ATI 新版驱动可以让 58xx 系正常跑 TessMark 了

Edison

[TessMark] OpenGL 4 Tessellation Benchmark: Comparative Table

Now that Catalyst 10.8 drivers have fixed the Radeon HD 58xx tessellation bug, I can publish comparative tables for TessMark, the OpenGL 4 tessellation benchmark. The following results are valid with the 0.2.x branch of TessMark. Use these results as a quick way to rank your graphics card regarding the GPU tessellation.

This post will be regularly updated.

1 – Geeks3D Results

Testbed:
- CPU: Core i7 960 @ 3.2GHz
- RAM: 4GB DDR3 Corsair Dominator
- Motherboard: GIGABYTE X58-A UD5
- Windows 7 64-bit
- R258.96 + Cat 10.8
- PSU: Corsair AX1200
- All graphics cards have been tested with default clocks.

Tessellation: moderate

97804 (1632FPS) – GeForce GTX 480 3-way SLI (R258.96)

81631 (1364FPS) – GeForce GTX 480 2-way SLI (R258.96)

48084 (804FPS) – EVGA GeForce GTX 480 (R258.96)

30512 (510FPS) – MSI N460GTX Cyclone 768D5 OC (R258.96)

29232 (487FPS) – MSI R5770 Hawk CrossFire (Cat 10.8)

27000 (450FPS) – ATI Radeon HD 5870 (Cat 10.8)

20366 (340FPS) – MSI R5770 Hawk (Cat 10.8)

5185 – Mobility Radeon HD 5470 (Cat 10.7)

Tessellation: normal

69952 (1167FPS) – GeForce GTX 480 3-way SLI (R258.96)

52648 (879FPS) – GeForce GTX 480 2-way SLI (R258.96)

29196 (487FPS) – EVGA GeForce GTX 480 (R258.96)

17452 (291FPS) – MSI N460GTX Cyclone 768D5 OC (R258.96)

13667 (228FPS) – MSI R5770 Hawk CrossFire (Cat 10.8)

8177 (136FPS) – ATI Radeon HD 5870 (Cat 10.8)

7515 (125FPS) – MSI R5770 Hawk (Cat 10.8)

2349 – Mobility Radeon HD 5470 (Cat 10.7)

Tessellation: extreme

34934 (583FPS) – GeForce GTX 480 3-way SLI (R258.96)

24837 (414FPS) – GeForce GTX 480 2-way SLI (R258.96)

13008 (217FPS) – EVGA GeForce GTX 480 (R258.96)

6729 (112FPS) – MSI N460GTX Cyclone 768D5 OC (R258.96)

4094 (68FPS) – MSI R5770 Hawk CrossFire (Cat 10.8)

2150 (36FPS) – ATI Radeon HD 5870 (Cat 10.8)

2114 (35FPS) – MSI R5770 Hawk (Cat 10.8)

820 – Mobility Radeon HD 5470 (Cat 10.7)

Tessellation: insane

7810 (130FPS) – GeForce GTX 480 2-way SLI (R258.96)

3963 (66FPS) – EVGA GeForce GTX 480 (R258.96)

1959 (33FPS) – MSI N460GTX Cyclone 768D5 OC (R258.96)

1028 (18FPS) – MSI R5770 Hawk CrossFire (Cat 10.8)

551 (10FPS) – MSI R5770 Hawk (Cat 10.8)

548 (10FPS) – ATI Radeon HD 5870 (Cat 10.8)

234 – Mobility Radeon HD 5470 (Cat 10.7)

2 – Results from around the Web

Tessellation: normal – 1024×768 fullscreen

65391 (1091FPS) – GeForce GTX 480 2-way SLI, 0X MSAA, (GPU core=700MHz, GPU mem=1848MHz, CPU=Intel Core i7 X980 @ 3.33GHz, source)

27910 (466FPS) – GeForce GTX 470, 0X MSAA, (GPU core=655MHz, GPU mem=1125MHz, CPU=AMD Phenom II X4 940, source)

24985 (416FPS) – GeForce GTX 460, 0X MSAA, (GPU core=920MHz, GPU mem=2200MHz, CPU=Intel Core Q9650 @ 3GHz, source)

18922 (315FPS) – GeForce GTX 460, 0X MSAA, (GPU core=675MHz, GPU mem=1800MHz, CPU=Intel Core Q9650 @ 3GHz, source)

9081 (151FPS) – Radeon HD 5870 (Cat 10.7a), 0X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

8679 (145FPS) – Radeon HD 5870 (Cat 10.7a), 8X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

Tessellation: extreme – 1024×768 fullscreen

10786 (180FPS) – GeForce GTX 470, 0X MSAA, (GPU core=655MHz, GPU mem=1125MHz, CPU=AMD Phenom II X4 940, source)

8787 (146FPS) – GeForce GTX 460, 0X MSAA, (GPU core=920MHz, GPU mem=2200MHz, CPU=Intel Core Q9650 @ 3GHz, source)

