內容目錄
ToggleAI工具演化超級快,工具眾多競爭激烈,對於電腦的負擔也越來越大,小編今天會從不同領域的AI應用來分析AI顯卡推薦排行,同時提供各種顯示卡實驗數據對比表格。
AI顯示卡|品牌差異
NVIDIA 輝達 AI顯卡
AI領域上,英偉達顯示卡憑藉CUDA技術,在處理速度和效率上是遠遠超過AMD的。 以AI繪畫工具DeepArt和AI語言學習軟體Descript為例,它們大量依賴CUDA核心,提供無與倫比的運算能力。 在專業軟體,如3D建模軟體犀牛等方面,英偉達顯示卡展現出更強的渲染能力。 在處理大型的場景渲染時,英偉達顯示卡能顯著減少渲染時間,提高工作效率。
CUDA技術是什麼?
CUDA,全稱為Compute Unified Device Architecture(統一計算架構),是由NVIDIA開發的一種平行運算平台和編程模型。它允許開發者使用NVIDIA的GPU來進行高效能的通用計算。簡單來說,CUDA技術讓原本主要用於圖形處理的GPU,也能夠處理其他複雜的計算任務,如科學計算和深度學習等。
CUDA提供了一套擴展的C語言編程接口,讓開發者能夠在GPU上實現平行處理。這意味著可以將數據密集型的計算任務分配到數千個平行處理核心上,從而大幅提升計算速度和效率。
此外,CUDA技術還支援多種編程語言和API,包括C++、Fortran、Python等,使其應用範圍更廣,並且能夠更容易地整合到各種不同的軟體和應用中。這使得CUDA成為了一個強大的工具,特別是在需要大量計算資源的領域中。
AMD 超威 AI顯卡
AMD顯示卡以其高性價比著稱,同價位的AMD顯示卡常提供比英偉達更好的性能表現。 在控制能耗方面,AMD顯示卡也是獨佔優勢的,相較於競品顯示卡能更有效的可控制能耗。 AMD也不斷推出創新技術,如RDNA架構,提升了其顯示卡的新能和效率,同時保持了競爭力的定價。
RDNA架構是什麼?
AMD的RDNA架構是一種高效能GPU微架構,專為提升遊戲和計算應用的效能而設計。RDNA代表Radeon DNA,它代表了AMD在圖形處理技術上的一個重大進步。
小編整理了此架構的四個特點,可以更清楚知道差異:
1.高效能遊戲體驗:RDNA架構旨在提供高效能的遊戲體驗,包括更快的處理速度和更高的畫面更新率。
2.光線追蹤技術:從RDNA(二代)開始,架構支援光線追蹤,這是一種能夠產生逼真光影效果的技術。
3.能效改善:RDNA架構在每瓦效能上有所提升,意味著相同的能量消耗可以達到更高的性能。
4.AI加速器:RDNA(三代)架構引入了AI加速器,進一步提升了處理效能,特別是在機器學習和人工智能應用中。
AI顯示卡|綜合排行表格 數據比較
- Inference Score: 這個分數反映了顯卡在推論階段的性能。推論是指使用已經訓練好的模型來進行數據分析和預測的過程。例如,在圖像識別、語音識別或其他AI應用中,模型會根據學習到的資料來處理新的輸入數據並給出結果。一個高的Inference Score意味著顯卡在這些任務上能夠更快地處理數據,並且更有效率。
- Training Score: 而訓練分數則是指顯卡在訓練階段的性能。訓練是機器學習模型學習和改進的過程,通過大量數據來調整模型參數,使其能夠準確地進行預測。這個過程通常需要大量的計算資源和時間。一個高的Training Score表示顯卡在進行模型訓練時更加強大和高效。
👇 下方表格可以往右滑,看更多數據比較
wdt_ID | wdt_created_by | wdt_created_at | wdt_last_edited_by | wdt_last_edited_at | 顯卡型號 | AI綜合評分 | Inference | Training | TF Version | 核心 | 頻率GHz | GPU加速 | 平台 | RAM GB |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Tesla V100 SXM2 32Gb | 35791 | 17761 | 18030 | 2.1.0 | 5120 (CUDA) | 1.29 / 1.53 | CUDA 10.1 | Debian 10 | 32 | ||||
2 | Tesla V100 PCIE 32Gb | 34394 | 16530 | 17865 | 2.1.0 | 5120 (CUDA) | 1.23 / 1.38 | CUDA 10.1 | Debian 10 | 32 | ||||
3 | Tesla V100 PCIE 16Gb | 34347 | 16511 | 17837 | 2.1.0 | 5120 (CUDA) | 1.25 / 1.38 | CUDA 10.1 | Red Hat 7.5 | 16 | ||||
4 | Tesla V100 SXM2 16Gb | 35086 | 17251 | 17836 | 2.1.0 | 5120 (CUDA) | 1.31 / 1.53 | CUDA 10.1 | Red Hat 7.5 | 16 | ||||
5 | NVIDIA TITAN RTX | 33339 | 16084 | 17255 | 2.1.0 | 4608 (CUDA) | 1.35 / 1.77 | CUDA 10.1 | Ubuntu 18.04 | 24 | ||||
6 | NVIDIA TITAN V | 33406 | 16192 | 17215 | 2.1.0 | 5120 (CUDA) | 1.20 / 1.46 | CUDA 10.1 | Ubuntu 18.04 | 12 | ||||
7 | NVIDIA Quadro GV100 | 33880 | 16748 | 17132 | 1.14.0 | 5120 (CUDA) | 1.13 / 1.63 | CUDA 10 | Debian 10 | 32 | ||||
8 | GeForce RTX 2080 Ti | 32870 | 16042 | 16828 | 2.1.0 | 4352 (CUDA) | 1.35 / 1.55 | CUDA 10 | Debian 10 | 11 | ||||
9 | NVIDIA Quadro RTX 8000 | 27651 | 13014 | 14637 | 2.1.0 | 4608 (CUDA) | 1.40 / 1.77 | CUDA 10.1 | Debian 10 | 48 | ||||
10 | NVIDIA Quadro GP100 | 25700 | 12264 | 13436 | 2.0.0 | 3584 (CUDA) | 1.30 / 1.44 | CUDA 10 | Red Hat 7.4 | 16 | ||||
11 | NVIDIA TITAN Xp | 24870 | 11948 | 12922 | 2.1.0 | 3840 (CUDA) | 1.41 / 1.58 | CUDA 10.2 | Debian 10 | 12 | ||||
12 | GeForce RTX 2080 SUPER | 24247 | 11513 | 12734 | 2.1.0 | 3072 (CUDA) | 1.65 / 1.82 | CUDA 10.1 | Windows 10 | 8 | ||||
13 | GeForce RTX 2070 SUPER | 24182 | 11472 | 12710 | 2.1.0 | 2560 (CUDA) | 1.61 / 1.77 | CUDA 10.2 | Ubuntu 18.04 | 8 | ||||
