Yellowbrick- An efficient Cloud Data Warehouse powered by Dell Technologies
Download PDFMon, 29 Jan 2024 23:20:57 -0000
|Read Time: 0 minutes
In the current economic climate, CIOs are rethinking their cloud strategy. They face challenges on several fronts - the need to continue innovating and driving growth while reducing the cost of cloud data programs and bringing tangible value. As cloud economics practices mature, private cloud and hybrid cloud are regaining strategic impetus. Organizations need the flexibility to manage data in private cloud, public cloud, co-lo, and at the edge. Yellowbrick delivers on this “Your Data Anywhere” vision.
Alongside new data management approaches such as data lakes, SQL based Data Warehouse technologies continue to prove their value as the primary business interface, with data lake vendors rushing to emulate their capabilities.
With Dell Technologies’ this solution is designed and optimized to provide an elastic data management platform for SQL analytics at any scale.
Business Challenges and Benefits
Yellowbrick data warehouse meets these challenges with a unique architecture designed to maximize efficiency with hardened security and simplified management. Yellowbrick delivers everything you would expect from a modern high-performance SQL cloud data warehouse.
It comes with cloud SaaS simplicity and elasticity with performance perfected through years of delivering value to customers in weeks and months and bills natively to exploit the power agility of the cloud.
Yellowbrick uniquely combines its MPP database software, and highly engineered systems design, with an agile elastic modern Kubernetes-based architecture that delivers high efficiency and maximizes performance in every deployment scenario.
Yellowbrick is engineered for maximum efficiency and price performance, supporting thousands of concurrent users on 1/5 of the cloud resources compare with competitors, maximizing data value with the simplicity and familiarity of SQL but with a unique pricing model that alleviates concerns over unpredictable cost overruns.
Who is Yellowbrick?
The Yellowbrick Data Warehouse is an elastic massively parallel processing (MPP) SQL database that runs on-premises, in the cloud, and at the network edge, it was designed for the most demanding batch real time and ad hoc and mixed workloads and can run complex queries at up to petabyte scale with guaranteed sub second response times. Yellowbrick is proven, providing business critical services at many large global enterprises with thousands of concurrent users. It is available on AWS, Azure, and Google Cloud as well as on-premises.
SQL Analytics for The Masses Cost-effectively supporting thousands of concurrent users running hundreds of concurrent ad-hoc queries, Yellowbrick leapfrogs competitors while still providing full elasticity with separate storage and compute. | |
Meet Mission-Critical Service Levels Intelligent workload management dynamically optimizes resources to ensure SLAs are consistently met without the need to scale out and spend more. | |
Ultimate Control of Data Security Yellowbrick’s data warehouse runs in your own cloud VPC or on-premises behind your firewall, allowing you to meet data sovereignty and governance requirements and pay for your own infrastructure. | |
Engineered for Extreme Efficiency and Performance Get answers faster with our Direct Data Path architecture. Yellowbrick runs mixed ad-hoc ETL, OLAP, and real-time streaming workloads delivering the maximum benefit from any underlying infrastructure platform. | |
Easy to Do Business With Optimize your costs with flexible on-demand or fixed subscription – Yellowbrick is invested in your success, not in emptying your wallet. Our NPS of 82 is a testament to our customer partnership model and support excellence. |
Figure 1 The Yellowbrick Advantage
Yellowbrick Overview
Designed to run complex mixed workloads and support ad-hoc SQL while computing correct answers on any schema, Yellowbrick offers massive scalability and supports vast numbers of concurrent users. This means our clients gain deeper, more meaningful insights into their customers more quickly than ever before possible, setting us apart from other cloud data warehouses (CDWs).
Figure 2 Yellowbrick Architecture
In an industry-first, full SQL-driven elasticity with separate storage and compute is available within your own cloud account as well as on-premises. Compute resources – elastic, virtual compute clusters (VCCs) – are created, resized, and dropped on-demand through SQL, and cache data persisted on shared cloud object storage. For example, ad-hoc users can be routed to one cluster, business-critical users to a second cluster, and more clusters created and dropped on demand for ETL processing.
