1: at the moment, and what about these so called scaling

260
00:11:14,440 --> 00:11:16,720
Speaker 1: rules that he's trying to combat in every feat, whether

261
00:11:16,800 --> 00:11:18,760
Speaker 1: or not it's about the scaling rules of how much

262
00:11:18,760 --> 00:11:21,199
Speaker 1: better llms can become with the amount of data on

263
00:11:21,280 --> 00:11:23,920
Speaker 1: compute you throw at it, but ultimately also the scaling

264
00:11:24,000 --> 00:11:26,600
Speaker 1: rules of China or whether that becomes something you can

265
00:11:26,679 --> 00:11:28,720
Speaker 1: still scale into in any way with an H twenty.

266
00:11:28,880 --> 00:11:30,320
Speaker 2: I'm glad he brought that up because one of the

267
00:11:30,360 --> 00:11:33,320
Speaker 2: questions that kun John mentioned was getting more detail on

268
00:11:33,400 --> 00:11:36,679
Speaker 2: views of more US restrictions, specifically with regard to China.

269
00:11:36,760 --> 00:11:40,240
Speaker 2: Read I know you've been doing some searching within the

270
00:11:40,360 --> 00:11:43,280
Speaker 2: press release. Are you seeing anything in there about China sales,

271
00:11:43,320 --> 00:11:45,000
Speaker 2: about a commentary about China.

272
00:11:44,800 --> 00:11:46,920
Speaker 5: No mention about China. So then Kunjohn is correct that

273
00:11:46,920 --> 00:11:48,560
Speaker 5: we that's definitely going to be a top of mind

274
00:11:49,200 --> 00:11:53,040
Speaker 5: topic for analysts on the call. You know right now

275
00:11:53,080 --> 00:11:54,960
Speaker 5: there's a lot of question marks looming around what they

276
00:11:55,000 --> 00:11:56,720
Speaker 5: can do there. It gets they get around twelve percent

277
00:11:56,760 --> 00:11:58,920
Speaker 5: of their sales over the past few years, most of

278
00:11:58,920 --> 00:12:01,439
Speaker 5: that coming across from game being automotive end data centers.

279
00:12:01,480 --> 00:12:03,600
Speaker 5: So I think there will be questions about, you know,

280
00:12:03,720 --> 00:12:06,760
Speaker 5: what sort of contingencies maybe they have in place if

281
00:12:07,080 --> 00:12:09,800
Speaker 5: any sort of rediction restrictions do come down.

282
00:12:10,000 --> 00:12:12,240
Speaker 3: Well, and Caroline, I mean jensenmong has made his way

283
00:12:12,280 --> 00:12:15,120
Speaker 3: to the way that's read. He said, conversations with President

284
00:12:15,160 --> 00:12:17,439
Speaker 3: Trump about these issues and his concerns.

285
00:12:17,679 --> 00:12:21,199
Speaker 1: Yeah, certainly, and that had been pre premeditated. He'd been

286
00:12:21,200 --> 00:12:23,800
Speaker 1: going to go and see President Trump. We know that

287
00:12:23,840 --> 00:12:25,960
Speaker 1: he'd wanted to. When our own ed Ludlow sat down

288
00:12:25,960 --> 00:12:28,400
Speaker 1: with him previously, he said, I'm ready willing and able

289
00:12:28,440 --> 00:12:30,400
Speaker 1: to go and meet President Trump, and certainly they wanted

290
00:12:30,440 --> 00:12:32,280
Speaker 1: me to go. She because remembers the end of the administration,

291
00:12:32,360 --> 00:12:35,560
Speaker 1: the Biden administration that's suddenly added all these curtailments to

292
00:12:35,559 --> 00:12:39,400
Speaker 1: how much they can place chips in other geographies, not

293
00:12:39,440 --> 00:12:41,360
Speaker 1: just China, because there's this worry that you sell into

294
00:12:41,360 --> 00:12:44,080
Speaker 1: a different nation to Singapore, and actually that's suddenly how

295
00:12:44,080 --> 00:12:47,199
Speaker 1: it's getting dispersed into China as well. So he's got

296
00:12:47,240 --> 00:12:50,319
Speaker 1: a lot to be debating. I do note in I mean,

297
00:12:50,320 --> 00:12:52,079
Speaker 1: I hate to be a cynic when you're looking at

298
00:12:52,120 --> 00:12:56,000
Speaker 1: these press releases, but he really only mentions US companies

299
00:12:56,040 --> 00:12:57,920
Speaker 1: at the beginning of how much that they're working with

300
00:12:58,000 --> 00:13:01,920
Speaker 1: the HWS is the Cisco's the verizons. I mean, it

301
00:13:02,000 --> 00:13:04,600
Speaker 1: really is a focus in on us.

302
00:13:04,760 --> 00:13:06,920
Speaker 3: It's just this week things I think now increasing a

303
00:13:07,000 --> 00:13:08,439
Speaker 3: partnership right with Cisco.

304
00:13:08,280 --> 00:13:10,679
Speaker 1: Exactly to get the networking better again, just get these

305
00:13:10,720 --> 00:13:12,240
Speaker 1: things into people's hands more efficiently.

306
00:13:12,400 --> 00:13:15,480
Speaker 2: We've talked about hop Hopper, We've talked a lot about Blackwell.

307
00:13:15,760 --> 00:13:17,360
Speaker 2: Is it too early for us to be talking about

308
00:13:17,360 --> 00:13:19,400
Speaker 2: Reuben I don't.

309
00:13:19,440 --> 00:13:23,640
Speaker 1: I mean, what in the next innovation, This is the

310
00:13:23,720 --> 00:13:26,240
Speaker 1: next year, this is the next I don't think Jadson

311
00:13:26,320 --> 00:13:27,960
Speaker 1: is ever too early to talk about what he's got

312
00:13:28,040 --> 00:13:28,720
Speaker 1: up next to sleep.

313
00:13:28,880 --> 00:13:31,200
Speaker 2: So that's that's the question. I mean, our investors looking

314
00:13:31,320 --> 00:13:33,240
Speaker 2: for you know, this is this is the This is

315
00:13:33,280 --> 00:13:36,040
Speaker 2: what happens with compute, right, You build something and then

316
00:13:36,040 --> 00:13:37,920
Speaker 2: you build something better, and you build something better. That

317
00:13:38,040 --> 00:13:40,280
Speaker 2: just happens over and over again, and it costs money.