14 | GeForce RTX 2080 | 23827 | 11242 | 12586 | 2.1.0 | 2944 (CUDA) | 1.52 / 1.71 | CUDA 10.2 | Ubuntu 18.04 | 8 | ||||
15 | GeForce GTX 1080 Ti | 24386 | 11914 | 12473 | 2.1.0 | 3584 (CUDA) | 1.58 / 1.60 | CUDA 10.2 | Debian 10 | 11 | ||||
16 | NVIDIA Tesla P100 | 22466 | 10662 | 11804 | 2.1.0 | 3584 (CUDA) | 1.19 / 1.33 | CUDA 10 | Ubuntu 18.04 | 16 | ||||
17 | NVIDIA Quadro RTX 5000 | 22366 | 10721 | 11646 | 2.0.0 | 3072 (CUDA) | 1.62 / 1.82 | CUDA 10.1 | Windows 10 | 16 | ||||
18 | GeForce RTX 2070 | 20723 | 9726 | 10997 | 2.1.0 | 2304 (CUDA) | 1.41 / 1.62 | CUDA 10.1 | Debian 10 | 8 | ||||
19 | NVIDIA Quadro P6000 | 21674 | 10927 | 10747 | 1.14.0 | 3840 (CUDA) | 1.51 / 1.65 | CUDA 10 | Debian 10 | 24 | ||||
20 | NVIDIA TITAN X Pascal | 20089 | 9714 | 10375 | 2.1.0 | 3584 (CUDA) | 1.42 / 1.53 | CUDA 10.1 | Windows 10 | 12 | ||||
21 | GeForce RTX 2060 SUPER | 19944 | 9599 | 10345 | 2.1.0 | 2176 (CUDA) | 1.47 / 1.65 | CUDA 10 | Ubuntu 18.04 | 8 | ||||
22 | NVIDIA Quadro RTX 4000 | 19066 | 9003 | 10063 | 2.0.0 | 2304 (CUDA) | 1.01 / 1.55 | CUDA 10.1 | Windows 10 | 8 | ||||
23 | NVIDIA TITAN Xp CE | 20103 | 10663 | 9440 | 1.12.0 | 3840 (CUDA) | 1.41 / 1.58 | CUDA 9 | Debian 9.9 | 12 | ||||
24 | GeForce RTX 2060 | 18392 | 8973 | 9419 | 2.1.0 | 1920 (CUDA) | 1.37 / 1.68 | CUDA 10.1 | Debian 10 | 6 | ||||
25 | GeForce RTX 2070 SUPER Laptop | 17412 | 8133 | 9278 | 2.1.0 | 2560 (CUDA) | 1.14 / 1.38 | CUDA 10.2 | Windows 10 | 8 | ||||
26 | GeForce RTX 2080 Max-Q | 17210 | 8240 | 8970 | 2.1.0 | 2944 (CUDA) | 0.76 / 1.10 | CUDA 10.2 | Windows 10 | 8 | ||||
27 | GeForce GTX 1080 | 17383 | 8439 | 8944 | 2.1.0 | 2560 (CUDA) | 1.61 / 1.73 | CUDA 10 | Ubuntu 18.04 | 8 | ||||
28 | GeForce GTX 1080 Laptop | 17134 | 8321 | 8813 | 2.1.0 | 2560 (CUDA) | 1.56 / 1.77 | CUDA 10 | Debian 9.9 | 8 | ||||
29 | NVIDIA Tesla P40 | 18006 | 9221 | 8785 | 1.14.0 | 3840 (CUDA) | 1.30 / 1.53 | CUDA 10 | Ubuntu 18.04 | 24 | ||||
30 | GeForce RTX 2070 Max-Q | 16479 | 7899 | 8580 | 2.1.0 | 2304 (CUDA) | 0.89 / 1.19 | CUDA 10.2 | Windows 10 | 8 | ||||
31 | NVIDIA Quadro RTX 5000 Max-Q | 16243 | 7761 | 8482 | 2.1.0 | 3072 (CUDA) | 0.60 / 1.35 | CUDA 10 | Windows 10 | 16 | ||||
32 | GeForce GTX 1070 Ti | 15865 | 7555 | 8310 | 2.1.0 | 2432 (CUDA) | 1.61 / 1.68 | CUDA 10.1 | Ubuntu 18.04 | 8 | ||||
33 | NVIDIA Quadro P5000 | 16238 | 7937 | 8301 | 2.0.0 | 2560 (CUDA) | 1.61 / 1.73 | CUDA 10.1 | Windows 10 | 16 | ||||
34 | GeForce GTX 980 Ti | 16038 | 7782 | 8256 | 2.1.0 | 2816 (CUDA) | 1.00 / 1.08 | CUDA 10.2 | CentOS 7.7 | 6 | ||||
35 | GeForce GTX TITAN X | 15351 | 7443 | 7908 | 2.1.0 | 3072 (CUDA) | 1.00 / 1.09 | CUDA 10 | Debian 10 | 12 | ||||
36 | GeForce RTX 2070 Laptop | 14925 | 7058 | 7867 | 2.1.0 | 2304 (CUDA) | 1.22 / 1.44 | CUDA 10.1 | Windows 10 | 8 | ||||
37 | GeForce RTX 2060 Laptop | 14924 | 7151 | 7774 | 2.1.0 | 1920 (CUDA) | 0.96 / 1.20 | CUDA 10.2 | Windows 10 | 6 | ||||
38 | GeForce GTX 1070 | 14830 | 7084 | 7745 | 2.1.0 | 1920 (CUDA) | 1.51 / 1.68 | CUDA 10.1 | Ubuntu 18.04 | 8 | ||||
39 | NVIDIA Tesla T4 | 14558 | 6923 | 7635 | 2.1.0 | 2560 (CUDA) | 0.59 / 1.59 | CUDA 10 | Ubuntu 18.04 | 16 | ||||
