Editor's note: This post is part of the AI Decoded series, which demystifies AI by making the technology more accessible, and showcases new hardware, software, tools and accelerations for RTX PC users.
Large language models are driving some of the most exciting developments in AI with their ability to quickly understand, summarize and generate text-based content.
These capabilities power a variety of use cases, including productivity tools, digital assistants, non-playable characters in video games and more. But they're not a one-size-fits-all solution, and developers often must fine-tune LLMs to fit the needs of their applications.
The NVIDIA RTX AI Toolkit makes it easy to fine-tune and deploy AI models on RTX AI PCs and workstations through a technique called low-rank adaptation, or LoRA. A new update, available today, enables support for using multiple LoRA adapters simultaneously within the NVIDIA TensorRT-LLM AI acceleration library, improving the performance of fine-tuned models by up to 6x.
Fine-Tuned for Performance LLMs must be carefully customized to achieve higher performance and meet growing user demands.
These foundational models are trained on huge amounts of data but often lack the context needed for a developer's specific use case. For example, a generic LLM can generate video game dialogue, but it will likely miss the nuance and subtlety needed to write in the style of a woodland elf with a dark past and a barely concealed disdain for authority.
To achieve more tailored outputs, developers can fine-tune the model with information related to the app's use case.
Take the example of developing an app to generate in-game dialogue using an LLM. The process of fine-tuning starts with using the weights of a pretrained model, such as information on what a character may say in the game. To get the dialogue in the right style, a developer can tune the model on a smaller dataset of examples, such as dialogue written in a more spooky or villainous tone.
In some cases, developers may want to run all of these different fine-tuning processes simultaneously. For example, they may want to generate marketing copy written in different voices for various content channels. At the same time, they may want to summarize a document and make stylistic suggestions - as well as draft a video game scene description and imagery prompt for a text-to-image generator.
It's not practical to run multiple models simultaneously, as they won't all fit in GPU memory at the same time. Even if they did, their inference time would be impacted by memory bandwidth - how fast data can be read from memory into GPUs.
Lo(RA) and Behold A popular way to address these issues is to use fine-tuning techniques such as low-rank adaptation. A simple way of thinking of it is as a patch file containing the customizations from the fine-tuning process.
Once trained, customized LoRA adapters can integrate seamlessly with the foundation model during inference, adding minimal overhead. Developers can attach the adapters to a single model to serve multiple use cases. This keeps the memory footprint low while still providing the additional details needed for each specific use case.
Architecture overview of supporting multiple clients and use-cases with a single foundation model using multi-LoRA capabilities In practice, this means that an app can keep just one copy of the base model in memory, alongside many customizations using multiple LoRA adapters.
This process is called multi-LoRA serving. When multiple calls are made to the model, the GPU can process all of the calls in parallel, maximizing the use of its Tensor Cores and minimizing the demands of memory and bandwidth so developers can efficiently use AI models in their workflows. Fine-tuned models using multi-LoRA adapters perform up to 6x faster.
LLM inference performance on GeForce RTX 4090 Desktop GPU for Llama 3B int4 with LoRA adapters applied at runtime. Input sequence length is 43 tokens and output sequence length is 100 tokens. LoRA adapter max rank is 64. In the example of the in-game dialogue application described earlier, the app's scope could be expanded, using multi-LoRA serving, to generate both story elements and illustrations - driven by a single prompt.
The user could input a basic story idea, and the LLM would flesh out the concept, expanding on the idea to provide a detailed foundation. The application could then use the same model, enhanced with two distinct LoRA adapters, to refine the story and generate corresponding imagery. One LoRA adapter generates a Stable Diffusion prompt to create visuals using a locally deployed Stable Diffusion XL model. Meanwhile, the other LoRA adapter, fine-tuned for story writing, could craft a well-structured and engaging narrative.
In this case, the same model is used for both inference passes, ensuring that the space required for the process doesn't significantly increase. The second pass, which involves both text and image generation, is performed using batched inference, making the process exceptionally fast and efficient on NVIDIA GPUs. This allows users to rapidly iterate through different versions of their stories, refining the narrative and the illustrations with ease.
This process is outlined in more detail in a recent technical blog.
LLMs are becoming one of the most important components of modern AI. As adoption and integration grows, demand for powerful, fast LLMs with application-specific customizations will only increase. The multi-LoRA support added today to the RTX AI Toolkit gives developers a powerful new way to accelerate these capabilities.
Most recent headlines
05/01/2027
Worlds first 802.15.4ab-UWB chip verified by Calterah and Rohde & Schwarz to be ...
14/11/2026
Longtime MLB Network employee David Maybrown passed away Thursday at the age of 57. David joined MLB Network in 2009 shortly after its launch, beginning as an i...
14/11/2026
Modular innovators introduce two guitar-based products
ADDAC System's recently introduced Four String Series makes it possible to integrate physical str...