318
00:13:40,400 --> 00:13:43,480
Speaker 2: It costs money. So how important is you know the

319
00:13:43,559 --> 00:13:47,240
Speaker 2: next generation what comes after blackwell apparently code named Rubin.

320
00:13:47,440 --> 00:13:50,000
Speaker 2: How important is that to where we are in the

321
00:13:50,040 --> 00:13:51,000
Speaker 2: timeline of Nvidia.

322
00:13:52,280 --> 00:13:56,040
Speaker 1: I think it's important to show that they are continuing

323
00:13:56,240 --> 00:13:59,719
Speaker 1: to just innovate ahead of any of the competition. That

324
00:14:00,120 --> 00:14:02,400
Speaker 1: the more it's more Lisa Sue who pays attention to

325
00:14:02,480 --> 00:14:06,120
Speaker 1: Ruben and things up like, yeah, how am I managing

326
00:14:06,200 --> 00:14:07,959
Speaker 1: to get any sort of in here in terms of

327
00:14:08,080 --> 00:14:10,559
Speaker 1: owning the data center space when it comes to just

328
00:14:10,679 --> 00:14:14,000
Speaker 1: the ability that Jensen has had to see around corners,

329
00:14:14,000 --> 00:14:15,679
Speaker 1: when it comes to the use and application of his

330
00:14:15,800 --> 00:14:17,360
Speaker 1: GPUs for gaming.

331
00:14:17,080 --> 00:14:19,200
Speaker 4: For crypto and then for generative AI.

332
00:14:19,760 --> 00:14:22,600
Speaker 1: But maybe Lisa Su starts to play the tack of okay,

333
00:14:22,680 --> 00:14:25,080
Speaker 1: but it's me with cheaper price points, and I wonder

334
00:14:25,080 --> 00:14:27,400
Speaker 1: whether the margin starts to creep into that narrative as

335
00:14:27,440 --> 00:14:30,480
Speaker 1: well as maybe yes, it's a deeply expensive piece of

336
00:14:30,560 --> 00:14:32,760
Speaker 1: kit with everyone offering cheaper alternatives.

337
00:14:32,840 --> 00:14:33,200
Speaker 6: What do you do?

338
00:14:33,280 --> 00:14:33,960
Speaker 4: How do you respond?

339
00:14:34,040 --> 00:14:36,000
Speaker 3: All right, folks, we're talking with Caroline Hyde, co host

340
00:14:36,000 --> 00:14:39,200
Speaker 3: of Bloomberg Technology on Bloomberg TV at eleven am Wall

341
00:14:39,280 --> 00:14:42,720
Speaker 3: Street Time. Let's bring into the conversation Mandy Seeing walking

342
00:14:42,800 --> 00:14:46,200
Speaker 3: into our studio Bloomberg Intelligence, Senior Tech Industry Analysis. He

343
00:14:46,240 --> 00:14:48,000
Speaker 3: walked in because we asked him to and we were

344
00:14:48,760 --> 00:14:51,400
Speaker 3: hoping you would. Hey, and video shares have been bouncing around.

345
00:14:51,920 --> 00:14:54,960
Speaker 3: Should any of us who are thinking about the AI industry?

346
00:14:55,080 --> 00:14:56,400
Speaker 3: Is this a dose of optimism?

347
00:14:57,440 --> 00:15:00,400
Speaker 7: It is? But when it comes to Nvidia specific I

348
00:15:00,480 --> 00:15:04,160
Speaker 7: would say, you know, the fact that networking revenue declines

349
00:15:04,240 --> 00:15:07,720
Speaker 7: sequentially for the second quarter in a row again in

350
00:15:07,840 --> 00:15:11,240
Speaker 7: Vidio is very high bar. And you know, given their

351
00:15:11,280 --> 00:15:14,800
Speaker 7: bed and raises, everyone expects them to do, you know,

352
00:15:14,960 --> 00:15:18,000
Speaker 7: well on every line. So networking going down, to me,

353
00:15:18,120 --> 00:15:21,680
Speaker 7: that's a sticking point, along with gross margins going down

354
00:15:22,200 --> 00:15:25,240
Speaker 7: and just the sequential increases that we have seen in

355
00:15:25,320 --> 00:15:29,280
Speaker 7: the past quarters. It looks like it's tapering again. It's

356
00:15:29,480 --> 00:15:32,560
Speaker 7: double digit. But when it comes to the overall growth rate,

357
00:15:32,680 --> 00:15:35,440
Speaker 7: it feels like it's decelerating. And the last data point

358
00:15:35,480 --> 00:15:40,520
Speaker 7: I would offer is the cloud exposure, the hyperscale cloud exposure.

359
00:15:40,800 --> 00:15:44,600
Speaker 7: It was forty five percent before now it's fifty percent Now.

360
00:15:44,760 --> 00:15:47,880
Speaker 7: What it tells you is they kind of we are exposed.

361
00:15:48,480 --> 00:15:51,520
Speaker 7: And so the fact that you know networking is going down,

362
00:15:51,640 --> 00:15:56,960
Speaker 7: it's a sign that you know, training versus inferencing component

363
00:15:57,560 --> 00:16:01,200
Speaker 7: that seems to be really in question here because networking

364
00:16:01,280 --> 00:16:06,040
Speaker 7: does well. If training is growing, if networking is going down,

365
00:16:06,240 --> 00:16:09,560
Speaker 7: that's a sign that they may be facing some sort

366
00:16:09,640 --> 00:16:13,960
Speaker 7: of challenges or starting to get more questions around the

367
00:16:14,000 --> 00:16:16,200
Speaker 7: scale of training. And I think that's a big question,

368
00:16:16,320 --> 00:16:16,840
Speaker 7: So go.

369
00:16:16,920 --> 00:16:19,240
Speaker 2: Deeper on that with training versus inference. We're talking a

370
00:16:19,240 --> 00:16:22,920
Speaker 2: little bit about it with Caroline context Amazon and what

371
00:16:23,000 --> 00:16:26,880
Speaker 2: Amazon has when it comes to in France. But what

372
00:16:27,000 --> 00:16:28,960
Speaker 2: do we learn in video within Vidia today.