Each data warehouse instance runs independently of one another. There is no single point of failure or metadata shared across instances. Global outages – when deployed with replication across multiple public clouds and/or on-premises – are impossible.
Yellowbrick is secure by default with no external network access to your database instance. Encryption of data at rest is standard with keys you manage. Columnar encryption, granular role-based access control, column masking, OAuth2, Active Directory, and Kerberos authentication are built in. Integrations with best-in-class enterprise data protection solutions secure PII data. Enterprise-class high availability, backups for data retention, and asynchronous replication for disaster recovery are standard. Management capabilities, Vantage offers significant value for your investment.
Yellowbrick powered by Dell Technologies
Yellowbrick and Dell share solutions that address a variety of data analytic use cases:
- Mission-critical Reporting and BI
- Data Warehouse modernization and consolidation
- Data-intensive B2B Apps and Data Monetization
- Hybrid Cloud Big Data Analytics
- Unified features store for data science and AI
- Multi-PB scale relational data lake
Symphony Retail AI
Symphony RetailAI serves the ever-changing consumer goods industry. That means they need to transfer terabytes of raw data to their 700 TB data warehouse and quickly convert it into easily digestible information for their consumers. Development and test, departmental data marts, self-service analytic workspaces for data scientists and developers, and edge/IoT computing.
TEOCO powered by Dell Technologies
TEOCO (The Employee-Owned Company) is a leading provider of telecom industry analytics and optimization solutions. The company provides intelligence about revenue assurance, network quality, and customer experience to more than 300 providers and customers. In addition to managing mountains of data for their clients, TEOCO also develops algorithms to transform raw data into actionable insights.
With these game-changing responsibilities in mind, TEOCO constantly strives to improve data warehouse innovation.
Some of the use cases <insert use case introduction>
Catalina Marketing powered by Dell Technologies
Catalina Marketing is the industry leader in consumer intelligence as well as in targeted instore and digital media. The company delivers an annual $6.1 billion in consumer value by pairing its exceptional analytics and insights with the richest buyer-history database in the world. To fulfill its mission, Catalina processes terabytes of data, transforming it into meaningful results so companies can optimize media planning to increase consumer engagement.
Catalina’s complex extract, transform, and load (ETL) processes required nightly conversions to produce data sets for querying and reporting. Plus, Catalina’s team of about 100 data scientists used advanced analytics and data-mining tools to perform large, ad hoc queries for a variety of customers.
Luis Velez, data engineering manager at Catalina explained that before Yellowbrick “It was an unsustainable environment in which we were not able to finish our data loads because we had 15 to 20 queries running at any given time.” “Every day, it was getting a little bit worse.” “Sometimes queries took hours, and other times they were simply killed so ETL processes could run,” says Aaron Augustine, executive director of data science at Catalina.
To achieve optimal results, Catalina incorporated Yellowbrick into its system, dividing the computing workload in half between the two platforms. Netezza would handle data processing, while Yellowbrick supported the consumption of processed data. During a three-week Proof of Technology (POT) exercise, Catalina found Yellowbrick’s single 10U, 30-node system performed 182X better than their current system. Catalina switched immediately.
The Enterprise Data Warehouse is powered by the Dell PowerEdge R660 server, together with Dell PowerSwitch networking and ECS storage featuring capacity, performance, and operational simplicity.
Dell Infrastructure Components
The following Dell components provide the foundation for the Yellowbrick private cloud solution.
Figure 3 Dell Yellowbrick Solution
Dell PowerEdge R660 Server is the ideal dual-socket 1U rack server based on Intel’s fourth-generation Xeon Scalable “Sapphire Rapids” processors for dense scale-out data center computing applications. Benefiting from the flexibility of 2.5” or 3.5” drives, the performance of NVMe, and embedded intelligence, it ensures optimized application performance in a secure platform.
The server is designed with a cyber-resilient architecture, integrating security deep into every phase in the life cycle. It has intelligent automation with integrated change management capabilities for update planning and seamless and zero-touch configuration. And it has built-in telemetry streaming, thermal management, and RESTful APIs with Redfish that offer streamlined visibility and control for better server management.