40 | GeForce GTX 1660 Ti Laptop | 14183 | 6713 | 7470 | 2.0.0 | 1536 (CUDA) | 1.45 / 1.59 | CUDA 10 | Ubuntu 18.04 | 6 | ||||
41 | GeForce GTX 1070 Max-Q | 13764 | 6491 | 7273 | 2.1.0 | 2048 (CUDA) | 1.22 / 1.38 | CUDA 10.1 | Debian 10 | 8 | ||||
42 | AMD Radeon VII | 19367 | 12173 | 7194 | 2.1.0 | 3840 | 1.40 / 1.75 | ROCm (OpenCL) | Debian 10 | 16 | ||||
43 | NVIDIA Quadro P4000 | 12924 | 6086 | 6837 | 2.0.0 | 1792 (CUDA) | 1.20 / 1.48 | CUDA 10.1 | CentOS 7.7 | 8 | ||||
44 | GeForce GTX 1080 Max-Q | 12175 | 5910 | 6265 | 2.0.0 | 2560 (CUDA) | 1.29 / 1.49 | CUDA 10.1 | Windows 10 | 8 | ||||
45 | GeForce GTX 1060 | 11399 | 5444 | 5955 | 2.1.0 | 1280 (CUDA) | 1.51 / 1.71 | CUDA 10.1 | Debian 10 | 6 | ||||
46 | NVIDIA Tesla P4 | 11083 | 5371 | 5712 | 2.1.0 | 2560 (CUDA) | 0.89 / 1.11 | CUDA 10 | Ubuntu 18.04 | 8 | ||||
47 | GeForce GTX 1660 SUPER | 12017 | 6327 | 5690 | 2.1.0 | 1408 (CUDA) | 1.53 / 1.75 | CUDA 10 | Windows 10 | 6 | ||||
48 | GeForce GTX 1060 Laptop | 10312 | 4887 | 5425 | 2.0.0 | 1280 (CUDA) | 1.40 / 1.67 | CUDA 10 | Debian 9.9 | 6 | ||||
49 | GeForce GTX 970 | 10228 | 4831 | 5398 | 2.1.0 | 1664 (CUDA) | 1.05 / 1.18 | CUDA 10 | Ubuntu 18.04 | 4 | ||||
50 | NVIDIA Quadro P3200 | 10112 | 4784 | 5328 | 2.0.0 | 1792 (CUDA) | 1.33 / 1.54 | CUDA 10 | Windows 10 | 6 | ||||
51 | GeForce GTX 1060 Max-Q | 9721 | 4637 | 5084 | 2.1.0 | 1280 (CUDA) | 1.06 / 1.48 | CUDA 10.1 | Windows 10 | 6 | ||||
52 | NVIDIA Tesla M60 | 9644 | 4685 | 4959 | 2.1.0 | 2048 x 2 (CUDA) | 0.56 / 1.18 | CUDA 10.1 | Windows 10 | 8 x 2 | ||||
53 | GeForce GTX TITAN BLACK | 8950 | 4001 | 4949 | 2.0.0 | 2880 (CUDA) | 0.89 / 0.98 | CUDA 10.1 | Windows 10 | 6 | ||||
54 | GeForce GTX 980M | 8423 | 3938 | 4485 | 1.12.0 | 1536 (CUDA) | 1.04 / 1.13 | CUDA 9 | Ubuntu 16.04 | 8 | ||||
55 | NVIDIA Quadro P2000 | 8145 | 3831 | 4315 | 2.1.0 | 1024 (CUDA) | 1.08 / 1.48 | CUDA 10.1 | Windows 10 | 5 | ||||
56 | NVIDIA Quadro M4000 | 7196 | 3416 | 3780 | 2.0.0 | 1664 (CUDA) | 0.77 | CUDA 10 | Ubuntu 18.04 | 8 | ||||
57 | GeForce GTX 1050 Ti | 7010 | 3231 | 3780 | 1.15.0 | 768 (CUDA) | 1.29 / 1.39 | CUDA 10 | Ubuntu 18.04 | 4 | ||||
58 | GeForce GTX 1050 Ti Laptop | 6924 | 3234 | 3690 | 2.1.0 | 768 (CUDA) | 1.49 / 1.62 | CUDA 10.1 | Ubuntu 18.04 | 4 | ||||
59 | NVIDIA Tesla K80 | 6582 | 2982 | 3600 | 2.1.0 | 2496 x 2 (CUDA) | 0.56 / 0.82 | CUDA 10.1 | Ubuntu 18.04 | 12 x 2 | ||||
60 | NVIDIA Tesla K40c | 6464 | 2896 | 3568 | 2.0.0 | 2880 (CUDA) | 0.75 / 0.88 | CUDA 10 | Debian 9.9 | 12 | ||||
61 | GeForce GTX 960 | 6308 | 2996 | 3312 | 2.1.0 | 1024 (CUDA) | 1.13 / 1.18 | CUDA 10 | Debian 10 | 4 | ||||
62 | GeForce GTX 1050 Laptop | 6093 | 2809 | 3284 | 2.0.0 | 640 (CUDA) | 1.35 / 1.49 | CUDA 10 | Debian 10 | 4 | ||||
63 | AMD Radeon RX 480 | 8090 | 4872 | 3218 | 2.1.0 | 2304 | 1.12 / 1.27 | ROCm (OpenCL) | Ubuntu 18.04 | 8 | ||||
64 | AMD Radeon RX 470 | 7704 | 4623 | 3081 | 2.1.0 | 2048 | 0.93 / 1.21 | ROCm (OpenCL) | Debian 10 | 4 | ||||
65 | NVIDIA Quadro P4200 | 5422 | 2473 | 2949 | 1.14.0 | 2304 (CUDA) | 1.23 / 1.65 | CUDA 10 | Debian 10 | 8 | ||||
66 | NVIDIA Quadro P1000 | 5090 | 2367 | 2724 | 2.0.0 | 640 (CUDA) | 1.27 / 1.48 | CUDA 10 | Debian 10 | 4 | ||||
67 | NVIDIA Quadro M2200 | 5173 | 2479 | 2694 | 2.1.0 | 1024 (CUDA) | 0.70 / 1.04 | CUDA 10 | Windows 10 | 4 | ||||
