New data center chips join ecosystem

At the Cloud Next ’24 conference in Las Vegas from April 9 to 11, Google Cloud announced a new Chrome enterprise subscription service and a series of generative AI add-ons for Google Workspace. Overall, Google Cloud is putting its Gemini generative AI applications into as many places as possible; for example, the company is working on providing Vertex AI infrastructure for other companies’ artificial intelligence and hardware, such as the new Axion CPU.

We attended a pre-briefing session to get an early look at new features and tools, including a generative AI video service for marketing and internal communications. Here’s a summary of what we believe to be the most impactful enterprise news from Google Cloud Next.

Chrome Enterprise Premium adds security controls

Enterprise browsers may be an emerging trend. Last year, Gartner predicted that by 2030, browsers designed for enterprise-grade security and support would be widely adopted. Google is launching Chrome Enterprise Premium. This enterprise layer adds an extra level of security to Chrome, enhancing it by:

• Enterprise control over policy enforcement, software updates and extension management.
• Incident and device reporting.
• Forensic Security Report.
• Context-aware access control.
• Threats and data protection.

Chrome Enterprise Premium is available today, April 9, wherever Chrome is available. Costs $6 per user per month.

Axion is Google’s first Arm-based CPU

Some Google Cloud services, such as BigQuery, will soon run on Google Axion processors, Google’s first custom Arm-based CPU for data centers (Figure A). Google Cloud says Axion’s performance is 50% higher than the current generation of x86-based virtual machines. Executors on Google Compute Engine, Google Kubernetes Engine, Dataproc, Dataflow, Cloud Batch, and more will be available later in 2024.

Figure A

Vertex AI will be rooted in Google search

Starting April 9, models trained in Vertex AI will be rooted in Google search. This may be controversial depending on the data used for training.

Grounded artificial intelligence is part of the search-augmentation generation trend, which avoids “illusions” by checking artificial intelligence against real information. Google Cloud said that the implementation will provide “fresh, high-quality information” for models based on Vertex AI.

Adds a host of artificial intelligence capabilities to Google security products

Google Cloud’s other announcements focused on adding Gemini to its security offerings, a move likely to compete with Microsoft’s thorough integration of Copilot into its security suite.

Gemini AI is now available in Google Security Operations (Picture B) and threat intelligence. In particular, from the end of April, Gemini will be able to assist with Chronicle Enterprise and Chronicle Enterprise Plus investigations. Starting April 9, security analysts can use Gemini to converse with Mandiant threat intelligence software in natural language.

Picture B

Security Command Center gets a boost from Gemini; as of April 9, the AI ​​feature in preview can scan for threats based on natural language prompts and summarize alerts and attack paths.

“Detection engineers can more easily create detections and playbooks, and security analysts can quickly find answers through intelligent summarization and natural language search,” Ronald Smalley, senior vice president of cybersecurity operations at Fiserv, said in a Google Cloud press release. Critical as SOC teams continue to manage increasing data volumes and need to detect, verify and respond to incidents faster.”

Gemini Cloud Assist in Google Cloud Security Services will now include the following AI-driven features in preview:

• Identity and access management recommendations for improving your IAM posture.
• Assist in creating encryption keys.
• Recommendations on how to implement confidential computing protection.

Create videos using AI and more in new Google Workspace products

Google launches a new workspace platform called Google Vids (Figure C), generative AI can help employees create promotional or information-rich videos.

Figure C

First, Google Vids will create a storyboard that the user can customize. From there, Google Vids can create videos and add voiceovers using preset AI voices or custom audio.

Google Vids will become a new platform, alongside document sheets and slideshows. The video will be available through Workspace Labs in June.

More Google Workspace announcements

• An artificial intelligence conference summarization and translation tool that costs $10 per user per month.
• The AI ​​Security add-on costs $10 per user per month for file classification and protection.
• In Gmail, artificial intelligence will power the Polish My Draft and Help Me Write tools, the latter available through voice commands on mobile devices.
• Google Gemini will be used to create cover images in chats and documents.
• Vertex AI will be integrated into Google Workspace.
• HubSpot extension for Gemini for Google Workspace.

Gemini 1.5 Pro launched on Vertex AI and data cloud

Google has announced a number of changes to its data cloud portfolio across repositories and data analytics. The biggest of these is the availability of Gemini 1.5 Pro in Vertex AI, which supports 1 million token context windows.

See: Everything you need to know about Google Cloud Platform. (Technology Republic)

Google continues to expand Gemini integration and functionality – today’s previews are Gemini in Looker and Gemini in BigQuery. Other new features in BigQuery are now available in preview or private preview, including Vertex AI file and message insights, vector embedding and matching, and more.

Other announcements (preview unless otherwise noted) include:

• Google Gemini in the repository.
• Vector support in the database and LangChain integration.
• AlloyDB’s vector support, natural language support, and model endpoint management.
• Workloads are isolated from access to the analytics engine in Bigtable Data Boost.
• Expanded disaster recovery and sub-two-second maintenance in Cloud SQL (publicly available).
• New execution sizes for Memorystore (general release) and support for two new persistence options in Memorystore (preview).
• Encryption keys managed by the customer in Cloud Firestore.

TPU v5p is generally available

Google’s TPU v5p AI accelerator is now generally available, allowing organizations to use it for AI inference and training. In addition, TPU v5p now supports Google Kubernetes Engine.

Advancing supercomputing and cloud computing

Google Cloud will work with NVIDIA hardware to provide support for Google’s AI training infrastructure. Announcements from the Cloud Infrastructure Group include:

• The A3 Mega VM, powered by NVIDIA H100 Tensor Core GPU for large-scale AI training, will be generally available in May.
• Hyperdisk ML, a new block storage service for AI inference/serving workloads, is now available in preview.
• JetStream, a throughput- and memory-optimized inference engine for LLM training, is now available on GitHub.
• Dynamic Workload Scheduler, a service that manages resources on Google Cloud to optimize AI workloads based on computing power, is now available in preview.
• Duet AI for Developers is now Gemini Code Assist.

Google Cloud Competitors

Google’s extensive product catalog means it has many competitors in different areas. Especially in terms of generative artificial intelligence ecosystem and enterprise support, it competes with and often interoperates with Microsoft’s Copilot AI, AWS, IBM and SAP. Axion chips in particular could help Google enter the data center chip field, which is dominated by Amazon and NVIDIA.

TechRepublic is reporting on Google Cloud Next ’24 remotely.