Dell ECS Storage is an enterprise-grade, cloud-scale, object storage platform that provides comprehensive protocol support for unstructured object and file workloads on a single modern storage platform. Either the ECS EX500 or EX5000 may be used depending on capacity requirements.
Dell PowerSwitch Networking switches are based on open standards to free the data center from outdated, proprietary approaches: They support future ready networking technology that helps you improve network performance, lower network management costs and complexity, and adopt new innovations in networking.
Why Dell Technologies
The technology required for data management and enterprise analytics is evolving quickly, and companies may not have experts on staff or who have the time to design, deploy, and manage solution stacks at the pace required. Dell Technologies has been a leader in the Big Data and advanced analytics space for more than a decade, with proven products, solutions, and expertise. Dell Technologies has teams of application and infrastructure experts dedicated to staying on the cutting edge, testing new technologies, and tuning solutions for your applications to help you keep pace with this constantly evolving landscape.
Dell Technologies is building a broad ecosystem of partners in the data space to bring the necessary experts, resources, and capabilities to our customers and accelerate their data strategy. We believe customers should be able to innovate using data irrespective of where it resides across on-premises, public cloud and edge. By partnering with Teradata, an industry leader in enterprise data management and analytics, we are creating optimized solutions for our customers.
Dell Technologies uniquely provides an extensive portfolio of technologies to deliver the advanced infrastructure that underpins successful data implementations. With years of experience and an ecosystem of curated technology and service partners, Dell Technologies provides innovative solutions, servers, networking, storage, workstations, and services that reduce complexity and enable you to capitalize on a universe of data.
Conclusion
Whether you want to expand your existing capabilities or get started with your first project, Yellowbrick powered by Dell Technologies offers XYZ. For more information about the solutions, please contact the Dell Technologies Teradata Solutions team by email.
Your company needs all tools and technologies working in concert to achieve success. Fast, effective systems that complement time management practices are crucial to making the most out of every employee hour. High-level data collection and processing that provides rich, detailed analytics can ensure your marketing campaigns strategically target your ideal customers and encourage conversion. To top it off, you need affordable products that meet your criteria and then some. After switching to Yellowbrick, our customers have seen dramatic gains in efficiency:
- Streamlined processes.
- Faster query times.
- Minimized data turnaround time.
- Richer, more accurate data.
- Increased customer growth.
- Affordable pricing with fixed-rate subscriptions for any deployment.
- No hidden fees or quotas.
- Predictable and reliable performance.
- Compatible with other components and applications.
- Highly capable system portability and accessibility.
- Innovative solutions.
- Little to no performance tuning.
- Ability to support a multitude of concurrent users.
Enjoy quick, easy, and supportive migration
At Yellowbrick, we are ready to provide you with simple, swift migration services. We complete most migrations in weeks, not months. Our 15-day proof of concept performance and operational testing period allows you to confirm that Yellowbrick is the right fit for your company. During this time, we will work closely with you to understand the requirements and scope a POC in your data center or in the cloud—whichever you prefer. We will set up a test instance, migrate your data, and integrate all necessary applications.
Since Yellowbrick is based on PostgreSQL, the world’s most advanced open-source database, and natively supports stored procedures, it works out of the box quickly. Our data solutions are also compatible with common industry tools, such as Tableau, MicroStrategy, SAS, and Microsoft Power BI, as well as Python and R programming languages. Coupled with one day of setup and one week of testing, your team can hit the ground running almost immediately.
Additionally, our broad partner network can help plan your transition, understand your data flows, and manage cutover with purpose-built tools and consulting services, so you can migrate from any platform.
Additional Resources
For more information, please see the following resources:
Related Documents
Powering Kafka with Kubernetes and Dell PowerEdge Servers with Intel® Processors
Mon, 29 Jan 2024 23:33:38 -0000
|Read Time: 0 minutes
Kafka with Kubernetes
At the top of this webpage are 3 PDF files outlining test results and reference configurations for Dell PowerEdge servers using both the 3rd Generation Intel® Xeon® processors and 4th Generation Intel Xeon processors. All testing was conducted in Dell Labs by Intel and Dell Engineers in October and November of 2023.
- “Dell DfD Kafka ICX” – highlights the recommended configurations for Dell PowerEdge servers using 3rd generation Intel® Xeon® processors.
