AI News from Nvidia, Microsoft, & Dell

Each week, TechTarget's news team and industry experts provide candid insight and analysis of the biggest IT news headlines of the week. On today's show, the experts discuss Google's new managed service for connecting public clouds; the latest AI news from Nvidia, Dell, and Microsoft; and Snowflake's Neeva acquisition. 

 

Watch Tech News this Week on the TechTarget News Bytes Channel:

Tech News this Week

 

Stories featured on Tech News this Week:

Dell enters generative AI market with Nvidia partnership

 

Episode Transcript

Antone Gonsalves (00:02):
Hi, and welcome to Tech News this week,
I'm your host techtargeteditorial news director and
Tonkin Sophos. On today's show,we'll discuss Google's new
managed service for connectingpublic clouds. The latest AI
news from Dell, Microsoft in thevideo, and snowflakes
acquisition of Neva for AIpowered search. This week,

(00:24):
Google Cloud launched a managedservice that provided provided
interconnectivity with the cloudinfrastructures of rivals, AWS,
Microsoft, Azure, and OracleCloud. Here to discuss the multi
cloud networking markets and howits evolving is analyst Bob

(00:44):
Lambert from tech targetsEnterprise Strategy Group.
Welcome, Bob.

Unknown (00:49):
Thanks, Anton. Great to be here.

Antone Gonsalves (00:52):
Yeah, great to have you. All right. So what's
driving Google Cloud to provideinterconnectivity with with
competitors? You know, what's,what is it hearing from its
customers?

Unknown (01:02):
Yeah, absolutely. I think what we've seen in our
research, is that, you know,the, the evolution to Cloud has
occurred at a fairly blisteringpace. You know, if we were to
look back 10 years ago, we wouldhave seen that maybe 17 20% of
organizations were using publiccloud infrastructure as a
service. Today, its use isvirtually ubiquitous, right? 95

(01:24):
plus percent of organizationsare using infrastructure as a
service. What's become moreinteresting over the last three
years, and this is just a, maybean indication of how fast things
are evolving. Just three yearsago, when we asked if
organizations were usingmultiple public clouds, like
infrastructure, onlyInfrastructure as a Service and
platform as a service, not notSAS, I don't want to include

(01:45):
that as part of our discussion.Three years ago, a majority of
organizations said they wereusing them. But when we asked
them and double click and said,Well, how are you using them,
the vast majority, were stillusing one and kicking the tires
with the others move forwardthree years to our latest
research. And what we found isthat about 85 90% of
organizations are using multipleclouds, and over half of them

(02:07):
are using them in a meaningfulway. So we're seeing this
transition to not only we'reusing the cloud, but we're using
multiple public, multipledifferent public cloud vendors.
And so that's really sparking aneed for organizations to be
able to connect those clouds upand do so in an operationally

(02:27):
efficient way. Right? Obviously,it's getting more complex if
you're starting to put differentapplications on different cloud
providers, even if it'sspreading out workloads across
different cloud providers. Sothat's really driving the need
for organizations to havesolutions that enable them to
connect more easily to thesemultiple public cloud
environments.

Antone Gonsalves (02:46):
Yeah, I think I mean, I saw I was looking at
the ESG study yesterday, and thenumber was like 97% of the IT IT
pros and networking folks thatyou you've surveyed. Were using
multiple clouds. That's that'sthat I was surprised. That's a
big number. What isinterconnectivity between the

(03:08):
public clouds, but what otherparts of this of this puzzle,
you know, have to be, are neededfor customers? I mean,
obviously, they don't haveeverything that they need at
this

Unknown (03:19):
point. Yeah, no, it's very intricate question, because
what we saw today in theannouncement this week from
Google, in their cross CloudConnect or interconnect service,
is that they're enablingorganizations to directly go
from a, you know, Google Cloudproperty over to an Azure or to
an AWS via, he looked like anEquinix, based on the demos that

