Falcon-180B on a single H200 GPU with INT4 AWQ, and 6.7x faster Llama-70B over A100

H200’s large capacity & high memory bandwidth, paired with TensorRT-LLM’s optimizations, maximizes inference performance.

Falcon-180B on a single H200 with INT4 AWQ

Falcon-180B, one of the largest & most accurate open source models available, can run on a single H200 GPU.

The 141GB of memory on H200, paired with TensorRT-LLM running INT4 AWQ with FP8, allows for the entire large language model to fit on a single GPU, where previously eight A100s were required. H200 Falcon-180B provides up to 800 tok/s and retains high accuracy.

Model Performance: H200’s large capacity & high memory bandwidth, utilizing INT4 AWQ to reduce memory footprint, allows for great performance on Falcon-180B on a single GPU.

Falcon-180B performance comparison

Preliminary measured Performance, subject to change. TP1 does not represent peak performance on H200. TensorRT-LLM v0.7a | Falcon-180B | 1xH200 TP1 | INT4 AWQ | BS: (in order) 256, 128

Model Accuracy: Often quantization can have adverse impacts on the accuracy of the model, however, TensorRT-LLM’s AWQ decreases memory footprint of the model by 4x while maintaining high accuracy.

Falcon-180B accuracy comparison

Preliminary measured accuracy, subject to change. TensorRT-LLM v0.7a | Falcon-180B | 1xH200 TP1 | INT4 AWQ

INT4 Activation-aware Weight Quantization (AWQ) (Lin et al., 2023) is a quantization technique which compresses the weights of an LLM down to 4bits based on their relative importance, and performs computation in FP16. This allows for AWQ to retain higher accuracy than other 4bit methods and reduce memory usage, but requires special kernels capable of handling the change in precision performantly.

TensorRT-LLM has implemented custom kernels for AWQ, and taken the technique a step further by performing FP8 computation on Hopper GPUs instead of the standard FP16.

Similar examples running Falcon-180B with quantization in TensorRT-LLM are available in examples/falcon.

Llama-70B on H200 up to 6.7x A100

TensorRT-LLM has improved its Group Query Attention (GQA) kernels, in the generation phase, providing up to 2.4x improvement on Llama-70B over TensorRT-LLM v0.5, achieving over 3,800 tok/s/gpu at up to 6.7x faster than A100.

H200 6.7x A100

Llama-70B H200 vs A100 comparison

Model

GPUs

Input Length

Output Length

Throughput (out tok/s/GPU)

Llama-70B

1

128

128

3,803

8

3,803

1

2048

2,941

8

3,163

1

4096

1,946

8

2,263

Preliminary measured performance, subject to change. TensorRT-LLM v0.7a | Llama2-70B | 1xH200 = TP1, 8xH200 = max TP/PP/DP config | FP8 | BS: (in order) 960, 960, 192, 560, 96, 640

TensorRT-LLM GQA now 2.4x faster on H200

Llama-70B H200 December vs Oct.

Preliminary measured performance, subject to change. TensorRT-LLM v0.7a vs TensorRT-LLM v0.6a | Llama2-70B | 1xH200 TP1 | FP8 | BS 192

Grouped Query Attention (GQA) (Ainslie et al., 2023), used in Llama-70B, is a variant of Multihead Attention (MHA) which groups key-value (KV) heads together, resulting in fewer KV heads than query (Q) heads. TensorRT-LLM has a custom implementation of MHA which supports GQA, multi-query attention (MQA) and standard MHA. It leverages Tensor Cores, including in the generation phase, and delivers great performance on NVIDIA GPUs.

Closing

These improvements will be published in the main branch soon, and will be included in the v0.7 & v0.8 releases.

Similar examples running Llama-70B in TensorRT-LLM are published in examples/llama.

For more information about H200, please see the H200 announcement blog.

Throughput is calculated as output tokens per second per gpu. out_tps=output_seqlen*batch_size/total_latency/tp

Glossary: | DP = Data Parallel ISL = Input Sequence Length | PP = Pipeline Parallel | OSL = Output Sequence Length | OOM = Out of Memory | TP = Tensor Parallel