373
00:16:29,440 --> 00:16:33,680
Speaker 7: Look, if you're setting up a giant cluster like Xai

374
00:16:33,880 --> 00:16:36,400
Speaker 7: has with you know, one hundred and two hundred thousand GPUs,

375
00:16:36,680 --> 00:16:38,760
Speaker 7: you need a lot of networking there. But if you

376
00:16:38,800 --> 00:16:41,240
Speaker 7: are setting up a data center with let's say a thousand,

377
00:16:41,320 --> 00:16:45,120
Speaker 7: two thousand GPUs, you don't need as much of Nvidia

378
00:16:45,160 --> 00:16:49,040
Speaker 7: as networking, and that's where you know, training versus inferencing

379
00:16:49,120 --> 00:16:53,600
Speaker 7: really matters, because if you know, companies are leaning towards

380
00:16:53,640 --> 00:16:58,840
Speaker 7: setting up more data centers with fewer chips. I mean

381
00:16:58,880 --> 00:17:00,800
Speaker 7: when I say, if you are still thousands of chips

382
00:17:00,840 --> 00:17:03,120
Speaker 7: but not to the scale of one hundred thousand, then

383
00:17:03,280 --> 00:17:05,720
Speaker 7: networking won't do as well. And that's the whole debate

384
00:17:05,760 --> 00:17:08,600
Speaker 7: about training versus inferencing. So to my mind, you know,

385
00:17:08,800 --> 00:17:13,000
Speaker 7: deep seek and all the recent developments from hyperscalers suggests

386
00:17:13,040 --> 00:17:17,040
Speaker 7: that you may not need giant clusters and that is

387
00:17:17,520 --> 00:17:21,240
Speaker 7: I think going to affect Nvidia going forward. So when

388
00:17:21,280 --> 00:17:22,920
Speaker 7: I look at their guide, when I look at the

389
00:17:23,040 --> 00:17:26,760
Speaker 7: fact that you know, networking declined sequentially, again, it's a

390
00:17:26,880 --> 00:17:30,760
Speaker 7: great print, but to me, the path seems to be

391
00:17:31,000 --> 00:17:33,560
Speaker 7: that it will slow down on the networking side, and

392
00:17:33,600 --> 00:17:36,080
Speaker 7: then you have to ask yourself where will the upside

393
00:17:36,160 --> 00:17:38,919
Speaker 7: come from. Maybe on the call they expand on starlink

394
00:17:39,119 --> 00:17:42,359
Speaker 7: and you know, the sovereigns that is a big pocket.

395
00:17:42,440 --> 00:17:45,480
Speaker 7: If that pans out in terms of training, like setting

396
00:17:45,600 --> 00:17:49,280
Speaker 7: up big clusters that Sam Altman has talked about, you know,

397
00:17:49,400 --> 00:17:52,160
Speaker 7: with projects started, then you will start to see networking

398
00:17:52,200 --> 00:17:55,000
Speaker 7: bounce but for me, there is a clear connection between

399
00:17:55,040 --> 00:17:58,000
Speaker 7: networking and size of cluster and also software. There was

400
00:17:58,119 --> 00:18:01,200
Speaker 7: nothing in the print around software, even though it's you know,

401
00:18:01,320 --> 00:18:04,080
Speaker 7: two billion dollars in revenue when you look at a

402
00:18:04,200 --> 00:18:07,040
Speaker 7: data center, you know, aligne, which with over one hundred

403
00:18:07,040 --> 00:18:10,320
Speaker 7: and fifteen billion dollars two billion is very great small,

404
00:18:10,440 --> 00:18:13,800
Speaker 7: but they didn't mention anything around software. So there's still

405
00:18:13,840 --> 00:18:17,639
Speaker 7: a chip company that depends on you know, companies buying

406
00:18:17,720 --> 00:18:22,040
Speaker 7: their their chips in bulk and also leering and networking.

407
00:18:22,160 --> 00:18:25,120
Speaker 7: So to me that the networking part is critical here.

408
00:18:25,240 --> 00:18:27,760
Speaker 3: You know, one of our on our live blog are

409
00:18:27,880 --> 00:18:31,360
Speaker 3: Carmin Rhinikey saying, you know, bringing kind of salesforce into

410
00:18:31,359 --> 00:18:33,600
Speaker 3: the mix, noting that shares of Salesforce are down more

411
00:18:33,640 --> 00:18:36,560
Speaker 3: than five percent in post market trading after it reported

412
00:18:36,720 --> 00:18:40,200
Speaker 3: we growth outlooks, stoking worry about its new AI product.

413
00:18:40,240 --> 00:18:41,920
Speaker 3: Like you know, you do think about kind of the

414
00:18:42,000 --> 00:18:45,520
Speaker 3: whole infrastructure, the whole community. Caroline, come on in the

415
00:18:45,560 --> 00:18:47,960
Speaker 3: conversation following off mandypin if you've got a question.

416
00:18:48,000 --> 00:18:51,040
Speaker 1: Well, I do think that that's interesting about agent KI

417
00:18:51,160 --> 00:18:53,960
Speaker 1: and this is where salesforce and agent force is in

418
00:18:54,040 --> 00:18:56,880
Speaker 1: the front and center and look, Jensen pays lip service

419
00:18:56,960 --> 00:18:59,040
Speaker 1: to it in the statement saying AI is advancing at

420
00:18:59,119 --> 00:19:02,040
Speaker 1: light speed as a GENTKI and physical AI set the

421
00:19:02,080 --> 00:19:04,080
Speaker 1: stage for the next wave, but he's got to keep

422
00:19:04,160 --> 00:19:07,680
Speaker 1: talking about scaling laws when it comes to training. That

423
00:19:07,880 --> 00:19:09,639
Speaker 1: is going back to a reasoning AI. He's trying to

424
00:19:09,640 --> 00:19:12,119
Speaker 1: find new ways to galvanize us all to be excited.

425
00:19:12,160 --> 00:19:14,399
Speaker 1: But if the agents aren't selling that well as far

426
00:19:14,480 --> 00:19:15,520
Speaker 1: as we thought about.

427
00:19:15,520 --> 00:19:18,520
Speaker 7: Is reasoning really that good for Nvidia? In my mind,

428
00:19:18,840 --> 00:19:23,200
Speaker 7: if the pivot has happened towards more reasoning, huh, that's

429
00:19:23,359 --> 00:19:26,320
Speaker 7: not a positive for Nvidia because when you're doing reasoning,

430
00:19:26,400 --> 00:19:30,359
Speaker 7: you are doing more inference compute. And I mean, based

431
00:19:30,400 --> 00:19:33,840
Speaker 7: on Jensen's past comments, Nvidia has a forty percent share

432
00:19:33,920 --> 00:19:38,200
Speaker 7: in inferencing. Sixty percent is training, right, So how can

433
00:19:38,280 --> 00:19:40,960
Speaker 7: it be better if you're doing more training, you know,

434
00:19:41,160 --> 00:19:44,800
Speaker 7: using that reasoning approach. So those are like very small nuggets,

435
00:19:45,160 --> 00:19:47,640
Speaker 7: but I think in the grand scheme of things they matter.