(03:41):
I saw, things like that. So ableto have that cross connect
capability. The otherinteresting piece around this
multiple, multiple public cloudenvironment, is the need for
organizations to be able toconnect to multiple different
clouds without having to learnall of the nomenclature and
tools and technology of eachcloud. So when you see companies
like Aviatrix, proximo, Cisco,VMware, all of these

(04:06):
organizations are all buildingout solutions that will enable
organizations to connect tomultiple public clouds, but do
so in an abstracted way. So theycan learn that tool and then
connect to any right so spinningone up is just as easy as the
next without having to gothrough the process of learning
all of the differenttechnologies, and all of the

(04:27):
different procedures that arerequired for those cloud
vendors.

Antone Gonsalves (04:31):
So also important to point out, I think
that, you know, what Google isproviding here, and I believe,
Microsoft, Microsoft, as yourand AWS has, you know, over
quote, cloud or some things havesimilar interconnectivity that
this is really cloud to cloud,you still have to get from the
data center to the cloud. Andthat's where, you know, I talked

(04:55):
to a company yesterday graphient. You know, that's, that's
what they do and they're notworried about, you know, what
the cloud vendors are doingbetween each other? It's more
getting that getting that datacenter, you know, a Cisco
environment or something to thecloud, you know, in the data
center?

Unknown (05:12):
No, I think I was thinking, right, you know, right
now organizations are thinkingabout how do they expand their
reach into the cloud and get tomultiple different ones.
Eventually, as they build outthose workloads or applications
that are distributed acrossmultiple public clouds, that
interconnection directly betweenthe clouds are going to be a lot
more important. And so I thinkwe're gonna see that coming,
especially when, you know,people like Oracle, where they

(05:32):
have their databases there, andthey want to connect to a front
end that somewhere else. So Ithink that that cloud to cloud
connectivity is going to beimportant. Obviously, right now,
a lot of organizations arefiguring out, they're still in
that process of how to migratethose applications to one or
more public clouds. So thosesolutions, like you talked about
graphene at the Aviatrix is, etcetera, the Cisco, VMware, all

(05:53):
those are focused on how do weenable organizations to more
easily move those applicationsand connect to multiple
different public clouds?

Antone Gonsalves (06:00):
And is it? Is it too early to talk about
disruption? In other words,vendors that could technology
and use today that could bereplaced as, as the market
evolves?

Unknown (06:11):
Yeah, it's a good question. It's a really good
question. I think it's, it'sstill very early. Like I said,
only three years ago, no one wasreally doing multi, you know,
multiple public clouds inearnest. And so we're only about
three years into this marketnow. And we're seeing it rapidly
evolve. It's great. I thinkcompetition is always great.
Whether that is you know thatthe hard part you have with
maybe a cloud service provideris that maybe enables them you

(06:35):
have to learn their theirsolution. And then you can
connect, but from what I'veseen, you still have to know the
other cloud providerstechnology. So there's going to
be some pluses and minuses, somethings that organizations need
to look at. Those, thoseexisting ones today are going to
enable that cloud to cloud aswell. So I think there's going
to be some there's going to besome good competition. And it's

(06:55):
interesting when you see theheavyweights like the cloud
providers that Cisco is theVMware is getting into the game,
it's certainly going to driveand drive the innovation, and
which is always going to be goodfor the end user because it's
going to drive helps to drivethat operational efficiency,
right make those tools simplerto use, have them become more
powerful, more sophisticated, sothat it can more easily connect
to multiple clouds, multiplepublic clouds and across

(07:17):
multiple public clouds.