68 | NVIDIA Tegra Xavier | 4625 | 2226 | 2399 | 2.0.0 | 512 (CUDA) | 0.85 / 1.38 | CUDA 10 | Ubuntu 18.04 | 16 | ||||
69 | NVIDIA Quadro M2000M (1) | 4093 | 1882 | 2211 | 1.14.0 | 640 (CUDA) | 1.04 / 1.20 | CUDA 10 | Fedora 30 | 4 | ||||
70 | NVIDIA Quadro M2000M (3) | 3966 | 1843 | 2123 | 1.14.0 | 640 (CUDA) | 1.04 / 1.20 | CUDA 10 | Windows 10 | 4 | ||||
71 | GeForce GTX 960M | 3928 | 1826 | 2102 | 2.1.0 | 640 (CUDA) | 1.10 / 1.18 | CUDA 10.1 | Windows 10 | 4 | ||||
72 | NVIDIA Quadro M2000M (2) | 3996 | 1837 | 2095 | 1.15.0 | 640 (CUDA) | 1.04 / 1.20 | CUDA 10 | Windows 10 | 4 | ||||
73 | NVIDIA Quadro K5000 | 3210 | 1502 | 1708 | 2.1.0 | 1536 (CUDA) | 0.71 | CUDA 10 | CentOS 7.6 | 4 | ||||
74 | Intel Xeon Gold 6148 | 4190 | 2588 | 1602 | 2.1.0 | 20 / 40 Threads | 2.4 / 3.7 | Intel MKL | CentOS 7.5 | 384 | ||||
75 | AMD Threadripper 3970X | 3829 | 2258 | 1571 | 2.1.0 | 32 / 64 Threads | 3.7 / 4.5 | Intel MKL | Ubuntu 18.04 | 128 | ||||
76 | Intel Core i9-9940X | 3262 | 1750 | 1512 | 1.15.0 | 14 / 28 Threads | 3.3 / 4.4 | Intel MKL | Debian 10 | 128 | ||||
77 | Intel Core i9-7940X | 3468 | 1963 | 1505 | 1.14.0 | 14 / 28 Threads | 3.1 / 4.3 | Intel MKL | Debian 11 | 32 | ||||
78 | Intel Xeon Gold 6130 | 3698 | 2211 | 1487 | 2.1.0 | 16 / 32 Threads | 2.1 / 3.7 | Intel MKL | CentOS 7.6 | 192 | ||||
79 | Intel Xeon E5-2620 v4 x 2 (1) | 2569 | 1108 | 1461 | 1.14-Conda | 16 / 32 Threads | 2.1 / 3.0 | Intel MKL | Debian 9.9 | 16 | ||||
80 | Intel Xeon Gold 6126 | 3514 | 2055 | 1459 | 2.0.0 | 12 / 24 Threads | 2.6 / 3.7 | Intel MKL | Red Hat 7.7 | 192 | ||||
81 | Intel Core i7-9700K | 2480 | 1038 | 1442 | 1.14-Conda | 8 / 8 Threads | 3.6 / 4.9 | Intel MKL | Debian 9.9 | 32 | ||||
82 | Intel Core i9-7900X | 2994 | 1598 | 1396 | 2.1.0 | 10 / 20 Threads | 3.3 / 4.3 | Intel MKL | Ubuntu 19.04 | 64 | ||||
83 | Intel Core i9-9960X | 3176 | 1800 | 1376 | 2.1.0 | 16 / 32 Threads | 3.1 / 4.4 | Intel MKL | Ubuntu 18.04 | 64 | ||||
84 | AMD Ryzen 9 3950X | 2973 | 1612 | 1361 | 2.1.0 | 16 / 32 Threads | 3.5 / 4.7 | Intel MKL | Debian 10 | 64 | ||||
85 | GeForce GTX 950M | 2512 | 1157 | 1355 | 2.1.0 | 640 (CUDA) | 0.99 / 1.12 | CUDA 10.1 | Ubuntu 18.04 | 4 | ||||
86 | Intel Xeon W-2195 | 3032 | 1708 | 1324 | 2.1.0 | 18 / 36 Threads | 2.3 / 4.3 | Intel MKL | Ubuntu 18.04 | 512 | ||||
87 | AMD Ryzen 9 3900X | 2832 | 1517 | 1315 | 2.1.0 | 12 / 24 Threads | 3.8 / 4.6 | Intel MKL | Debian 10 | 32 | ||||
88 | Intel Xeon E5-2698 v4 | 3000 | 1764 | 1236 | 1.14.0 | 20 / 40 Threads | 2.2 / 3.6 | Intel MKL | Ubuntu 18.04 | 512 | ||||
89 | Intel Core i9-9820X | 2498 | 1274 | 1224 | 2.0.0 | 10 / 20 Threads | 3.3 / 4.1 | Intel MKL | Ubuntu 18.04 | 64 | ||||
90 | Intel Xeon Gold 6248 | 2978 | 1755 | 1223 | 1.15.0 | 20 / 40 Threads | 2.5 / 3.9 | Intel MKL | Ubuntu 16.04 | 192 | ||||
91 | Intel Xeon Gold 6134 | 2477 | 1273 | 1204 | 1.15.0 | 8 / 16 Threads | 3.2 / 3.7 | Intel MKL | Ubuntu 18.04 | 22 | ||||
92 | AMD EPYC 7302 | 2665 | 1464 | 1201 | 1.14.0 | 16 / 32 Threads | 3.0 / 3.3 | Intel MKL | Ubuntu 18.04 | 128 | ||||
93 | Intel Core i7-8700 (1) | 2025 | 847 | 1178 | 1.14-Conda | 6 / 12 Threads | 3.2 / 4.6 | Intel MKL | Debian 9.9 | 32 | ||||