14/11/2026
Marcelo Dai's Live Session Captured with URSA Cine Immersive
Brie Clayton November 14, 2026
0 Comments
Artist takes viewers on an immersive journe...
14/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
14/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
14/11/2026
Newest Lenses, Horizon Anamorphic Seminar & Guided Lens Character Experience Highlight Creative Possibilities
At the 2026 Toru CAMERIMAGE Film Festival, ZEISS...
14/11/2026
Sonnet Announces Black Friday Sale on Select Products
Brie Clayton November 14, 2026
0 Comments
WHAT: Sonnet Technologies is having a sale on a select...
14/11/2026
Berklee Alumni Shine at the 2026 Latin Grammy Awards This year's winners included songwriters, engineers, and performers who contributed to acclaimed reco...
13/11/2026
The UIM E1 Series wraps its third season this week with the Bombay Sapphire E1 Miami GP, Nov. 11-14, on Biscayne Bay at the Miami Seaplane Base on Watson Island...
13/11/2026
In-venue and creative video staffers at the professional and collegiate level ha...
13/11/2026
Singer-songwriter Benson Boone will headline entertainment for the Discover NHL Winter Classic between the Utah Mammoth and Colorado Avalanche at Rice-Eccles St...
13/11/2026
Blackmagic Design has announced the Blackmagic Ambisonic Microphone, a 7th order ambisonic microphone array that captures spatial audio for immersive video prod...
13/11/2026
SMPTE has announced that The Studio B&H and Riedel Communications will provide the video infrastructure and communications technology for the Solutions Hub at t...
13/11/2026
The Hollywood Professional Association (HPA) has announced the theme and program focus for the 2027 HPA Tech Retreat, which takes place Feb. 14-18, 2027, in La ...
13/11/2026
Football Manager 27 Touch is available on Apple Arcade starting today. The game ...
13/11/2026
MotoGP has announced a two-year media rights agreement with ESPN covering the Un...
13/11/2026
Apple Sports, the free iPhone app, has added the Ultimate Fighting Championship (UFC) to its lineup of supported sports, in time for Saturday's UFC 334. The...
13/11/2026
The NBA and Fanatics Studios, the joint venture between Fanatics and entertainme...
13/11/2026
ESPN held its annual EDGE Innovation Conference last night and offered up an ove...
13/11/2026
Regional sports network gives Ithaca College, SUNY Cortland students a platform to thrive
One of the biggest games in Division III college football will be pla...
13/11/2026
Studer-inspired 500-series PWM compressor unveiled
The HLabs range already includes a couple of Studer-inspired offerings in the form of the PRE169 and EQ16...
13/11/2026
Download Now!
To view the white paper, please enter the following details:
--
First Name* First Name
Company*
Country* Please select your countryA...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Blue Lucy, the workflow orchestration and integration platform for broadcasters, rights owners and localisation companies, today announced an integration with c...
13/11/2026
Krotos announces Black Friday offers on Studio Pro and Studio Max
Brie Clayton November 12, 2026
0 Comments
First-year annual plans start at US$195.99...
13/11/2026
Blackmagic Design Announces Blackmagic Ambisonic Microphone
Brie Clayton November 12, 2026
0 Comments
New 7th order ambisonic microphone with 108 MEMS...
13/11/2026
Meet the Berklee Student Setting Planetary Motion to Music With his compositional platform SoniKosmos, Javier Roqueni aims to make the universe the ultimate i...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , er...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , er...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Share
{alert(Link copied)}).catch(err => {console.error(Failed to copy: , err)}) data-platform=copy_link data-action= class=font-ui-heading py-2 text-sm tex...
13/11/2026
Amagi, the media industry cloud company, today announced a partnership with Golden Times Media Technology (stock code 920021.BJ), a Beijing-based media and tech...
13/11/2026
CGI (TSX: GIB.A) (NYSE: GIB), one of the largest independent IT and business consulting services firms in the world, together with its Malaysian partner, Cedar ...
13/11/2026
SMPTE, the home for media professionals, technologists and engineers, today announced that The Studio B&H and Riedel Communications will provide the video infra...
13/11/2026
iWedia, a global leader in connected TV software solutions, will present its software-first approach to DRAM optimization at Android TV APAC Summit 2026 in Bang...
13/11/2026
Beginning today, November 13, Apple HomeView is available for purchase at Apple Store locations, on the Apple Store online, and in the Apple Store app in select...
13/11/2026
From pediatric research to smart hospitals, AI is helping care teams spend less time on routine tasks and more time with patients.
Health systems and startups...
12/11/2026
Industry leaders examine how Dynamic Media Facility, Media eXchange Layer, and c...
12/11/2026
The junior Tar Heel is embracing the technical and creative challenges of live production while emerging as a leader and mentor in the control room...
12/11/2026
Gracenote, Nielsen's content intelligence business, has released its fifth a...
Editor's note: This post is part of the AI Decoded series, which demystifies AI by making the technology more accessible, and showcases new hardware, software, tools and accelerations for RTX PC users.