Antone Gonsalves (07:20):
Alright, so tech companies are rolling out
AI across product lines toattract customers and capture
the attention of Wall Street.This week, Vidya said it would
release supercomputing hardwarefor running AI based
applications. The announcementfollowed Microsoft releasing AI
developer tools, and partnershipdeals between Microsoft Dell

(07:44):
Technologies, and the video, nohit here to give us the why
behind the news is techtargeteditorials AI reporter Esther
Zhao. Alright, so let's startwith Microsoft, tell us how
Microsoft is helping developerstap into a

Unknown (08:01):
Yeah, so what they announced last week is kind of
like they're widening theirplugin partner, they're kind of
adopted the same calling plug inpartner as open AI who they're
actually partnered with. And sowhat they're doing is they
developers now have access tolike 50 plugins, through their a
new platform from Microsoft. Andso the way they kind of view

(08:23):
plugins are kind of like digitalassistants. So you can also as a
developer, create and test yourown plugin as well. And also,
like you are using the plugin tokind of be a hands on feet, so
to speak. When it comes topractical applications of stuff
like performing like practicalthings, like oh, yeah, let's
connect this logins chat to BT,and connected to other

(08:47):
applications like, I don't know,like Reddit and stuff like that
annoys or even like travelapplications, in order to be
able to do real attack likemaybe book a travel and stuff
like that.

Antone Gonsalves (09:00):
Yeah, that's always been me since it's, since
its inception, it's always beenMicrosoft strong point is its
developers and bringing theminto the whatever ecosystem
they're trying to build. They'realso they're also doing
something they're integratingtheir Bing search engine with
chat GPT you know, what's new onthat front?

Unknown (09:22):
Yeah, so since chat GPT was created I think a lot of
people have criticized the factthat number one, it's not up to
date, I believe it was like 2021that it was up to date which is
more than a year ago. So now ifI haven't been Interjet shabby
GBC being that is default searchengine that makes chatty PT a
lot more up to date so to speak.So now you if you as an X chat

(09:46):
GPT like what happened in thenews yesterday, because Bing
searches now as default searchengine that will make it up to
date, so that eliminates that.Weakness, I think is the word
when it comes to that TPC. Sothat's that Heelan

Antone Gonsalves (10:01):
Yes. And Microsoft was a Microsoft that
also updated its the data thatit has in chat GPT it's not
whereas open AI is, if you forthe free version is, I think
stops at 2021. But Microsoft hasupdated the data.

Unknown (10:24):
Basically, what I what I was saying is the fact that
now Microsoft has allowed thefree version, I believe, from
what I remember, Chad TPP Plususers now have access to like
that updated version. So that'sthe paid version, the free
version, I think it's coming onlater on.

Antone Gonsalves (10:42):
Okay. All right. So the on the
partnerships, Microsoft, AzureCloud has a new partnership. In
the video, Dell Technologieslaunched the new video
partnership called Projecthelix. You know, I'm gonna give
us a summary of each of those.

Unknown (10:58):
Yes, sir. So Microsoft, obviously they have a
partnership with open AI. So oneof the things that was asked to
me was like, Why do you needanother partnership with Nvidia
and Nvidia obviously is a strongas strong in the market. And so
like that allows Microsoft tokind of stretch his wings a
little bit more. So now it'sintegrated with Nvidia AI

(11:19):
enterprise AI. So they can alsohave quantum NVIDIA GPUs as part
of their stack. So that's goodfor Microsoft, in terms of Dell
and Nvidia and Nvidia for Dellis kind of entering the
generative AI space in terms ofthey're doing it a little bit
different in that they are theyhave generative AI capabilities

(11:41):
for onpremise, those who areinterested in gaining those
people loop capabilities and onpremise and I think one analyst
basically said, Oh, now they'reoffering generative AI in a box,
so to speak. And so now,enterprises who are interested
in stuff like that have accessto Dell, Dell hardware, as well
as Nvidia hardware and software.So they call it project Felix.

(12:05):
And so it gives generative AIcapabilities to onpremise users.
So you don't want to use in acloud, you want to use on
premise. It's an interestingbecause I think I was I think
this was the first I've heard ofsomething.

Antone Gonsalves (12:18):
Huh? Yeah. It makes sense that Dell would want
to say the hardware software ina box and you're good to go. And
I assume that they may have somea subscription fee involved
where you pay?