436
00:19:47,960 --> 00:19:51,200
Speaker 2: You mentioned agentic AGENTIC AI. You got to explain what

437
00:19:51,359 --> 00:19:53,560
Speaker 2: that is for people, especially in the context of what

438
00:19:53,720 --> 00:19:56,480
Speaker 2: you talked about with Panos Pane earlier and what Amazon

439
00:19:56,560 --> 00:19:57,399
Speaker 2: released today, and.

440
00:19:57,440 --> 00:20:02,360
Speaker 1: I think this is how ultimately starts being more helpful

441
00:20:02,520 --> 00:20:06,760
Speaker 1: that a not only will your chatbot basically go away

442
00:20:06,800 --> 00:20:10,040
Speaker 1: and take longer thoughts about things, thoughts quote unquote, you know,

443
00:20:10,320 --> 00:20:12,920
Speaker 1: trying to or personify these these pits of tar and

444
00:20:13,000 --> 00:20:15,680
Speaker 1: it is real, it is real, and person fall in

445
00:20:15,760 --> 00:20:20,320
Speaker 1: love with it. But she if it's ass then is

446
00:20:20,400 --> 00:20:22,800
Speaker 1: going to go off and start actioning stuff for you.

447
00:20:23,080 --> 00:20:24,840
Speaker 1: Not only are they going to tell you the best restaurant,

448
00:20:24,880 --> 00:20:26,840
Speaker 1: they go book it for you. They're going to potentially

449
00:20:26,880 --> 00:20:31,200
Speaker 1: start booking flights, paying for them. Agents can take ultimately

450
00:20:31,480 --> 00:20:34,880
Speaker 1: still relatively basic needs, but run with them and complete

451
00:20:34,920 --> 00:20:38,119
Speaker 1: the task. And that's ultimately where Salesforce and these enterprise

452
00:20:38,200 --> 00:20:40,760
Speaker 1: software companies have started to really see. They're in the

453
00:20:40,960 --> 00:20:43,680
Speaker 1: question for many which has is always being asked is

454
00:20:43,720 --> 00:20:46,200
Speaker 1: how they price for it? And Salesforce sort of stick

455
00:20:46,240 --> 00:20:48,440
Speaker 1: the cell and sells on a limb and started saying, well,

456
00:20:48,520 --> 00:20:50,440
Speaker 1: this is how much going to pay per per user

457
00:20:50,520 --> 00:20:52,400
Speaker 1: or indeed per amount that you actually use.

458
00:20:52,480 --> 00:20:54,440
Speaker 4: But all of this is the return on AI that

459
00:20:54,520 --> 00:20:56,040
Speaker 4: we keep questioning in so many ways.

460
00:20:56,280 --> 00:20:59,359
Speaker 7: I mean, you have to ask yourself, why is Nvidia's

461
00:20:59,400 --> 00:21:00,760
Speaker 7: growth smart going down?

462
00:21:01,000 --> 00:21:01,119
Speaker 6: Right?

463
00:21:01,800 --> 00:21:04,600
Speaker 7: You know, obviously they release a new product that should

464
00:21:04,640 --> 00:21:07,960
Speaker 7: command premium pricing, and look they guided to that. So

465
00:21:08,160 --> 00:21:11,200
Speaker 7: they're saying in the second half you will see gross

466
00:21:11,240 --> 00:21:14,880
Speaker 7: margins rebound, but they're going down for a reason. Either

467
00:21:15,600 --> 00:21:19,440
Speaker 7: they're paying up more for foundry costs to DSMC or

468
00:21:19,680 --> 00:21:23,840
Speaker 7: they're just incanctivizing their customers to buy more. And you know,

469
00:21:24,359 --> 00:21:27,040
Speaker 7: given black Belt did so well in the quarter, there

470
00:21:27,119 --> 00:21:30,119
Speaker 7: could be some you know pricing aspect to it in

471
00:21:30,240 --> 00:21:33,200
Speaker 7: terms of just driving demand and you know, shipping morels.

472
00:21:33,280 --> 00:21:35,440
Speaker 4: Is that because we're getting more competition as well in

473
00:21:35,480 --> 00:21:35,959
Speaker 4: the future.

474
00:21:36,400 --> 00:21:40,399
Speaker 7: I mean CSPs, the four CSPs make fifty percent of

475
00:21:40,480 --> 00:21:45,520
Speaker 7: their revenue hyperscalers, the hyperscalers.

476
00:21:44,920 --> 00:21:48,920
Speaker 3: So all you're talking about your alphabet, you're talking about

477
00:21:48,920 --> 00:21:51,680
Speaker 3: your Microsoft microsycroso.

478
00:21:51,960 --> 00:21:55,040
Speaker 7: Yeah, and Amazon so and it was forty five percent

479
00:21:55,600 --> 00:21:58,920
Speaker 7: last quarter, So that exporture grew, right, I mean, and

480
00:21:59,200 --> 00:22:01,879
Speaker 7: they're shipping more blackwell. So when you connect it to

481
00:22:02,359 --> 00:22:06,399
Speaker 7: clearly there is and these are customers that are buying

482
00:22:06,480 --> 00:22:10,040
Speaker 7: chips in bulk, so they need some pricing, you know.

483
00:22:10,119 --> 00:22:13,000
Speaker 3: To Does that play to the margin that they're getting

484
00:22:13,000 --> 00:22:13,320
Speaker 3: a break?

485
00:22:14,680 --> 00:22:17,320
Speaker 7: I would assume? So, right, Yeah, if you are a

486
00:22:17,440 --> 00:22:20,600
Speaker 7: customer that is paying, you know, twenty billion dollars a

487
00:22:20,760 --> 00:22:23,400
Speaker 7: year or two in Vidia, you want some pricing break

488
00:22:23,440 --> 00:22:24,840
Speaker 7: compared to the other customers, not.

489
00:22:24,920 --> 00:22:25,800
Speaker 3: Just a free basket.

