Dell and Snowflake Simplify Data Access

[jonwayne235]

Today’s press release seems significant.

DELL generated $101 billion in revenue last year.
$34 billion of the total revenue came from infrastructure solutions, which includes storage and on premises sales.
So they must have many significant large customers with large troves of data that they could work on SNOW.

May 2, 2022 01:00 PM ET (BZ Newswire) – News
Dell Technologies World – Dell Technologies (NYSE:DELL) and Snowflake (NYSE:SNOW), the Data Cloud company, will work together to connect data from Dell’s industry leading enterprise storage portfolio with the Snowflake Data Cloud.

A first of its kind, this collaboration will help customers have greater flexibility operating in multi-cloud environments, meet data sovereignty requirements, and easily turn data into insights.

Once available, Dell and Snowflake customers will be able to use on-premises data stored on Dell object storage with the Snowflake Data Cloud while keeping their data local or seamlessly copying it to public clouds.

The companies will pursue product integrations and joint go-to-market efforts in the second half of 2022.

“As the industry leader in data storage, Dell Technologies provides customers with the tools needed to derive insights from their data wherever it resides,” said Jeff Boudreau, president, Dell Technologies Infrastructure Solutions Group. “With the industry’s first collaboration between on-premises storage systems and the Snowflake Data Cloud, we are able to expand Dell’s SaaS provider ecosystem for data insights on-premises and in public clouds.”

“Snowflake’s mission is to mobilize the world’s data by empowering organizations to eliminate silos and bring data together to unlock more value and deeper insights,” said Christian Kleinerman, Snowflake senior vice president of product. “This collaboration with Dell will allow organizations to gain more value from their on-premises data while leveraging the performance and simplicity of Snowflake’s platform and the powerful collaboration capabilities of the Snowflake Data Cloud.”

[5thhorseman]

Dell Infrastructure Solutions revenue has been flat/dropping for about four years now, which is what you’d expect from an on-prem solutions provider. So this seems more of a win for Snowflake than Dell as it attracts customers who need a hybrid on-prem/cloud solution due to data sovereignty issues. For Dell it’s just another feature to help retain customers.

https://www.statista.com/statistics/992556/dell-revenue-by-s…

Caitlin Gordon (Dell): “Customers want to keep some of the data in their data center, but they still want to use that with Snowflake’s cloud-based analytics,” Gordan said. “By not changing where the analytics run but keeping the data in place, we’ve been able to collaborate to enable Snowflake’s external table capability to actually read the data that exists on our on-premise object storage capabilities—which is either ECS or ObjectScale—so that you can actually read that data, keep it in place, but leverage that with other cloud-based data that you have in the Snowflake platform.”

https://www.crn.com/news/cloud/dell-snowflake-form-industry-…

This hybrid capability already exists via Snowflake’s external tables feature. It allows one to transparently include data from other sources when querying data in Snowflake. However, “querying data stored external to the database is likely to be slower than querying native database tables” as per Snowflake’s own documentation. So look for the “product integrations” to address this performance issue before the joint go-to-market efforts.

https://docs.snowflake.com/en/user-guide/tables-external-int…

[chang88]

One of Snowflake’s unique feature is that they let analyst query data that come from different sources. For example, you can have one table where the source is AWS S3, and another that came from Azure, and the analysts can join them on the fly in the same script to get the report they want within Snowflake and not having to involve engineering to move the data to a single cloud platform.

There’s been a lot of talks about going “multi-cloud” recently. One reason is because of companies merging and combining infrastructures, the other is that some companies are any to avoid complete lock-in to a specific cloud. I don’t think this will be a major contribution to Snowflake’s revenue in itself. But this is a big plus when choosing data warehouse during a large scale migration - you want to be able to keep the current analytics running without interruption when you move the backend data that powers the analytics across platforms. So this is one unique competitive advantage to Snowflake.

I think this extension of support to on prem data for Dell falls into the same category. If I want to move my data from Dell clusters to AWS, then this lets me do that in a manner that cause least interruption to analytical workflow.

[brittlerock]

I’ve been retired for over ten years now, so my observations may not be very accurate, but FWIW before I retired the cloud was just beginning to be something that was being considered. At that time I worked within a relatively small group of IT folks in the Enterprise Architecture group at The Boeing Company. I had no direct responsibilities related to the emerging cloud, but the guy who had this responsibility was a colleague and we often discussed how Boeing might benefit from this technology.

At the time it was pretty much a given that there were at least two classes of data that were unlikely to be migrated to the cloud and would most likely remain on premises for the foreseeable future. These two classes were operational data such as the active databases of transactional systems like ERP, HR, finance, etc. This data is highly volatile, the applications that use it require near instantaneous response times and system availability must be very highly reliable 24/7/365. I could be wrong (and there’s several people who are regulars on this board who can correct me if I am), but it seems unlikely that this data class would be migrated to Snowflake. While it’s true that this data is usually mirrored to remote storage devices in order to provide disaster recovery, I think it unlikely that Snowflake would find application in this capacity.

