NVIDIA RTX PRO 6000 Blackwell Server vs NVIDIA TITAN Ada

Comparison of NVIDIA RTX PRO 6000 Blackwell Server with 96 GB GDDR7 and 24,064 cores vs NVIDIA TITAN Ada with 48 GB GDDR6X and 18,432 cores.

Loading...

Performance Rating

A100 A100
H200 H200
MI325X MI325X

NVIDIA RTX PRO 6000 Blackwell Server

57.5

NVIDIA RTX PRO 6000 Blackwell Server

57.5
RX 7900 XTX RX 7900 XTX
MI250 MI250
Instinct MI300X Instinct MI300X

NVIDIA TITAN Ada

NVIDIA TITAN Ada

Contents:

Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering Benchmarks Additional

Memory

Memory Size

🔥 +100% 96 ГБ
48 GB

Memory Type

GDDR7 GDDR6X

Memory Bandwidth

🔥 +56% 1.79 TB/s
1.15 TB/s

Memory Bus Width

512 бит 384 бит

ML Performance

FP16 (Half Precision)

🔥 +36% 126.0 TFLOPS
92.9 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +36% 126.0 TFLOPS
92.9 TFLOPS

FP64 (Double Precision)

🔥 +36% 1.968 TFLOPS
1.4515 TFLOPS

CUDA Cores

🔥 +31% 24,064
18,432

RT Cores

🔥 +31% 188
144

Architecture & Compatibility

GPU Architecture

Blackwell 2.0 Ada Lovelace

SM (Streaming Multiprocessor)

🔥 +31% 188
144

PCIe Version

PCIe 5.0 x16 PCIe 4.0 x16

ML Software Support

CUDA Version

🔥 12.0
8.9

Clocks & Performance

Base Clock

🔥 1,590
2,235

Boost Clock

🔥 +4% 2,617
2,520

Memory Clock

🔥 +17% 1,750
1,500

Power Consumption

TDP/TGP

🔥 -25% 600 W
800 W

Recommended PSU

🔥 -17% 1000 W
1200 W

Power Connector

1x 16-pin 2x 16-pin

Rendering

Texture Units (TMU)

🔥 +31% 752
576

ROP

🔥 +31% 188
144

L2 Cache

🔥 +33% 128 MB
96 MB

Benchmarks

MLPerf, llama2-70b-99.9 (fp4)

3 250 tokens/s

MLPerf, llama3.1-8b (fp4)

5 758 tokens/s

Geekbench AI, FP16

53 322 points

Geekbench AI, INT8

28 264 points

Geekbench AI, FP32

37 299 points

MLPerf, mixtral-8x7b (fp8)

3 767 tokens/s

Additional

Slots

Dual-slot Quad-slot

Release Date

March 18, 2025 No

Display Outputs

4x DisplayPort 2.1b
1x HDMI 2.1
3x DisplayPort 1.4a

Renting is cheaper than buying