- “Dell DfD Kafka SPR” – highlights the recommended configurations for Dell PowerEdge servers using 4th generation Intel® Xeon® processors.
- “Dell DfD Kafka Kubernetes Test Report” – Highlights the results of performance testing on both configurations with comparisons that demonstrate the performance differences between them.
Solution Overview
The Apache® Software Foundation developed Kafka as an Open Source solution to provide distributed event store and stream processing capabilities. Apache Kafka uses a publish-subscribe model to enable efficient data sharing across multiple applications. Applications can publish messages to a pool of message brokers, which subsequently distribute the data to multiple subscriber applications in real time.
Kafka is often deployed for mission-critical applications and streaming analytics along with other use cases. These types of workloads require leading-edge performance which places significant demand on hardware.
There are five major APIs in Kafka[i]:
- Producer API – Permits an application to publish streams of records.
- Consumer API – Permits an application to subscribe to topics and process streams of records.
- Connect API – performs the reusable producer and consumer APIs that can link the topics to the existing applications.
- Streams API – This API converts the input streams to output and produces the result.
- Admin API – Used to manage Kafka topics, brokers, and other Kafka objects.
Kafka with Dell PowerEdge and Intel processor benefits
The introduction of new server technologies allows customers to deploy solutions using the newly introduced functionality, but it can also provide an opportunity for them to review their current infrastructure and determine if the new technology might increase performance and efficiency. Dell and Intel recently conducted testing of Kafka performance in a Kubernetes environment and measured the performance of two different compression engines on the new Dell PowerEdge R760 with 4th generation Intel® Xeon® Scalable processors and compared the results to the same solution running on the previous generation R750 with 3rd generation Intel® Xeon® Scalable processors to determine if customers could benefit from a transition.
Some of the key changes incorporated into 4th generation Intel® Xeon® Scalable processors include:
- Quick Assist Technology (QAT) to accelerate data compression and encryption.
- Support for 4800 MT/s DDR5 memory
Raw performance: As noted in the report, our tests showed a 72% producers’ latency decrease with gzip compression and a 62% producers’ latency decrease with zstd compression.
Conclusion
Choosing the right combination of Server and Processor can increase performance and reduce time, allowing customers to react faster and process more data. As this testing demonstrated, the Dell PowerEdge R760 with 4th Generation Intel® Xeon® CPUs significantly outperformed the previous generation.
- The Dell PowerEdge R760 with 4th Generation Intel® Xeon® Scalable processors delivered:
- 62% faster processing using zstd compression
- 72% faster procession using gzip compression
- 4th Generation Intel® Xeon® Scalable processors benefits are the results of:
- Innovative CPU microarchitecture providing a performance boost
- Introduction of DDR5 memory support
[i] https://en.wikipedia.org/wiki/Apache_Kafka
Llama-2 on Dell PowerEdge XE9640 with Intel Data Center GPU Max 1550
Fri, 12 Jan 2024 18:04:24 -0000
|Read Time: 0 minutes
Part two is now available: https://infohub.delltechnologies.com/p/expanding-gpu-choice-with-intel-data-center-gpu-max-series/
| MORE CHOICE IN THE GPU MARKET
We are delighted to showcase our collaboration with Intel® to introduce expanded options within the GPU market with the Intel® Data Center GPU Max Series, now accessible via Dell™ PowerEdge™ XE9640.
The Intel® Data Center GPU Max Series is Intel® highest performing GPU with more than 100 billion transistors, up to 128 Xe cores, and up to 128 GB of high bandwidth memory. Intel® Data Center GPU Max Series pairs seamlessly with Dell™ PowerEdge™ XE9640, Dell™ first liquid-cooled 4-way GPU platform in a 2u server.
Dell™ recently announced partnerships with both Meta and Hugging Face to enable seamless support for enterprises to select, deploy, and fine-tune AI models for industry specific use cases anchored by Llama 2 7B Chat from Meta.
We put Dell™ PowerEdge™ XE9640 and Intel® Data Center GPU Max Series to the test with the Llama-2 7B Chat model. In doing so, we tested the tokens per second and the number of concurrent users that can be supported while scaling up to four GPUs. Dell™ PowerEdge™ XE9640 and Intel® Data Center GPU Max Series showcased a strong scalability and met target end user latency goals.
