“Software developers, data scientists, artists, designers and engineers need powerful AI and graphics performance to push the boundaries of visual computing and simulation, helping tackle incredible industry challenges,” said Bob Pette, vice president of enterprise platforms at NVIDIA. “Bringing NVIDIA Blackwell to workstations and servers will take productivity, performance and speed to new heights, accelerating AI inference serving, data science, visualization and content creation.”
Sitting at the top of the heap is the RTX Pro 6000 Blackwell Server Edition, which is designed for enterprise data center deployments. Externally, that entails a passively-cooled thermal design that can run 24/7 in data center environments. Customers also have the ability to configure up to eight GPUs for servers.
Additionally, NVIDIA says the Server Edition part can be combined with its vGPU software to run AI workloads across virtualized environments, giving remote workers access to high-performance virtual workstations. Support for this feature will arrive in the second half of 2025.
According to NVIDIA, its new RTX Pro 6000 Blackwell Server Edition delivers a multi-fold increase in performance across a whole bunch of enterprise workloads compared to its previous generation Ada Lovelace-based L40S GPU. More specifically, it’s claiming the Blackwell part delivers 5x higher large language model (LLM) inference throughput for agentic AI applications, close to 7x faster genomics sequencing, 3.3x faster text-to-video generation, and 2x speed increases for both inference for recommender systems and rendering.
“The RTX Pro 6000 Blackwell Server Edition also demonstrates game-changing acceleration for genomic analysis and drug discovery inference workloads, enabled by a new class of dynamic programming instructions,” NVIDIA says.
“On a single RTX Pro 6000 Blackwell Server Edition GPU, Fastq2bam and DeepVariant—elements of the NVIDIA Parabricks pipeline for germline analysis—run up to 1.5x faster compared with using an L40S GPU, and 1.75x faster compared with using an NVIDIA H100 GPU,” NVIDIA adds.
Another example shared by NVIDIA is OpenFold2, an AI model that predicts protein structures for drug discovery research. According to NVIDIA, its RTX Pro 6000 Blackwell Server Edition GPU enables up to a 4.8x inference performance boost compared to its L40S GPUs.
That’s the sole data center solution within the RTX Pro Blackwell family. Directly underneath sits the desktop lineup, which consists of the RTX Pro 6000 Blackwell Workstation Edition, RTX Pro 6000 Blackwell Max-Q Workstation Edition, RTX Pro 5000 Blackwell, RTX Pro 4500 Blackwell, and RTX Pro 4000 Blackwell.
Additionally, the RTX Pro 6000 data center and desktop GPUs, as well as the 5000 series desktop GPUs, feature multi-instance GPU (MIG) technology. This allows a single GPU to be securely partitioned in up to four instances (6000 series) or two instances (5000 series), which is obviously not something that can be accomplished on NVIDIA’s GeForce cards.
For workers on the go, NVIDIA also announced several laptop variants, including the RTX Pro 5000 Blackwell, RTX Pro 4000 Blackwell, RTX Pro 3000 Blackwell, RTX Pro 2000 Blackwell, RTX Pro 1000 Blackwell, and RTX Pro 5000 Blackwell. Memory configurations top out at up to 24GB of GDDR7 (RTX Pro 5000 Blackwell), while all of the laptop SKUs support the latest implementation of Max-Q.
We won’t rehash Blackwell from an architectural overview, but suffice say, everything that makes Blackwell tick (and then some) is included in all of these GPUs. Things like fourth generation RT cores (with up to a 2x performance uplift over the previous generation), fifth generation Tensor cores capable of up to 4,000 AI trillion operations per second and DLSS 4 multi-frame generation, DisplayPort 2.1 support to drive high resolution displays at up to 4K/480Hz and 8K/165Hz, and so forth.
What about availability? NVIDIA says the RTX Pro 6000 Blackwell Server Edition will be available soon in server configs from various data center partners, such as Cisco, Dell, HP, Lenovo, and Supermicro. Additionally, cloud service and GPU cloud providers (AWS, Google Cloud, Microsoft Azure, and CoreWeave) will be the first to offer instances powered by the same GPU sometime later this year. We’ll also see data center platforms tap the 6000 series part from the likes of ASUS, Gigabyte, Ingrasys, Quanta Cloud Technology, and others.
The RTX Pro 6000 Blackwell Workstation Edition and RTX Pro 6000 Blackwell Max-Q Workstation Edition will both show up in April from partners like PNY and TD SYNNEX, followed by wider availability in May from manufacturers like BOXX, Dell, HP, Lambda, and Lenovo.
Moving down the list, the RTX Pro 5000, RTX Pro 4500, and RTX Pro 4000 Blackwell GPUs will launch in the summer time frame from BOXX, Dell, HP, and Lenovo.
Finally, the RTX Pro Blackwell laptop GPUs will be available sometime later this year from Dell, HP, Lenovo, and Razer.