Unknown (12:31):
Yeah, yeah. I mean, yeah, it's what I've heard is
expensive, upfront costs, but ascheaper in the long run.

Antone Gonsalves (12:44):
All right. Finally, in the business
analytics market, snowflakerecently agreed to require Neva
Neva was a search engine, fueledby generative AI. Also Alteryx
added an AI engine to hisanalytics cloud. Here to
explain. The news is techtargeteditorials, business analytics

(13:07):
reporter Eric Hamadan. Alright,so how will Neva help data
management vendors snowflake,catch up with competitors, my
understanding is that snowflakewas falling behind with AI
services.

Unknown (13:23):
Well, I wouldn't say it was really falling behind it,
you know, it had some it hadsome AI capabilities. What it
really does is it bringssnowflake in line in terms of
generative AI. And then six orsix and a half months since open
AI. first launched Chet GPT,there's really been a slew of
data management and analyticsvendors that have either

(13:46):
integrated with open AI or othergenerative AI platforms or
develop their own generative AI.Got like Informatica and
Databricks, on the datamanagement side, Tableau
thoughtspot Power BI on theanalytics side. But really a lot
of what what they've announced,has been in just the last two

(14:08):
months or so. And all of whatthey have announced, is still in
preview. It's really justtheoretical. At this point, no
one has seen any finishedproducts on the debt data
management and analytics realm.So by acquiring Neva agreeing to
acquire Neva snowflake is reallyjust bringing itself in line

(14:30):
with those who have alreadyrevealed integrations or
internal developments that arein the works and slated for
general release sometime.Vaguely in 2023.

Antone Gonsalves (14:44):
Yeah, yeah, it seems like that seems common
right now in the industry wherethe company's vendors are making
the announcement that AI iscoming. They want to stay ahead.

Unknown (14:55):
I think it gives them a headline it gets them in the
news shows that they're at leastin an in on the game. But in
terms of what it will actuallyproduce what will actually see,
you know, it's wait and see.

Antone Gonsalves (15:09):
We'll have to wait and see. Sure. Now Neva
Neva didn't have a successfulproduct right before the
acquisition. So before snowflakecame to the rescue. So what was
the status of Neva before this?Yeah,

Unknown (15:25):
it was a search engine, that it had lofty goals of
competing with Google. But itwas going to be subscription
based, so that it didn't have tocharge for ads, you're just
gonna raise money throughsubscriptions. And it was going
to use generative AI topersonalize responses using
machine learning so that itwould learn about the its

(15:47):
subscribers and reallypersonalize the responses. But
it just didn't gain anytraction. So on May 23, I
believe it was that it said itwas going to suspend its search
engine and focus exclusively onJanuary of AI. And then it was
the very next day that snowflakerevealed that it was under

(16:07):
agreement to acquire Neva. So itlooked like the suspension of
the search engine was really aprecursor, you know, part of the
acquisition is really justacquiring capabilities. It
wasn't looking to acquirecustomers,

Antone Gonsalves (16:25):
and then trying to go head to head
against Google was probably notthe best strategy for a startup.

Unknown (16:32):
It had its you know, it had slight differentiation, but
it certainly wasn't enough toattract a large customer base.

Antone Gonsalves (16:39):
Correct. All right. So switching to Alteryx.
What are the AI capabilitiesthat companies bring into its
analytics cloud?

Unknown (16:47):
So it's kind of the same thing as all the other
ones. It's, it had its ownprevious pre existing AI
capabilities. And its move isgoing to be to combine those
with generative AI that it'sbringing in through
integrations. And it's, it'sit's doing, it's building a
search engine that it'sbuilding, excuse me, an engine

(17:09):
that it's calling aid in AI din. And it's really just a
combination of pre existing AIcapabilities with generative AI
capabilities brought in from theoutside.

Antone Gonsalves (17:21):
All right, that wraps up this week's show.
Thanks for watching and enjoythe weekend.