94 | Intel Core i9-9900K | 2402 | 1230 | 1172 | 1.14.0 | 8 / 16 Threads | 3.6 / 5.0 | Intel MKL | Ubuntu 19.10 | 32 | ||||
95 | Intel Core i7-9800X | 2508 | 1354 | 1154 | 1.14.0 | 8 / 16 Threads | 3.8 / 4.4 | Intel MKL | Ubuntu 18.04 | 32 | ||||
96 | Intel Xeon W-2135 | 2321 | 1201 | 1120 | 1.14.0 | 6 / 12 Threads | 3.7 / 4.5 | Intel MKL | Debian 10 | 128 | ||||
97 | Intel Core i7-9700KF | 2303 | 1204 | 1099 | 1.14.0 | 8 / 8 Threads | 3.6 / 4.9 | Intel MKL | Ubuntu 18.04 | 64 | ||||
98 | Intel Core i7-7820X | 2354 | 1271 | 1083 | 2.1.0 | 8 / 16 Threads | 3.6 / 4.3 | Intel MKL | Red Hat 7.6 | 128 | ||||
99 | AMD EPYC 7502 | 2723 | 1686 | 1037 | 1.14.0 | 32 / 64 Threads | 2.5 / 3.4 | Intel MKL | Ubuntu 18.04 | 1024 | ||||
100 | AMD Ryzen 7 3800X | 2206 | 1188 | 1018 | 2.1.0 | 8 / 16 Threads | 3.9 / 4.5 | Intel MKL | Debian 11 | 16 | ||||
101 | Intel Core i7-8086K | 2001 | 1009 | 992 | 1.14.0 | 6 / 12 Threads | 4.0 / 5.0 | Intel MKL | Debian 9.9 | 32 | ||||
102 | Intel Core i7-9700F | 1928 | 952 | 976 | 1.15.0 | 8 / 8 Threads | 3.0 / 4.7 | Intel MKL | Debian 10 | 32 | ||||
103 | AMD Threadripper 1920X | 1896 | 940 | 956 | 2.1.0 | 12 / 24 Threads | 3.5 / 4.0 | Intel MKL | Ubuntu 19.04 | 64 | ||||
104 | Intel Core i7-8700 (2) | 1859 | 903 | 956 | 1.14 | 6 / 12 Threads | 3.2 / 4.6 | Intel MKL | Ubuntu 18.04 | 32 | ||||
105 | AMD Ryzen 5 3600 | 1905 | 990 | 915 | 1.14.0 | 6 / 12 Threads | 3.6 / 4.2 | Intel MKL | Ubuntu 16.04 | 16 | ||||
106 | AMD EPYC 7601 | 2113 | 1199 | 914 | 2.1.0 | 32 / 64 Threads | 2.2 / 3.2 | Intel MKL | Debian 10 | 256 | ||||
107 | Intel Core i7-6850K | 1732 | 821 | 911 | 2.1.0 | 6 / 12 Threads | 3.6 / 3.8 | Intel MKL | Debian 10 | 128 | ||||
108 | GeForce 940MX | 1672 | 774 | 898 | 1.14.0 | 512 (CUDA) | 0.80 / 0.86 | CUDA 10 | RHEL 7.7 | 4 | ||||
109 | Intel Xeon W-2123 | 1761 | 882 | 879 | 2.1.0 | 4 / 8 Threads | 3.6 / 3.9 | Intel MKL | Ubuntu 18.04 | 32 | ||||
110 | Intel Xeon E5-2620 v4 x 2 (2) | 1841 | 968 | 873 | 1.14.0 | 16 / 32 Threads | 2.1 / 3.0 | Intel MKL | Debian 9.9 | 16 | ||||
111 | Intel Core i5-8400 | 1706 | 848 | 859 | 1.14.0 | 6 / 6 Threads | 2.8 / 4.0 | Intel MKL | Ubuntu 18.04 | 32 | ||||
112 | AMD Ryzen 7 3700X | 1883 | 1026 | 857 | 1.14.0 | 8 / 16 Threads | 3.6 / 4.4 | Intel MKL | Ubuntu 19.10 | 32 | ||||
113 | Intel Core i7-5930K | 1652 | 798 | 854 | 1.14.0 | 6 / 12 Threads | 3.5 / 3.7 | Intel MKL | Ubuntu 18.04 | 32 | ||||
114 | Intel Core i7-6700K (1) | 1448 | 608 | 840 | 1.14-Conda | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.9 | 32 | ||||
115 | Intel Xeon D-2183IT | 1769 | 938 | 831 | 1.14.0 | 12 / 24 Threads | 2.1 / 3.0 | Intel MKL | CentOS 7.6 | 256 | ||||
116 | AMD Ryzen 5 3500X | 1760 | 929 | 831 | 2.1.1 | 6 / 6 Threads | 3.6 / 4.1 | Intel MKL | Debian 10 | 16 | ||||
117 | Intel Xeon E5-2650 v4 | 1644 | 823 | 821 | 2.0.0 | 12 / 24 Threads | 2.2 / 2.9 | Intel MKL | CentOS 7.4 | 128 | ||||
118 | Intel Core i7-8850H | 1558 | 737 | 821 | 2.1.0 | 6 / 12 Threads | 2.6 / 4.3 | Intel MKL | Debian 10 | 32 | ||||
119 | Intel Core i7-4790K | 1400 | 582 | 818 | 1.14-Conda | 4 / 8 Threads | 4.0 / 4.4 | Intel MKL | Debian 9.9 | 16 | ||||
120 | Intel Core i7-9750H | 1567 | 751 | 816 | 1.14.0 | 6 / 12 Threads | 2.6 / 4.5 | Intel MKL | Ubuntu 18.04 | 32 | ||||