“Scalers AI™ ran eight concurrent processes of Llama-2 7B Chat with Dell™ PowerEdge™ XE9640 and Intel® Data Center GPU Max Series for a total throughput of >107 tokens per second, achieving our end user token latency target of 100 milliseconds”
Chetan Gadgil, CTO at Scalers AI
| LLAMA-2 7B CHAT MODEL
Large Language Models (LLMs) are powerful deep learning architectures that have been pre-trained on large datasets such as OpenAI ChatGPT. We have chosen to test Llama-2 7B Chat because it is an open source model that can be leveraged for commercial use cases, such as coding, functional tasks, and even creative tasks.
For inference testing in Large Language Models such as Llama-2 7B Chat, GPUs are incredibly useful due to their parallel processing architecture which can handle Llama-2's massive parameter sets. To efficiently handle expanding datasets, powerful GPUs such as Intel® Data Center GPU Max 1550 are critical.
| ARCHITECTURE
We started our testing environment with Dell™ PowerEdge™ XE9640 with four Intel® Data Center GPU Max 1550, running on Ubuntu 22.04.
We used Hugging Face Optimum, an extension of Transformers that provides a set of performance optimization tools to train and run models on targeted hardware, ensuring maximum efficiency. For Intel® Data Center GPU Max 1550, we selected the Optimum-Intel package. Optimum-intel integrates libraries provided by Intel® to accelerate end-to-end pipelines on Intel®. With Optimum-intel you can optimize your model to Intel® OpenVINO™ IR format and attain enhanced performance using the Intel® OpenVINO™ runtime.
Dell™ PowerEdge™ XE9640 Intel® Data Center GPU Max 1550
Source: https://www.dell.com/ Source: https://www.intel.com
| SYSTEM SET-UP SETUP
1. Installation of Drivers
To install drivers for the Intel® Data Center GPU Max Series, we followed the steps here.
2. Verification of Installation
To verify the installation of the drivers, we followed the steps here.
3. Installation of Docker
To install Docker on Ubuntu 22.04.3., we followed the steps here.
| RUNNING THE LLAMA-2 7B CHAT MODEL
1. Set up a Docker container for all our dependencies to ensure seamless deployment and straightforward replication:
sudo docker run --rm -it --privileged --device=/dev/dri --ipc=host intel/intel-extension-for-pytorch:xpu-max-2.0.110-xpu bash
2. To install the Python dependencies, our Llama-2 7B Chat model requires:
pip install openvino==2023.2.0
pip install transformers==4.33.1
pip install optimum-intel==1.11.0
pip install onnx==1.15.0
3. Access the Llama-2 7B Chat model through HuggingFace:
huggingface-cli login
4. Convert the Llama-2 7B Chat HuggingFace model into Intel® OpenVINO™ IR format using Intel® Optimum to export it:
from optimum.intel import OVModelForCausalLM
from transformers import AutoTokenizer
model_id = "meta-llama/Llama-2-7b-chat-hf"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = OVModelForCausalLM.from_pretrained(model_id, export=True) model.save_pretrained("llama-2-7b-chat-ov")
tokenizer.save_pretrained("llama-2-7b-chat-ov")
5. Run the code snippet below to generate the text with the Llama-2 7B Chat model:
import time
from optimum.intel import OVModelForCausalLM
from transformers import AutoTokenizer, pipeline
model_name = "llama-2-7b-chat-ov"
input_text = "What are the key features of Intel's data center GPUs?"
max_new_tokens = 100
# Initialize and load tokenizer, model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = OVModelForCausalLM.from_pretrained(model_name, ov_config= {"INFERENCE_PRECISION_HINT":"f32"}, compile=False)
model.to("GPU")
model.compile()
# Initialize HF pipeline
text_generator = pipeline( "text-generation", model=model, tokenizer=tokenizer, return_tensors=True, )
# Inference
start_time = time.time()
output = text_generator( input_text, max_new_tokens=max_new_tokens ) _ = tokenizer.decode(output[0]["generated_token_ids"])
end_time = time.time()
# Calculate number of tokens generated
num_tokens = len(output[0]["generated_token_ids"])
inference_time = end_time - start_time
token_per_sec = num_tokens / inference_time
print(f"Inference time: {inference_time} sec")
print(f"Token per sec: {token_per_sec}")
| ENTER PROMPT
What are the key features of Intel® Data Center GPUs?
