NVIDIA's cuLitho to Speed Up Computational Lithography for 2nm and Beyondby Anton Shilov on March 27, 2023 1:00 PM EST
Production of chips using leading-edge process technologies requires more compute power than ever. To address requirements of 2nm nodes and beyond, NVIDIA is rolling out its cuLitho software library that uses the company's DGX H100 systems based on H100 GPUs and promises to increase performance available to mask shops within a reasonable amount of consumed power by 40 times.
Modern process technologies push wafer fab equipment to its limits and often require finer resolution than is physically possible, which is where computational lithography comes into play. The primary purpose of computational lithography is to enhance the achievable resolution in photolithography processes without modifying the tools. To do so, CL employs algorithms that simulate the production process, incorporating crucial data from ASML's equipment and shuttle (test) wafers. These simulations aid in refining the reticle (photomask) by deliberately altering the patterns to counteract the physical and chemical influences that arise throughout the lithography and patterning steps.
There are several computational lithography techniques, including Resolution Enhancement Technology (RET), Inverse Lithography Technology (ILT, a method to reduce manufacturing variations by utilizing non-rectangular shapes on the photomask), Optical Proximity Correction (OPC, a technique for improving photolithography by correcting image inaccuracies resulting from diffraction or process-related impacts), and Source Mask Optimization (SMO). All of them are widely used at today's fabs.
Meanwhile, compute-expensive techniques like inverse lithography technology and source mask optimization are specific to a given design. They have to be implemented individually for each chip to ensure appropriate resolution and avoid yield-limiting hotspots. Synthesis of reticles that use RET, ILT, OPC, and SMO involves the usage of computational lithography. As nodes get thinner, the complexity of computations increases, and compute horsepower becomes a bottleneck for mask shops as each modern chip uses dozens of reticles. For example, NVIDIA's H100 uses 89 of them.
NVIDIA says that computational lithography currently consumes tens of billions of CPU hours every year and, therefore, enormous power. Meanwhile, highly parallel GPUs like NVIDIA's H100 promise higher performance at lower cost and power consumption. In particular, NVIDIA says that 500 of its DGX H100 systems packing 4000 of its H100 GPUs (that consume 5 MW of power) and using computational lithography software that uses cuLitho can offer the performance of 40,000 CPU-based systems which consume 35 MW that TSMC uses today. The company also goes on to say that mask makers can produce 3 – 5 times more reticles per day using nine times less power than they use today once they start relying on GPU-accelerated computational lithography, another claim that requires verification by actual mask shops, but which gives a basic understanding where the company wants to go.
"With lithography at the limits of physics, NVIDIA's introduction of cuLitho and collaboration with our partners TSMC, ASML, and Synopsys allows fabs to increase throughput, reduce their carbon footprint and set the foundation for 2nm and beyond."
While performance targets set by NVIDIA are impressive, it should be noted that the cuLitho software library for computational lithography must be incorporated in software offered by ASML, Synopsys, and TSMC well used by their partners, among mask shops. For current-generation lithography (think 7 nm, 5 nm, and 3 nm-class nodes), mask shops already use CPU-based computational lithography solutions and will continue to do so for at least a while. This is perhaps why NVIDIA is discussing its computational lithography efforts in context with next-generation 2 nm-class nodes and beyond. Meanwhile, it makes sense to expect foundries and mask shops to at least try deploying NVIDIA's cuLitho for some of their upcoming 3 nm-class nodes to increase yields and performance. TSMC, for example, will start to qualify cuLitho in mid-2023, so expect the platform to be available to the company's customers beginning in 2024.
"Computational lithography, specifically optical proximity correction, or OPC, is pushing the boundaries of compute workloads for the most advanced chips," said Aart de Geus, chief executive of Synopsys. "By collaborating with our partner NVIDIA to run Synopsys OPC software on the cuLitho platform, we massively accelerated the performance from weeks to days! The team-up of our two leading companies continues to force amazing advances in the industry."
An official statement by NVIDIA states that "A fab process change often requires an OPC revision, creating bottlenecks." "cuLitho not only helps remove these bottlenecks, but it also makes possible novel solutions and innovative techniques like curvilinear masks, high NA EUV lithography, and sub-atomic photoresist modeling needed for new technology nodes."
Extra compute horsepower available for computational lithography applications will come in particularly handy for the next generation of production nodes that will use High-NA lithography scanners and will mandate the usage of ILT, OPC, and SMO to consider physical peculiarities of lithography scanners and resists to ensure decent yields, low variation (i.e., foreseeable performance and power consumption), and predictable costs. Meanwhile, computational costs for RET, ILT, OPC, and SMO will inevitably increase at 2 nm and beyond, so it looks like NVIDIA will introduce its cuLitho platform at a good time.
Post Your CommentPlease log in or sign up to comment.
View All Comments
James5mith - Monday, March 27, 2023 - linkDoes any of this help them keep chips consuming less than 1kW of power? Because it feels like that is where we are headed. Efficiency be damned, ramp up power for performance. Seems like too long ago Intel promised no feature would be included if it didn't offer 2%+ performance gain for less than 1% power increase. Reply
Eliadbu - Monday, March 27, 2023 - linkWhile we will still see improvement in perf/watt it will be lower than we've used to. It already slower than past decade that is why power consumption went up - to allow better generational improvements that would not other wise would be possible with current nodes. I fear that the day that will see no improvement in performance/watt with new node is not that far away, or it will be delayed by slower release of new node.
We will need to find replacement for CMOS technology. Reply
Diogene7 - Tuesday, March 28, 2023 - linkAccording to some technology benchmarks done by Intel, Intel seems to do research in spintronics related technology (they used the term MESO) as a technology successor to CMOS.
As of 2023, as there is already some investments done in MRAM manufacturing and a lot of ongoing research to develop manufacturing tools (like at Belgian research institute IMEC), at this point in time (March 2023), I would think that ithrre is a growing ressources allocation momentum toward spintronics, which seems to give this technology a good probability to develop further (cost going down with scale…) Reply
brucethemoose - Monday, March 27, 2023 - linkPart of that is a business choice. Clock targets *could* be lower, but most vendors seemingly figured that customers don't really care about power consumption if it makes the end products faster, cheaper to purchase and less supply constrained.
Apple's A/M series goes against that trend though. I think they really are targeting "minimum task energy." Reply
Oxford Guy - Monday, March 27, 2023 - linkI am typing this on an 'M1 Pro' Macbook Pro. It's 'overbuilt' for the chip. Instead of going for a super thin and light form factor, Apple gave this machine a thick metal case (both in the thickness of the profile but also in the solidity of the metal).
The benefit of that is peace and quiet. Personally, I'd much rather have a laptop that's a bit heavier and thicker in order to not hear fans. Reply
mode_13h - Tuesday, March 28, 2023 - linkWhat does that have to do with anything in the posts you just replied to? Seems like nothing more than an advert, to me. Reply
PeachNCream - Tuesday, March 28, 2023 - linkMainly it was written to troll you in specific. Reply
Skeptical123 - Tuesday, March 28, 2023 - linkso it's your alt account... Reply
mode_13h - Thursday, March 30, 2023 - linkI'm sure Peaches has better things to do than troll me. Reply
mode_13h - Thursday, March 30, 2023 - linkI don't think so, because we've hardly ever had any exchanges about Apple products. There would be better ways to troll me, if that were his objective.
I think he's probably feeling chuffed about his fancy new Mac and struggling to refrain from blabbing about it at every opportunity. Honestly, I'm happy for him if he likes it. I just don't appreciate the digression. Reply