121 | Intel Core i7-7700K | 1568 | 759 | 809 | 1.14.0 | 4 / 8 Threads | 4.2 / 4.5 | Intel MKL | Debian 9.9 | 32 | ||||
122 | Intel Core i9-9980HK | 1633 | 826 | 807 | 2.1.0 | 8 / 16 Threads | 2.4 / 5.0 | Intel MKL | macOS Catalina | 32 | ||||
123 | AMD EPYC 7451 | 1769 | 964 | 805 | 2.0.0 | 24 / 48 Threads | 2.3 / 3.2 | Intel MKL | Ubuntu 18.04 | 256 | ||||
124 | AMD Ryzen 7 2700X | 1499 | 721 | 778 | 2.0.0 | 8 / 16 Threads | 3.7 / 4.3 | Intel MKL | Ubuntu 18.04 | 32 | ||||
125 | Intel Core i7-8750H | 1476 | 701 | 776 | 2.2.0 | 6 / 12 Threads | 2.2 / 4.1 | Intel MKL | Ubuntu 18.04 | 32 | ||||
126 | Intel Core i7-6700K (2) | 1413 | 655 | 699 | 2.1.0 | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.11 | 32 | ||||
127 | Intel Core i7-6700K (3) | 1354 | 655 | 699 | 1.14.0 | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.9 | 32 | ||||
128 | NVIDIA Tegra X2 | 1412 | 720 | 693 | 2.0.0 | 256 (CUDA) | 0.85 / 1.47 | CUDA 10 | Ubuntu 18.04 | 4 | ||||
129 | Intel Core i9-9880H | 1471 | 786 | 685 | 1.14.0 | 8 / 16 Threads | 2.3 / 4.8 | Intel MKL | Debian 10 | 16 | ||||
130 | Intel Xeon E5-2695 v2 | 1255 | 578 | 678 | 2.1.0 | 12 / 24 Threads | 2.4 / 3.2 | Intel MKL | Ubuntu 18.04 | 128 | ||||
131 | Intel Xeon E3-1505M v5 (1) | 1125 | 449 | 676 | 1.14-Conda | 4 / 8 Threads | 2.8 / 3.7 | Intel MKL | Fedora 30 | 16 | ||||
132 | Intel Core i7-6700K (4) | 1241 | 577 | 664 | 2.0.0 | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.11 | 32 | ||||
133 | Intel XeonSilver 4208 | 1232 | 589 | 643 | 1.14.0 | 8 / 16 Threads | 2.1 / 3.2 | Intel MKL | Ubuntu 18.04 | 32 | ||||
134 | Intel XeonSilver 4110 | 1197 | 556 | 642 | 1.14.0 | 8 / 16 Threads | 2.1 / 3.0 | Intel MKL | Ubuntu 18.04 | 96 | ||||
135 | Intel Xeon E5-2620 v3 | 1254 | 614 | 640 | 1.14.0 | 6 / 12 Threads | 2.4 / 3.2 | Intel MKL | Debian 10 | 24 | ||||
136 | AMD Ryzen 7 2700 | 1226 | 589 | 637 | 2.1.0 | 8 / 16 Threads | 3.2 / 4.1 | Intel MKL | Debian 10 | 16 | ||||
137 | Intel Core i7-8569U | 1168 | 539 | 629 | 1.15.0 | 4 / 8 Threads | 2.8 / 4.7 | Intel MKL | macOS Catalina | 16 | ||||
138 | AMD Ryzen 7 1700X | 1214 | 589 | 625 | 1.14.0 | 8 / 16 Threads | 3.4 / 3.8 | Intel MKL | Ubuntu 18.04 | 32 | ||||
139 | Intel Core i5-7400 | 1189 | 564 | 625 | 1.14.0 | 4 / 4 Threads | 3.0 / 3.5 | Intel MKL | Ubuntu 18.04 | 32 | ||||
140 | Intel Xeon E3-1231 v3 | 1175 | 569 | 607 | 2.2.0 | 4 / 8 Threads | 3.4 / 3.8 | Intel MKL | Debian 10 | 32 | ||||
141 | IBM POWER9 | 1563 | 960 | 604 | 2.1.0 | 32 / 128 Threads | 2.9 / 3.8 | Intel MKL | Debian 10 | 320 | ||||
142 | Intel Xeon E5-2650 v2 | 1256 | 653 | 603 | 1.14.0 | 8 / 16 Threads | 2.6 / 3.4 | Intel MKL | Debian 9.9 | 128 | ||||
143 | AMD Ryzen 7 1700 | 1164 | 563 | 602 | 1.14.0 | 8 / 16 Threads | 3.0 / 3.7 | Intel MKL | Ubuntu 19.10 | 32 | ||||
144 | Intel Core i5-6500 | 1151 | 554 | 597 | 1.15.0 | 4 / 4 Threads | 3.2 / 3.6 | Intel MKL | Debian 10 | 16 | ||||
145 | Intel Core i7-4770K | 1152 | 566 | 587 | 1.14.0 | 4 / 8 Threads | 3.5 / 3.9 | Intel MKL | Debian 9.9 | 16 | ||||
146 | AMD Ryzen 5 2600 | 1108 | 523 | 585 | 1.14.0 | 6 / 12 Threads | 3.4 / 3.9 | Intel MKL | Elementary 5.1 | 32 | ||||
147 | Intel Xeon E5-1620 v3 | 1054 | 477 | 577 | 2.0.0 | 4 / 8 Threads | 3.5 / 3.6 | Intel MKL | Debian 9.9 | 82 | ||||
