HP and NVIDIA: Building the Missing Link in AI Solutions

I’m at HP Amplify this week where HP prepares its partner sales channel for the coming year. I’ve long known that HP was ahead of its peers in enterprise-grade and printer security and in its partner program, but this was the first I’d heard of its new AI Studio, AI PC and AI workstation offerings that were created to give HP a competitive advantage as companies roll out internal AI solutions.

What makes HP’s approach different from the other companies I cover, particularly Dell, is that there is an actual solution, not just a vendor throwing hardware into the market. It was fascinating to hear Intel, NVIDIA, AMD, Google, and Microsoft stand up and praise HP’s effort at the event, but it wasn’t until I was briefed on AI Studio, a joint project with NVIDIA (that NVIDIA’s CEO Jensen Huang is unusually excited about) that I realized why HP was so far ahead.

Let’s talk about what makes HP’s approach to AI uniquely different from its peers with a particular focus on why NVIDIA, the clear leader in AI, is so excited about HP’s AI Studio offering.

Partners

The current big problem with AI is that few IT organizations are AI experts, and that is also true of companies that resell tech products. Yet, IT organizations look to their technology providers for help in understanding AI. Right now, it’s a blind-leading-the-blind problem that causes cost overruns and disappointing results.

I’ve been particularly frustrated with one of HP’s larger peers, who seems to think throwing hardware at the problem is the way to fix it. It’s not. You need to first understand AI by deploying it yourself and then train your partner sales channel so they too can provide adequate support.

At Amplify, HP announced AI training and certification and master class programs that weren’t conceived until HP became proficient in AI use (HP also demonstrated broad use of AI internally in its announcements this week). This both helps partners spin up their own AI programs internally and helps its clients come up to speed on the technology before forcing a bunch of hardware they weren’t prepared for onto them.

Granted, this program will take time to roll out, and you’ll want to ask any sales/deployment partner if they have this certification and how far along they are in understanding and deploying AI internally before selecting them. If you do, you’ll avoid the pain many are experiencing as the company learns AI on your nickel.

What I found fascinating was the CEO of Google was on stage (virtually) saying this is exactly the kind of program partners needed without yet knowing that this was the program HP was launching.

AMD vs. Intel vs. Qualcomm

While the three chip vendors were on stage, only two of the vendors announced products with their parts in them (Qualcomm-based products will launch later in the year). Intel had the most AI PC design wins, as was expected. However, on the mobile workstation side, HP indicated that while AMD and Intel products were priced the same, the AMD mobile workstation products had significantly better battery life and around 30% better performance.

This was interesting because, traditionally, AMD is the lower cost but lower performing option, but in workstations, AMD outperformed Intel at the same price, and those who use workstations really care about performance. In addition, they had a new tower workstation with AMD Threadripper and NVIDIA that will potentially rival Lenovo’s offering (Lenovo took over market leadership for workstations sometime back when it made a similar move that was unchallenged by either Dell or HP at the time).

These new workstations are specifically focused on AI projects.

AI Studio

AI Studio is a unique software offering developed in conjunction with NVIDIA to create a collaborative, coordinated tool to help data scientists build Large Language Models (LLMs) that form the data foundation for generative AI. Building a large LLM can cost upwards of $100M and is often plagued by poor coordination between data scientists and other partners who often don’t work that well together. HP’s AI Studio provides an AI-backed framework to help manage an LLM creation process. I was told that NVIDIA’s CEO was over the moon about this offering because it corrected a huge tool shortage problem in the AI creation space.

Wrapping Up

In their class, HP is currently in the lead in three areas. First, it has the strongest security effort in its Wolf Security offering, but it’s only for enterprises (Wolf’s impact on SMB and, particularly, consumer products either doesn’t exist yet or is very light). Second, its partner effort is second to none, which is kind of sad given that before Dell bought it, EMC had the top partner effort at the time. And third, HP has the leading AI solution that includes not only desktop hardware leadership (HP Inc. doesn’t make servers), but the leading collaboration and workstation creation solutions, as well (LLMs virtually all start out on workstations).

While the hardware works today as does AI Studio, HP’s partner channel has not yet spun up its support efforts, but partners are pivoting to certification and training now, suggesting that by year’s end, many should be up to speed. Since this is the first comprehensive AI effort of this type, that means HP’s peers, even if they emulate what HP has done, won’t be ready until well into 2025.