GeekRadar / Il banco di confronto

CONFRONTA.

GPU

NVIDIA

RTX PRO 4500 Blackwell

Processore grafico

Memoria video: 32 GB^*
Bus di memoria: 256 bit^*
Banda di memoria nominale: 896 GB/s^*

Non sono ancora disponibili benchmark associati a questa scheda.

Elaborazione e piattaforma

Architettura: Blackwell 2.0^*
Nodo produttivo: 5 nm^*
Frequenza base GPU: 1.635 MHz^*
Frequenza boost GPU: 2.407 MHz^*

Memoria e archiviazione

Memoria video: 32 GB^*
Tipo memoria video: GDDR7^*
Bus di memoria: 256 bit^*
Banda di memoria nominale: 896 GB/s^*

Configurazione di sistema

Specifiche dettagliate

Specifiche 1 · 23 campi

Architettura GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

Blackwell^*

Architettura GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

Blackwell^*

Picco teorico: FP32 senza unità tensoriali denso per GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

51^* TFLOPS

Picco teorico: FP32 senza unità tensoriali denso per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

51^* TFLOPS

Picco teorico calcolato: FP4 tensoriale denso per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

808,5^* TFLOPS

Ambito e condizioni

Calcolo: 1617 / 2^*

Picco teorico: FP4 tensoriale con sparsità per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

1.617^* TFLOPS

Configurazione fisica — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

PCIe, full height, dual slot (4.4 in x 10.5 in L)^*

Configurazione fisica — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

PCIe, FHFL, single slot, passive (4.4 in H x 10.5 in L)^*

Supporto hardware nativo per l’interconnessione fra GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

No^*

Interfaccia PCIe dichiarata — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

PCIe Gen 5 x16^*

Interfaccia PCIe dichiarata — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

PCIe Gen 5 x16^*

Potenza massima della scheda — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

200^* W

Potenza massima della scheda — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

165^* W

Banda massima della memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

896^* GB/s

Banda massima della memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

800^* GB/s

Ampiezza dell’interfaccia di memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

256^* bit

Ampiezza dell’interfaccia di memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

256^* bit

Capacità di memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

32^* GB

Capacità di memoria per GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

32^* GB

Tecnologia della memoria — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

GDDR7^*

Tecnologia della memoria — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

GDDR7^*

Modello GPU — NVIDIA RTX PRO 4500 Blackwell Server Edition — 32 GB GDDR7

NVIDIA RTX PRO 4500 Blackwell Server Edition^*

Modello GPU — NVIDIA RTX PRO 4500 Blackwell Workstation Edition — 32 GB GDDR7

NVIDIA RTX PRO 4500 Blackwell Workstation Edition^*

Specifiche 2 · 59 campi

Brand: NVIDIA^*
Top BUS WIDTH: 256^* bit
Graphics Processor Process Type: 4N FinFET^*
Graphics Processor GPU Name: GB203^*
Graphics Processor Architecture: Blackwell 2.0^*
Render Config SM Count: 82^*
Graphics Features CUDA: 12^*
Render Config Tensor Cores: 328^*
Graphics Processor Process Size: 5^* nm
Graphics Processor Transistors: 45,600 million^*
Graphics Processor Die Size: 378^* mm²
Graphics Card Release Date: Mar 18th, 2025^*
Graphics Card Generation: Blackwell PRO W (x000)^*
Graphics Card Production: Active^*
Name: RTX PRO 4500 Blackwell^*
Graphics Card Bus Interface: PCIe 5.0 x16^*
Clock Speeds Memory Clock: 1750 MHz 28 Gbps effective^*
Memory Memory Size: 32^* GB
Memory Memory Type: GDDR7^*
Memory Memory Bus: 256^* bit
Memory Bandwidth: 896^* GB/s
Render Config TMUs: 328^*
Top GRAPHICS PROCESSOR: GB203^*
Render Config ROPs: 112^*
Theoretical Performance Pixel Rate: 269.6 GPixel/s
Theoretical Performance Texture Rate: 789.5 GTexel/s
Board Design Slot Width: Dual-slot^*
Board Design TDP: 200^* W
Board Design Suggested PSU: 550^* W
Board Design Outputs: 4x DisplayPort 2.1b^*
Graphics Features DirectX: 12 Ultimate (12_2)^*
Graphics Features OpenGL: 4,6^*
Graphics Features OpenCL: 3^*
Graphics Features Vulkan: 1,4^*
source file: nvidia2025.json^*
Graphics Processor Foundry: TSMC^*
Graphics Card Predecessor: Workstation Ada^*
Recommended Resolutions: Good:1920x1080 | Good:2560x1440 | Good:3840x2160^*
Board Design Length: 267 mm 10.5 inches^*
Board Design Power Connectors: 1x 16-pin^*
Graphics Processor Density: 120.6M / mm²^*
Board Design Height: 111 mm 4.4 inches^*
Board Design Width: 40 mm 1.6 inches^*
Top TMUS: 328^*
Top ROPS: 112^*
Clock Speeds Base Clock: 1.635^* MHz
Clock Speeds Boost Clock: 2.407^* MHz
Graphics Features Shader Model: 6.8^*
Top CORES: 10.496^*
Render Config Shading Units: 10.496^*
Render Config L2 Cache: 64^* MB
Top MEMORY SIZE: 32^* GB
Theoretical Performance FP32 (float): 50,53^* TFLOPS
Theoretical Performance FP64 (double): 789.5 GFLOPS (1:64)
Render Config L1 Cache: 128 KB (per SM)^*
Graphics Card Announced: Mar 17th, 2025^*
Theoretical Performance FP16 (half): 50.53 TFLOPS (1:1)
Top MEMORY TYPE: GDDR7^*
Render Config RT Cores: 82^*

* Dati provvisori