148 | Intel Core i5-8279U | 1085 | 511 | 574 | 1.14.0 | 4 / 8 Threads | 2.4 / 4.1 | Intel MKL | macOS Catalina | 8 | ||||
149 | Intel Core i5-6400 | 1091 | 523 | 568 | 1.14.0 | 4 / 4 Threads | 2.7 / 3.3 | Intel MKL | Ubuntu 18.04 | 32 | ||||
150 | Intel Core i7-6700K (5) | 1091 | 529 | 562 | 2.0.0-Intel | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.11 | 32 | ||||
151 | Intel Core i5-9300H | 1087 | 528 | 559 | 1.14.0 | 4 / 8 Threads | 2.4 / 4.1 | Intel MKL | Ubuntu 16.04 | 12 | ||||
152 | AMD Ryzen 5 1600 | 1050 | 496 | 554 | 2.1.0 | 6 / 12 Threads | 3.2 / 3.6 | Intel MKL | Debian 10 | 16 | ||||
153 | Intel Core i7-7700HQ | 993 | 451 | 542 | 1.15.0 | 4 / 8 Threads | 2.8 / 3.8 | Intel MKL | Ubuntu 16.04 | 16 | ||||
154 | Intel Core i7-8565U | 958 | 424 | 534 | 2.1.0 | 4 / 8 Threads | 1.8 / 4.6 | Intel MKL | Ubuntu 18.04 | 32 | ||||
155 | Intel Xeon E3-1505M v5 (2) | 950 | 423 | 527 | 1.14.0 | 4 / 8 Threads | 2.8 / 3.7 | Intel MKL | Fedora 30 | 16 | ||||
156 | Intel Core i5-1035G1 | 970 | 444 | 526 | 1.14.0 | 4 / 8 Threads | 1.0 / 3.6 | Intel MKL | Ubuntu 18.04 | 16 | ||||
157 | Intel Core i3-8100 | 1062 | 539 | 524 | 1.14.0 | 4 / 4 Threads | 3.6 | Intel MKL | Ubuntu 18.04 | 16 | ||||
158 | Intel Xeon E5-2623 v4 | 1012 | 515 | 497 | 1.14.0 | 4 / 8 Threads | 2.6 / 3.2 | Intel MKL | Debian 9.9 | 32 | ||||
159 | Intel Core i7-7820HQ | 969 | 475 | 494 | 1.14.0 | 4 / 8 Threads | 2.9 / 3.9 | Intel MKL | macOS Mojave | 16 | ||||
160 | Intel Core i7-3930K | 882 | 390 | 492 | 1.14.0 | 6 / 12 Threads | 3.2 / 3.8 | Intel MKL | Gentoo (4.20) | 32 | ||||
161 | Intel Core i7-8550U | 834 | 374 | 460 | 1.14.0 | 4 / 8 Threads | 1.8 / 4.0 | Intel MKL | Ubuntu 18.04 | 16 | ||||
162 | Intel Core i7-6700K (6) | 790 | 330 | 460 | 1.12.0 | 4 / 8 Threads | 4.0 / 4.2 | AVX | Debian 9.9 | 32 | ||||
163 | NVIDIA Carmel (ARMv8.2-A) | 819 | 386 | 433 | 2.0.0 | 8 / 8 Threads | 2.26 | ARM NEON | Ubuntu 18.04 | 16 | ||||
164 | Intel Core i3-7100 | 815 | 383 | 432 | 1.14.0 | 2 / 4 Threads | 3.9 | Intel MKL | Ubuntu 18.04 | 8 | ||||
165 | Intel Core i7-4870HQ | 822 | 392 | 430 | 2.0.0 | 4 / 8 Threads | 2.7 / 3.6 | Intel MKL | macOS Mojave | 16 | ||||
166 | Intel Core i7-4870HQ | 750 | 338 | 412 | 2.0.0 | 4 / 8 Threads | 2.5 / 3.7 | Intel MKL | macOS Catalina | 16 | ||||
167 | Intel Core i7-3820 | 665 | 289 | 376 | 1.14.0 | 4 / 8 Threads | 3.6 / 3.8 | Intel MKL | Ubuntu 19.04 | 32 | ||||
168 | Intel Core i7-7500U | 664 | 290 | 374 | 2.1.0 | 2 / 4 Threads | 2.7 / 3.5 | Intel MKL | Ubuntu 18.04 | 16 | ||||
169 | Intel Xeon E3-1505M v5 (4) | 564 5 | 197 | 367 | 2.0.0-Conda | 4 / 8 Threads | 2.8 / 3.7 | Intel MKL | Windows 10 | 16 | ||||
170 | Intel Core i7-6700K (7) | 711 | 358 | 353 | 1.13.1-Intel | 4 / 8 Threads | 4.0 / 4.2 | Intel MKL | Debian 9.9 | 32 | ||||
171 | Intel Core i7-4850HQ | 629 | 290 | 339 | 2.0.0 | 4 / 8 Threads | 2.3 / 3.7 | Intel MKL | macOS Catalina | 16 | ||||
172 | Intel Core i7-3770K | 612 | 273 | 339 | 1.14-Conda | 4 / 8 Threads | 3.5 / 3.9 | Intel MKL | Debian 9.9 | 16 | ||||
173 | Intel Core i5-7300U | 623 | 289 | 334 | 1.14.0 | 2 / 4 Threads | 2.6 / 3.5 | Intel MKL | Ubuntu 18.04 | 16 | ||||