The other class of data that was retained on-prem and would not be a likely candidate for Snowflake storage was highly protected information, such as black project data (this is data related to secret government projects) and internal proprietary and secret information.

My point is that these are the classes of data that are most likely stored on-prem on Dell equipment and least likely to be migrated to Snowflake.

The data used primarily for analysis was primarily held on Teradata servers, although a fair amount was also held on general purpose UNIX servers. Snowflake has disclosed that their migration from Teradata is very low (I think on the order of 3%). The long and short of it is that I don’t think this announcement represents a big uptick in Snowflake data.

[coldmountain]

Times have changed a great deal since you were working. I just this minute finished a demo for an ERP system that we are considering for our (medium sized manufacturing) company that is web-based, that is, hosted in azure. All of the short listed options were also cloud based. Our HR system is also cloud based.

I have posted about this in the past on Snowflake threads, but another product we use – a route accounting software – I learned is hosted on snowflake, and the company that made that also has a warehouse management system that runs on snowflake. The latter was particularly surprising to me, because that is a motion control system, something that is generally considered ONLY the province of PLCs, on site and close to the action.

[squidboyhq]

Brittlerock,

I think it unlikely that Snowflake would find application in this [transactional] capacity.

While it’s true that SNOW isn’t geared toward housing transactional databases, there are oodles of use cases where it makes sense for a transactional database to be synced on a regular basis to SNOW for data modeling/analysis/AI/ML purposes. You don’t need current-up-to-the-microsecond data when you’re performing YOY trend analysis, for example.

would not be a likely candidate for Snowflake storage was highly protected information

I’m sure you’re right there are some super-duper double-secret project datasets out there that may never be entrusted to a cloud-based 3rd party like SNOW, I’d guess that’s a tiny slice of the gov’t data world and virtual noise from a SNOW TAM perspective. Since we aren’t likely to obtain any concrete stats about these datasets, I guess guessing will have to do.

I’ve been in SAAS development for over 20 years now. When I first started developing for the cloud around Y2K one of the biggest hurdles the company I worked for faced was convincing prospective customers that cloud-based software could be just as secure, and likely more secure, than software running on servers in their home office basement or wherever. Eventually the tide turned on that and within about 5 years several government agencies were among our largest and most lucrative customers.

In any case, based on this link it appears SNOW meets or exceeds multiple security compliance standards (unless they’re touting these reports’ availability but don’t comply - but I doubt that’s the case), so it would seem to me they would be an acceptable environment to host all but the most sensitive data:

https://www.snowflake.com/snowflakes-security-compliance-rep…

Cheers,

Eric
long SNOW

[jonwayne235]

I came across an incredibly detailed and long write up here (https://siliconangle.com/2022/05/07/may-not-know-dell-snowfl…) published today about SNOW and Dell’s deal.

I would greatly appreciate if the more tech-inclined folks here could read it and comment their insights.

A couple excerpts:

Cloud-native ISVs such as Snowflake are realizing that despite their cloud-only dogma, they have to grit their teeth and deal with on-premises data or risk getting shut out of evolving data architectures.

…Sometimes organizations can’t or won’t move data into the cloud for a variety of reasons, like data sovereignty, compliance, security policies, culture… whatever. So with this solution, data can remain in place on-prem or it can be migrated into the public cloud for the full cloud-native Snowflake experience… **but the data is read only. That means you can’t do… data manipulation language…**But the data can be queried. However, the performance of those queries…will almost certainly be slower…at the end of the day, it’s going to run faster in the cloud. You can be almost certain of that.

They also note that “after 13 straight surveys of astounding Net Scores, Snowflake has finally broken trend, with Net Score dropping below the 70% mark among all respondents.”

[Picture is here: https://d2axcg2cspgbkk.cloudfront.net/wp-content/uploads/Bre…

“Net Score is measured by asking customers if they’re adding the platform new (that’s the lime green in the bar pointing from Snowflake in the graph); are they increasing spend by 6% or more – that’s the forest green; is spending flat (the gray); is their spend decreasing by 6% or worse (pinkish); or are they decommissioning the platform (bright red, which is essentially zero for Snowflake). Subtract red from green and you get Net Score.

Now what’s somewhat interesting is that Snowflake’s Net Score overall in the survey is a 68 – still huge – but its Net Score inside Dell accounts drops to the low 60s. Nonetheless, this chart tells you why Snowflake: Its highly elevated spending momentum combined with an increasing presence in the market over the past two years makes it a perfect initial data platform partner for Dell. As well, Snowflake’s presence in Dell accounts is below the Snowflake average, which gives Snowflake an incentive to capture more Dell customers and expand its TAM.

Ok, there is still a LOT more to unpack from the article but I won’t do a great job summarizing or deciphering it short of just copying the whole article.

Look forward to anyone else’s comments.