Output
Intel® Data Center GPUs are designed to provide high levels of performance and power efficiency for a wide range of applications including machine learning, artificial intelligence and high performance computing.
Some of the key features of Intel® Data Center GPUs include:
1. High performance Intel® Data Center GPUs are designed to provide high levels of performance for demanding workloads, such as deep learning and scientific simulations.
2. Power efficiency.
| PERFORMANCE RESULTS & ANALYSIS
Figure: Comparing GPU vs CPU Performance
During the evaluation of the GPU configurations performance, we observed that a machine with a single GPU achieved a throughput of ~13 tokens per second across two concurrent processes. With two GPUs, we noted ~13 tokens per second across four concurrent processes for a total throughput of ~54 tokens per second. With four GPUs, we observed a total throughput of ~107 tokens per second supporting eight processes concurrently. The latency per process remains well below Scalers AI™ target of 100 milliseconds, despite an increase in the number of concurrent processes.
As latency represents the time a user must wait before task completion, it is a critical metric for hardware selection on large language models. This evaluation underscores the significant impact of GPU parallelism on both throughput and user response time. The scalability from one GPU to four GPUs reflects a significant enhancement in computational power, enabling more concurrent processes at nearly the same latency.
Our results demonstrate that Dell™ PowerEdge™ XE9640 with four Intel® Data Center GPU Max 1550 is up to the task of running Llama-2 7B Chat and meeting end user experience targets.
Number of GPUS | Throughput (Tokens/second) | Number of processes | Token Latency (ms) |
1 | 26.83 | 2 | 74.55 |
2 | 53.81 | 4 | 74.34 |
3 | 80.35 | 6 | 74.68 |
4 | 107.55 | 8 | 74.38 |
Table: Results after taking different number of GPUs
*Performance varies by use case, model, application, hardware & software configurations, the quality of the resolution of the input data, and other factors. This performance testing is intended for informational purposes and not intended to be a guarantee of actual performance of an AI application.
| ABOUT SCALERS AI™
Scalers AI™ specializes in creating end-to-end artificial intelligence (AI) solutions to fast track industry transformation across a wide range of industries, including retail, smart cities, manufacturing, insurance, finance, legal and healthcare. Scalers AI™ industry offerings include custom large language models and multimodal platforms supporting voice, image, and text. As a full stack AI solutions company with solutions ranging from the cloud to the edge, our customers often need versatile common off the shelf (COTS) hardware that works well across a range of workloads.
| Dell™ PowerEdge™ XE9640 Key specifications
MACHINE | Dell™ PowerEdge™ XE9640 |
Operating system | Ubuntu 22.04.3 LTS |
CPU | Intel® Xeon® Platinum 8468 |
MEMORY | 512Gi |
GPU | Intel® Data Center GPU Max 1550 |
GPU COUNT | 4 |
SOFTWARE STACK | Intel® OpenVINO® - 2023.2.0 transformers - 4.33.1 optimum-intel - 1.11.0" xpu-smi - 1.2.22.20231025 |
| HUGGING FACE OPTIMUM
Learn more: https://huggingface.co
| TEST METHODOLOGY
The Llama-2 7B Chat FP32 model is exported into the Intel® OpenVINO™ format and then tested for text generation (inference) using Hugging Face Optimum. Hugging Face Optimum is an extension of Hugging Face transformers and Diffusers that provides tools to export and run optimized models on various ecosystems including Intel® OpenVINO™.
For performance tests, 20 iterations were executed for each inference scenario out of which initial five iterations were considered as warm-up and were discarded for calculating Inference time (in seconds) and tokens per second. The time collected includes encode-decode time using tokenizer and LLM inference time.
Read part two: https://infohub.delltechnologies.com/p/expanding-gpu-choice-with-intel-data-center-gpu-max-series/