490
00:22:26,200 --> 00:22:28,320
Speaker 1: It was a question that I couldn't answer fully because

491
00:22:28,359 --> 00:22:29,639
Speaker 1: I don't have the depth of knowledge that you do

492
00:22:29,800 --> 00:22:33,360
Speaker 1: Man deep And ultimately, the issue of getting to Blackwell

493
00:22:33,440 --> 00:22:35,800
Speaker 1: and the heating issues that we've seen and the ultimate rollout,

494
00:22:36,160 --> 00:22:38,360
Speaker 1: has that cost them any margin? And does it cost

495
00:22:38,400 --> 00:22:40,280
Speaker 1: them to have top going back and fixing things or

496
00:22:40,400 --> 00:22:43,200
Speaker 1: is it on someone else to keep using ensuring it's working.

497
00:22:43,400 --> 00:22:46,280
Speaker 7: So there were concerns around Blackwell, and look, if you're

498
00:22:46,280 --> 00:22:49,520
Speaker 7: a customer, you would want to wait until they resolve

499
00:22:49,600 --> 00:22:52,879
Speaker 7: all these issues. But in Media moved at a pretty

500
00:22:52,920 --> 00:22:55,800
Speaker 7: fast pace in terms of fixing those issues and actually

501
00:22:56,000 --> 00:22:58,760
Speaker 7: getting eleven billion dollars of revenue in the quarter. So

502
00:22:58,920 --> 00:23:01,199
Speaker 7: the fact that they were to do it so quickly

503
00:23:01,880 --> 00:23:04,920
Speaker 7: shows good execution. But they're shipping a lot of these

504
00:23:05,000 --> 00:23:08,879
Speaker 7: chips to the hyperscalers, who again I feel, have to

505
00:23:08,960 --> 00:23:13,320
Speaker 7: be incentivized to take those bulk shipments, and that could

506
00:23:13,359 --> 00:23:15,479
Speaker 7: have played in terms of the lower margins.

507
00:23:15,520 --> 00:23:18,960
Speaker 2: And we're the concerns about overbuild answered where they put

508
00:23:19,000 --> 00:23:19,320
Speaker 2: to rest.

509
00:23:19,520 --> 00:23:23,560
Speaker 7: I mean they look at all of their suppliers TSMC,

510
00:23:24,000 --> 00:23:26,760
Speaker 7: Micron on the HBM side, all of them are still

511
00:23:26,800 --> 00:23:29,680
Speaker 7: supply constraints, so I can't imagine, you know, there is

512
00:23:29,760 --> 00:23:33,240
Speaker 7: any chance of an overbuild here. There's clearly more demand

513
00:23:33,359 --> 00:23:36,800
Speaker 7: than supply and all the suppliers are continuing to expand capacity.

514
00:23:37,080 --> 00:23:39,480
Speaker 7: So that is the least of my concern here. It's

515
00:23:39,560 --> 00:23:43,000
Speaker 7: more the competition and the lower gross margins and the

516
00:23:43,119 --> 00:23:46,240
Speaker 7: training versus inferencing is the key point that I hope

517
00:23:46,440 --> 00:23:48,760
Speaker 7: he addresses on the call in relation to deep sea.

518
00:23:48,920 --> 00:23:50,879
Speaker 3: I'm just again going to people who maybe are not

519
00:23:51,080 --> 00:23:54,359
Speaker 3: so in it, like you guys are the training versus inferencing,

520
00:23:54,680 --> 00:23:56,760
Speaker 3: Like what do they need to kind of understand. I

521
00:23:56,840 --> 00:23:58,240
Speaker 3: want to go back there because I do feel like

522
00:23:58,320 --> 00:23:59,480
Speaker 3: that's the big factor here.

523
00:24:00,080 --> 00:24:04,119
Speaker 7: Building one million GPU clusters right now, the largest sized

524
00:24:04,119 --> 00:24:08,000
Speaker 7: cluster is a two hundred K cluster that XCAI has built,

525
00:24:08,080 --> 00:24:09,919
Speaker 7: and I think META is also a building, and.

526
00:24:09,960 --> 00:24:12,119
Speaker 2: That's for training or for enfortan training training.

527
00:24:12,160 --> 00:24:15,560
Speaker 7: You build a one million or one hundred thousand chip cluster,

528
00:24:16,000 --> 00:24:18,560
Speaker 7: you're using it for training, You're not doing inferencing because

529
00:24:18,640 --> 00:24:22,560
Speaker 7: inferencing patterns are very volatile, and inferencing is more tied

530
00:24:22,600 --> 00:24:26,719
Speaker 7: to usage. You know, if I am using a chat

531
00:24:26,800 --> 00:24:29,959
Speaker 7: gipt during the day, the traffic will be more, at

532
00:24:30,080 --> 00:24:31,680
Speaker 7: night it will be less. So I'm not going to

533
00:24:31,960 --> 00:24:35,560
Speaker 7: keep the chips under utilized. I want one hundred thousand

534
00:24:35,680 --> 00:24:39,360
Speaker 7: chips to be utilized at maximum capacity. That's why training

535
00:24:39,760 --> 00:24:42,240
Speaker 7: is what is used for, you know, the large GPU clusters.

536
00:24:42,359 --> 00:24:45,000
Speaker 2: But to that point, even to the inference point we learn,

537
00:24:45,280 --> 00:24:47,320
Speaker 2: I think today that deep seek is sort of doing

538
00:24:47,440 --> 00:24:50,399
Speaker 2: surge pricing when it comes to when you're using the

539
00:24:50,560 --> 00:24:53,040
Speaker 2: app because they want people to use it not during

540
00:24:53,119 --> 00:24:54,760
Speaker 2: peak times because that is expensive.

541
00:24:54,960 --> 00:24:57,040
Speaker 7: Yeah, I mean look, and that goes to the under

542
00:24:57,119 --> 00:24:59,760
Speaker 7: supply aspect, right, so the market is still under some

543
00:25:00,359 --> 00:25:03,119
Speaker 7: and that's where wherever you can find capacity, you use it.

544
00:25:03,359 --> 00:25:05,119
Speaker 3: We know, you guys both have to go. Caroline, So

545
00:25:05,320 --> 00:25:07,320
Speaker 3: you're sitting down with Jensen, what do you want to know?

546
00:25:08,200 --> 00:25:10,399
Speaker 1: I want to know about For me, I want to

547
00:25:10,440 --> 00:25:13,280
Speaker 1: know about China, and I also want to know about

548
00:25:13,480 --> 00:25:17,200
Speaker 1: this capacity issue that we keep talking about and how

549
00:25:18,000 --> 00:25:21,520
Speaker 1: ultimately at Easy is with the amount that demand he

550
00:25:21,600 --> 00:25:22,840
Speaker 1: still go going all.