174 | Intel Core i5-7360U | 608 | 281 | 327 | 1.15.0 | 2 / 4 Threads | 2.3 / 3.6 | Intel MKL | macOS Catalina | 8 | ||||
175 | Intel Core i5-8250U | 583 | 262 | 321 | 2.0.0 | 4 / 8 Threads | 1.6 / 3.4 | Intel MKL | Debian 10.2 | 16 | ||||
176 | Intel Core i7-3720QM | 569 | 249 | 320 | 2.1.0 | 4 / 8 Threads | 2.6 / 3.6 | Intel MKL | macOS Catalina | 16 | ||||
177 | Intel Core i7-2600K | 566 | 251 | 316 | 1.14.0 | 4 / 8 Threads | 3.4 / 3.8 | Intel MKL | Debian 9.9 | 16 | ||||
178 | Intel Core i7-3635QM | 568 | 253 | 315 | 2.1.0 | 4 / 8 Threads | 2.4 / 3.4 | Intel MKL | Fedora 31 | 8 | ||||
179 | Intel Core i7-2700K | 566 | 259 | 307 | 1.14-Conda | 4 / 8 Threads | 3.5 / 3.9 | Intel MKL | Debian 9.9 | 16 | ||||
180 | Intel Core i7-6700K (8) | 486 | 186 | 300 | 1.5.0 | 4 / 8 Threads | 4.0 / 4.2 | no AVX | Debian 9.9 | 32 | ||||
181 | Intel Xeon E3-1505M v5 (3) | 612 5 | 313 | 299 | 1.14.0 | 4 / 8 Threads | 2.8 / 3.7 | Intel MKL | Windows 10 | 16 | ||||
182 | Intel Core i7-5500U | 517 | 224 | 293 | 2.1.0 | 2 / 4 Threads | 2.4 / 3.0 | Intel MKL | Ubuntu 18.04 | 16 | ||||
183 | Intel Xeon E3-1505M v5 (5) | 518 5 | 240 | 278 | 2.0.0 | 4 / 8 Threads | 2.8 / 3.7 | Intel MKL | Windows 10 | 16 | ||||
184 | Intel Core i7-4510U | 434 | 189 | 245 | 2.1.0 | 2 / 4 Threads | 2.0 / 3.1 | Intel MKL | Debian 10 | 16 | ||||
185 | AMD Ryzen 3 1200 | 502 5 | 288 | 214 | 1.14.0 | 4 / 4 Threads | 3.8 | Intel MKL | Windows 10 | 8 | ||||
186 | Intel Core i5-6200U | 414 5 | 204 | 210 | 2.2.0 | 2 / 4 Threads | 2.3 / 2.8 | Intel MKL | Windows 10 | 8 | ||||
187 | Intel Xeon E3-1505M v5 (6) | 378 5 | 173 | 205 | 1.13.1 | 4 / 8 Threads | 2.8 / 3.7 | AVX | Windows 10 | 16 | ||||
188 | AMD FX-8300 | 346 | 150 | 196 | 2.2.0 | 8 / 8 Threads | 3.3 / 3.9 | Intel MKL | Ubuntu 20.04 | 16 | ||||
189 | Intel Core i7-950 (1) | 328 | 133 | 195 | 1.14-Conda | 4 / 8 Threads | 3.06 / 3.33 | Intel MKL | Debian 9.9 | 16 | ||||
190 | Intel Core i5-4258U | 382 | 192 | 190 | 2.1.0 | 2 / 4 Threads | 2.4 / 2.9 | Intel MKL | macOS Mojave | 8 | ||||
191 | Intel Core i7-950 (2) | 293 | 111 | 182 | 1.5.0 | 4 / 8 Threads | 3.06 / 3.33 | no AVX | Debian 9.9 | 16 | ||||
192 | Intel Core i7-930 | 288 | 110 | 178 | 1.5.0 | 4 / 8 Threads | 2.80 / 3.06 | no AVX | Debian 9.9 | 16 | ||||
193 | Zhaoxin KaiXian KX-U6780A | 220 | 93 | 127 | 1.14.0 | 8 / 8 Threads | 2.7 | Intel MKL | Ubuntu 18.04 | 16 | ||||
194 | Intel Core i3-3110M | 180 5 | 79 | 101 | 2.2.0 | 2 / 4 Threads | 2.4 | Intel MKL | Windows 10 | 12 | ||||
195 | GeForce GT 1030 | 2750 1 | 1375 | 0 | 1.12.0 | 384 (CUDA) | 1.23 / 1.47 | CUDA 9 | Debian 9.9 | 2 | ||||
196 | GeForce GT 710 | 470 1 | 235 | 0 | 1.12.0 | 96 (CUDA) | 0.78 / 0.80 | CUDA 9 | Debian 9.9 | 2 |
AI繪圖|圖片生成速度 數據(每秒幾張)
AI繪圖是小編最愛的功能之一,Midjourney、Stable diffusion等都非常好用,小編測試了一份顯示卡運行SD的出圖速度及價格,新手小白可能知道AI圖片生成吃電腦配置,吃顯示卡配置但怎麼選擇顯示卡卻不太清楚,下方這張表格,一定能幫助到你。
👇 本表格都是入門款,價格實惠,若追求高規格可以參考上一段文中的排行選擇
其他AI繪圖 Q&A
長期使用AI繪圖,該選擇雲端主機或自架主機
【自架主機】
優點:本地部署後想怎麼搞就怎麼搞,不用擔心後續其他成本(電費除外)
缺點:前期一次性投入成本高(一台電腦落地)
【雲端主機】
優點:按月付費,一次性成本低,有一般電腦就能操作
缺點:雲端服務部署複雜,沒有專業人士支援或動手能力較差可能折騰的沒脾氣出圖數量根據套餐量來決定,有可能套餐用完了,還要出圖就要增加套餐了
【小編建議】
自架主機貴在顯示卡,其實說到底,貴在時間,我的1050ti 4g顯卡也能出圖,大概10分鐘左右出1張而且顯卡熱的感覺快冒煙boom了,如果貴的買不起(4090 )但又想有快速的出圖體驗(和4090差不多),能接受一定風險的話,可以考慮一下2080Ti,如果想更便宜,動手能力強,可以考慮特斯拉 p40