551
00:25:22,760 --> 00:25:24,440
Speaker 3: Right, same thing for you, man, Deep, you're sitting down

552
00:25:24,440 --> 00:25:24,800
Speaker 3: with Jensen.

553
00:25:24,880 --> 00:25:28,919
Speaker 7: Hey, right, training, Just tell me your training share right now?

554
00:25:29,040 --> 00:25:31,360
Speaker 7: Is there still sixty forty? Has it changed? And what's

555
00:25:31,440 --> 00:25:31,960
Speaker 7: the outlook?

556
00:25:32,200 --> 00:25:32,520
Speaker 3: All right?

557
00:25:32,600 --> 00:25:33,200
Speaker 4: Guys, you rock.

558
00:25:33,600 --> 00:25:36,000
Speaker 3: Thank you so much. Man Deep seeing Bloomberg Intelligence senior

559
00:25:36,040 --> 00:25:39,200
Speaker 3: tech industry analyst here in studio along with Caroline Hide

560
00:25:39,240 --> 00:25:42,359
Speaker 3: Coast of Bloomberg Technology and Bloomberg TV. Check out Man

561
00:25:42,400 --> 00:25:44,880
Speaker 3: Deep's research. It'll be on the Bloomberg tomorrow. Caroline will

562
00:25:44,920 --> 00:25:47,199
Speaker 3: be all over it on Bloomberg Technology. That comes your

563
00:25:47,200 --> 00:25:50,000
Speaker 3: way at eleven am Wall Street time on Bloomberg TV.

564
00:25:50,440 --> 00:25:53,200
Speaker 2: All right, well, you heard Caroline mention China just now,

565
00:25:53,280 --> 00:25:54,840
Speaker 2: and that's kind of what we want to talk about

566
00:25:54,880 --> 00:25:57,960
Speaker 2: with Michael Shepherd. He's Bloomberg News senior editor for Technology

567
00:25:58,000 --> 00:26:00,320
Speaker 2: and Strategic Industries. He joins us from our wash Rington,

568
00:26:00,600 --> 00:26:04,200
Speaker 2: DC bureau. As Red Brown pointed out to us, Michael

569
00:26:04,359 --> 00:26:06,560
Speaker 2: just a few moments ago, no mention of China in

570
00:26:06,680 --> 00:26:09,920
Speaker 2: the press release. We are waiting for Jensen to take

571
00:26:10,040 --> 00:26:15,400
Speaker 2: questions from analysts about that. What are the concerns around potential.

572
00:26:15,560 --> 00:26:18,800
Speaker 2: I don't want to use the term sanctions, but restrictions

573
00:26:19,119 --> 00:26:20,760
Speaker 2: on what in Nvidia can sell to China.

574
00:26:21,560 --> 00:26:24,000
Speaker 6: Well, that's a better way of putting it, Tim, because

575
00:26:24,359 --> 00:26:27,119
Speaker 6: really what in Vidia has confronted it's not some it

576
00:26:27,240 --> 00:26:32,120
Speaker 6: sanctions or penalties, but limits on the types and quantities

577
00:26:32,160 --> 00:26:34,760
Speaker 6: of chips that can sell to Chinese entities. And of

578
00:26:34,840 --> 00:26:39,600
Speaker 6: course this is in response to the challenge that the

579
00:26:39,680 --> 00:26:44,440
Speaker 6: world's second largest economy poses to the US in advanced technology,

580
00:26:44,760 --> 00:26:49,000
Speaker 6: but especially in this emerging area of artificial intelligence. And

581
00:26:49,560 --> 00:26:53,200
Speaker 6: Donald Trump upon taking office, he made it a national

582
00:26:53,280 --> 00:26:57,720
Speaker 6: priority to preserve dominance and artificial intelligence. But that was

583
00:26:57,880 --> 00:27:01,159
Speaker 6: also a theme during the Biden administration, and some of

584
00:27:01,200 --> 00:27:05,520
Speaker 6: these restrictions that have really been deviled in Vidia at

585
00:27:05,600 --> 00:27:09,119
Speaker 6: times date to the Biden era and President Joe Biden,

586
00:27:09,320 --> 00:27:11,840
Speaker 6: a week before leaving office and handing over the reigns

587
00:27:11,880 --> 00:27:15,879
Speaker 6: of power to Donald Trump, imposed a three tiered system

588
00:27:16,000 --> 00:27:21,600
Speaker 6: of limits on country's access to sensitive chips for AI

589
00:27:21,840 --> 00:27:25,639
Speaker 6: of the likes of which that in Nvidia produces. And

590
00:27:25,840 --> 00:27:29,120
Speaker 6: so we are waiting to see what Jensen Wang will

591
00:27:29,200 --> 00:27:32,400
Speaker 6: have to say about this, what sort of challenge as

592
00:27:32,440 --> 00:27:34,959
Speaker 6: opposed to the business, and what sorts of steps has

593
00:27:35,000 --> 00:27:37,120
Speaker 6: he taken then maybe try to talk Donald Trump out

594
00:27:37,160 --> 00:27:37,280
Speaker 6: of it.

595
00:27:39,640 --> 00:27:41,600
Speaker 3: Has he been so has he been to the White

596
00:27:41,640 --> 00:27:42,320
Speaker 3: House yet.

597
00:27:43,240 --> 00:27:45,840
Speaker 6: And he was one of the last of the big

598
00:27:46,040 --> 00:27:49,280
Speaker 6: tech executives to pay a visit to the White House

599
00:27:49,320 --> 00:27:52,399
Speaker 6: and see Donald Trump. There was no visit by Jensen

600
00:27:52,520 --> 00:27:56,280
Speaker 6: Wang to mar Lago during the transition, and it wasn't

601
00:27:56,359 --> 00:28:00,439
Speaker 6: until January thirty first, that week when deep seek really

602
00:28:00,640 --> 00:28:04,119
Speaker 6: shook the markets, and especially in Vidia. It was a

603
00:28:04,280 --> 00:28:07,000
Speaker 6: Friday that he went in to go meet with the

604
00:28:07,119 --> 00:28:12,119
Speaker 6: President and talk generally about artificial intelligence, of course, but

605
00:28:12,240 --> 00:28:15,560
Speaker 6: you had to imagine that export controls came up. Nvidia

606
00:28:15,680 --> 00:28:19,520
Speaker 6: certainly has railed against them. We saw statements coming from

607
00:28:19,840 --> 00:28:23,280
Speaker 6: their spokespeople, as you know, we heard the Biden folks

608
00:28:23,680 --> 00:28:27,080
Speaker 6: talk more about them. And then as the Trump team

609
00:28:27,240 --> 00:28:30,600
Speaker 6: came in and saw a deep Seek's breakthrough and thought, hey,

610
00:28:30,800 --> 00:28:33,399
Speaker 6: we have to do something ourselves. We may need to

611
00:28:33,640 --> 00:28:36,720
Speaker 6: further tighten these restrictions. Well they weighed in again. But

612
00:28:36,840 --> 00:28:39,840
Speaker 6: Jensen Wong has been quiet about it. So this will

613
00:28:39,840 --> 00:28:41,320
Speaker 6: be a moment for him to say something.

614
00:28:41,760 --> 00:28:43,560
Speaker 2: You know, I think it was Caroline, I can't remember

615
00:28:43,600 --> 00:28:45,560
Speaker 2: exactly who it was on our on our Eric Carrol

616
00:28:45,600 --> 00:28:48,280
Speaker 2: who mentioned that in the press release, the names that

617
00:28:48,360 --> 00:28:52,640
Speaker 2: are mentioned cloud service providers aws Core Weave, Google Cloud Platform,

618
00:28:52,960 --> 00:29:01,719
Speaker 2: Microsoft Aser, Oral Cloud, Oracle Cloud Infrastructure, AWS, Cisco companies,

619
00:29:01,760 --> 00:29:04,959
Speaker 2: all US companies, And I don't know if it's if

620
00:29:05,040 --> 00:29:07,600
Speaker 2: it's a political statement or if it's just a statement

621
00:29:07,680 --> 00:29:11,120
Speaker 2: of the fact that these most prominent companies are based

622
00:29:11,160 --> 00:29:14,640
Speaker 2: in the US. Is there a read into this, Michael,

623
00:29:14,680 --> 00:29:17,280
Speaker 2: that this is a company that's making sure it's touting

624
00:29:17,520 --> 00:29:18,720
Speaker 2: its American connections.

625
00:29:19,080 --> 00:29:21,760
Speaker 6: So many companies, Tim McCarroll, are trying to get on

626
00:29:21,800 --> 00:29:23,800
Speaker 6: the right side of Trump in that sense, and in

627
00:29:23,920 --> 00:29:27,960
Speaker 6: this case, the facts may be conveniently aligned to whatever

628
00:29:28,160 --> 00:29:32,720
Speaker 6: Trump and the new administration may want to hear about

629
00:29:32,960 --> 00:29:37,960
Speaker 6: American companies buying from an American provider, namely in Vidio.

630
00:29:38,040 --> 00:29:40,800
Speaker 6: But of course, so many of Nvidia's chips are made

631
00:29:40,840 --> 00:29:44,640
Speaker 6: in Taiwan by TSMC, so that has to be another

632
00:29:44,840 --> 00:29:48,440
Speaker 6: area of concern for Jensen Wong as he hears Donald

633
00:29:48,440 --> 00:29:52,920
Speaker 6: Trump talk about potential tariffs on semiconductor imports. He hasn't

634
00:29:52,960 --> 00:29:55,680
Speaker 6: imposed him aet he hasn't really announced him yet, but

635
00:29:55,800 --> 00:29:58,160
Speaker 6: it is something that we are watching for in the

636
00:29:58,280 --> 00:30:01,360
Speaker 6: coming weeks, and of course, as it will be delivering

637
00:30:01,760 --> 00:30:06,040
Speaker 6: this address to Congress next Tuesday, evening, and that could

638
00:30:06,040 --> 00:30:08,640
Speaker 6: be a moment where we hear him talk in somewhat

639
00:30:08,720 --> 00:30:13,080
Speaker 6: more detail about this. Remember, the semiconductor supply chain is

640
00:30:13,240 --> 00:30:19,760
Speaker 6: so interconnected globally. You have foundries in Taiwan making chips

641
00:30:19,840 --> 00:30:22,680
Speaker 6: that are designed here in the US, and and so

642
00:30:22,880 --> 00:30:25,520
Speaker 6: all the production and design are happening in a lot

643
00:30:25,640 --> 00:30:29,280
Speaker 6: of different places. So when you see tariffs and export

644
00:30:29,400 --> 00:30:33,000
Speaker 6: controls imposed, you know, the trade flows and the impacts

645
00:30:33,160 --> 00:30:37,040
Speaker 6: really ricochet through the supply chain and all the way through.

646
00:30:37,640 --> 00:30:39,840
Speaker 6: You know, when you get to end use products like

647
00:30:40,200 --> 00:30:42,880
Speaker 6: computers and like cars, you know where we're going beyond

648
00:30:42,920 --> 00:30:45,120
Speaker 6: the data centers that we think of with Nvidio, But

649
00:30:45,600 --> 00:30:49,960
Speaker 6: speaking more broadly on the semiconductor sector, the ripple effect

650
00:30:50,080 --> 00:30:53,080
Speaker 6: through the supply chain will be really strong.

651
00:30:53,280 --> 00:30:53,600
Speaker 4: Michael.

652
00:30:53,720 --> 00:30:56,200
Speaker 3: And you know, certainly something you guys and you are,

653
00:30:56,440 --> 00:30:58,440
Speaker 3: you know, on top of on a regular basis in

654
00:30:58,520 --> 00:31:00,040
Speaker 3: terms of things coming out of the White Houspital. Do

655
00:31:00,120 --> 00:31:03,040
Speaker 3: you wonder about when it comes to ai and what

656
00:31:03,480 --> 00:31:06,720
Speaker 3: President Trump wants to see. Obviously he wants to protect

657
00:31:07,160 --> 00:31:10,520
Speaker 3: the US industry, the US chip industry, but so many

658
00:31:10,560 --> 00:31:13,680
Speaker 3: of these big tech companies here in the United States

659
00:31:13,720 --> 00:31:17,200
Speaker 3: are global players. It's a fine dance, right in terms

660
00:31:17,280 --> 00:31:21,680
Speaker 3: of their top line, their revenue, their growth. Yes, the

661
00:31:21,800 --> 00:31:25,040
Speaker 3: US market obviously important, but global markets are important as well.

662
00:31:25,880 --> 00:31:28,120
Speaker 6: That's a great point, Carol, because when you think about

663
00:31:28,240 --> 00:31:32,480
Speaker 6: one of the best known US chip companies, it also

664
00:31:32,640 --> 00:31:37,080
Speaker 6: happens to be one that is struggling existentially right now,

665
00:31:37,120 --> 00:31:40,600
Speaker 6: and that is Intel. And it has sales overseas, but

666
00:31:40,920 --> 00:31:46,040
Speaker 6: including in China. So any export restrictions, any move that

667
00:31:46,280 --> 00:31:49,640
Speaker 6: could affect its ability to do business in China could

668
00:31:49,760 --> 00:31:52,880
Speaker 6: be bad for Intel and in Vidia as well. You

669
00:31:52,960 --> 00:31:55,280
Speaker 6: know the You know, it's not like Nvidia has no

670
00:31:55,600 --> 00:32:00,320
Speaker 6: business in China, but the most sophisticated chips are or

671
00:32:00,440 --> 00:32:04,600
Speaker 6: off limits. Now the administration is wighing weather to lower

672
00:32:04,680 --> 00:32:09,760
Speaker 6: the threshold on those limits, making even less advanced ships

673
00:32:09,840 --> 00:32:13,320
Speaker 6: made by Nvidia and other makers too off limits for

674
00:32:13,480 --> 00:32:18,160
Speaker 6: Chinese buyers, barring a license of permission from the US

675
00:32:18,240 --> 00:32:21,040
Speaker 6: Commerce Department. So they are thinking about this, and we're

676
00:32:21,080 --> 00:32:24,000
Speaker 6: looking at a series of restrictions too, and these include

677
00:32:24,600 --> 00:32:28,440
Speaker 6: getting Japan and the Netherlands to get companies like Tokyo

678
00:32:28,520 --> 00:32:32,240
Speaker 6: Electron and ASML. These are the businesses that make the

679
00:32:32,400 --> 00:32:37,760
Speaker 6: semiconductor manufacturing equipment, these huge photo lithography machines that are

680
00:32:37,840 --> 00:32:41,120
Speaker 6: really finicky and expensive. What they want is they want

681
00:32:41,240 --> 00:32:46,840
Speaker 6: Japan and the Netherlands to limit maintenance of these machines

682
00:32:46,880 --> 00:32:49,520
Speaker 6: that have already been sold to China. You know, they're

683
00:32:49,560 --> 00:32:53,200
Speaker 6: looking at every which way to contain China in this area.

684
00:32:53,440 --> 00:32:54,840
Speaker 3: Michael, hang on for a second. I just want to

685
00:32:54,880 --> 00:32:57,400
Speaker 3: recap because we are seeing shares of Nvidia up about

686
00:32:57,440 --> 00:33:02,000
Speaker 3: two point two percent now in the postmarket trade, after

687
00:33:02,080 --> 00:33:05,360
Speaker 3: market trade. They've been bouncing around a lot gains and losses,

688
00:33:05,400 --> 00:33:07,840
Speaker 3: but it does look like they're kind of holding at

689
00:33:07,880 --> 00:33:10,800
Speaker 3: this level again, a two point two percent gain. Keep

690
00:33:10,840 --> 00:33:13,160
Speaker 3: in mind, I'm looking at our Ian King his reporting

691
00:33:14,080 --> 00:33:17,440
Speaker 3: and video giving solid quarterly results and a bullish revenue

692
00:33:17,480 --> 00:33:20,680
Speaker 3: forecast for the current period, even if the numbers didn't

693
00:33:20,760 --> 00:33:23,800
Speaker 3: reach the blowout level that some investors were banking on.

694
00:33:23,920 --> 00:33:26,680
Speaker 3: And of course we're waiting for some headlines tim here

695
00:33:27,040 --> 00:33:30,200
Speaker 3: as the call will get underway soon with the C

696
00:33:30,360 --> 00:33:32,560
Speaker 3: suite and Jensen Wong of Nvidia so big.

697
00:33:32,680 --> 00:33:35,040
Speaker 2: I think the big question that I have is how

698
00:33:35,120 --> 00:33:37,440
Speaker 2: much Insight will get And we don't know this yet, Michael,

699
00:33:37,480 --> 00:33:39,800
Speaker 2: but how much Insight will get on the call into

700
00:33:40,320 --> 00:33:46,239
Speaker 2: invidious relationship with China and to what extent restrictions are

701
00:33:46,360 --> 00:33:47,760
Speaker 2: set to affect the company.

702
00:33:49,360 --> 00:33:52,720
Speaker 6: And that is a key question for investors going into

703
00:33:52,800 --> 00:33:57,080
Speaker 6: this because when Deep Seek emerged onto the scene, and

704
00:33:57,200 --> 00:33:59,840
Speaker 6: remember Deep Seek was a trigger for the Trump administration

705
00:34:00,080 --> 00:34:04,080
Speaker 6: to think about tougher restrictions on Nvidia's chips, and it

706
00:34:04,200 --> 00:34:09,080
Speaker 6: was also an existential moment for in Vidia itself. You know,

707
00:34:09,200 --> 00:34:12,719
Speaker 6: Deepseeak was able to produce its chatbot on the cheap,

708
00:34:12,960 --> 00:34:16,800
Speaker 6: undermining this whole idea that hey, we need Manhattan Project

709
00:34:16,960 --> 00:34:21,600
Speaker 6: level CAPEX to be able to produce these breakthroughs in AI,

710
00:34:21,760 --> 00:34:24,560
Speaker 6: and of course to do that you need tons and

711
00:34:24,680 --> 00:34:29,839
Speaker 6: tons of Nvidia accelerators. Deep Seek raise questions about, well,

712
00:34:30,000 --> 00:34:33,920
Speaker 6: maybe we don't need quite so many work quite so much. Nonetheless,

713
00:34:34,000 --> 00:34:38,280
Speaker 6: we've seen the hyperscalers all, you know, reinforce their commitment

714
00:34:38,440 --> 00:34:42,120
Speaker 6: to heavy CAPEX tens and tens of billions of dollars

715
00:34:42,760 --> 00:34:45,880
Speaker 6: in plans for data centers and chip purchases.

716
00:34:46,360 --> 00:34:49,000
Speaker 3: All right, So appreciate being able to get this perspective

717
00:34:49,040 --> 00:34:50,440
Speaker 3: and roll it in. It's such a big part of

718
00:34:50,480 --> 00:34:53,880
Speaker 3: the Nvidia story. Michael Sheppard, thank you, as always, Senior

719
00:34:53,960 --> 00:34:57,400
Speaker 3: editor for Technology and Strategic Industries out there in our

720
00:34:57,440 --> 00:34:59,080
Speaker 3: Bloomberg News Bureau in DC,