6518 (109FPS) – GeForce GTX 460, 0X MSAA, (GPU core=675MHz, GPU mem=1800MHz, CPU=Intel Core Q9650 @ 3GHz, source)

2300 (39FPS) – Radeon HD 5870 (Cat 10.7a), 0X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

Tessellation: insane – 1024×768 fullscreen

8011 (133FPS) – GeForce GTX 480 2-way SLI, 0X MSAA, (GPU core=700MHz, GPU mem=1848MHz, CPU=Intel Core i7 X980 @ 3.33GHz, source)

2475 (41FPS) – GeForce GTX 460, 0X MSAA, (GPU core=920MHz, GPU mem=2200MHz, CPU=Intel Core Q9650 @ 3GHz, source)

1821 (31FPS) – GeForce GTX 460, 0X MSAA, (GPU core=675MHz, GPU mem=1800MHz, CPU=Intel Core Q9650 @ 3GHz, source)

580 (10FPS) – Radeon HD 5870 (Cat 10.7a), 0X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

Tessellation: normal – 1680×1050 fullscreen

39135 (653FPS) – GeForce GTX 460, 0X MSAA (GPU core=880MHz, GPU mem=2125MHz, CPU=Core i5 750 @ 2.67GHz, source)

34705 (579FPS) – GeForce GTX 480, 0X MSAA (GPU core=843MHz, GPU mem=1980MHz, CPU=Core i7 920 @ 2.67GHz, source)

9678 (161FPS) – Radeon HD 5850 (Cat 10.8), 0X MSAA, (GPU core=1000MHz, GPU mem=1250MHz, CPU=AMD Phenom II X6 1090T, source)

8476 (141FPS) – Radeon HD 5850 (Cat 10.8), 0X MSAA, (GPU core=900MHz, GPU mem=1200MHz, CPU=Core i5 750 @ 2.67GHz, source)

Tessellation: insane – 1680×1050 fullscreen

3595 (60FPS) – GeForce GTX 470, 4X MSAA (GPU core=780MHz, GPU mem=1750MHz, CPU=Core i5 750 @ 2.67GHz, source)

3171 (60FPS) – GeForce GTX 470, 4X MSAA (GPU core=607MHz, GPU mem=1674MHz, CPU=Core 2 Quad Q9400 @ 2.66GHz, source)

319 (6FPS) – Radeon HD 5500 Series, 0X MSAA (GPU core=550MHz, GPU mem=400MHz, CPU=Pentium Dual Core E5300 @ 2.60GHz, source)

Tessellation: extreme – 1920×1080 fullscreen

22074 (368FPS) – GeForce GTX 470, 0X MSAA, (GPU core=700MHz, GPU mem=1700MHz, CPU=Core i7 920 @ 2.67GHz, source)

915 (16FPS) – Radeon HD 5970, 8X MSAA, (GPU core=725MHz, GPU mem=1000MHz, CPU=AMD Phenom II X4 955, source)

Tessellation: insane – 1920×1080 fullscreen

5469 (91FPS) – GeForce GTX 470, 16X MSAA, (GPU core=830MHz, GPU mem=1800MHz, CPU=Core i5 750 @ 2.67GHz, source)

915 (16FPS) – Radeon HD 5970, 8X MSAA, (GPU core=725MHz, GPU mem=1000MHz, CPU=AMD Phenom II X4 955, source)

Tessellation: normal – 1920×1200 fullscreen

35637 (595FPS) – GeForce GTX 460, 0X MSAA (GPU core=880MHz, GPU mem=2125MHz, CPU=Core i5 750 @ 2.67GHz, source)

26641 (445FPS) – GeForce GTX 480, 0X MSAA (GPU core=700MHz, GPU mem=1848MHz, CPU=Core i7 960 @ 3.20GHz, source)

9507 (158FPS) – Radeon HD 5850 (Cat 10.8), 0X MSAA, (GPU core=1000MHz, GPU mem=1250MHz, CPU=AMD Phenom II X6 1090T, source)

8634 (144FPS) – Radeon HD 5870 (Cat 10.7a), 0X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

7740 (129FPS) – Radeon HD 5870 (Cat 10.7a), 8X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

Tessellation: insane – 1920×1200 fullscreen

7740 (129FPS) – Radeon HD 5870 (Cat 10.7a), 8X MSAA, (GPU core=900MHz, GPU mem=1250MHz, CPU=Intel Core Q9550 @ 2.83GHz, source)

5781 (97FPS) – GeForce GTX 480, 16X MSAA, (GPU core=860MHz, GPU mem=1999MHz, CPU=Intel Core i7 X980 @ 3.33GHz, source)

yellowfly

反正吵架的人基本都会选择性无视的

goodayoo

然后年底等HD6系列出来时，肯定又能当很久单 ...
我只想请教这个结论是如何得出的？
LockheedMartin 发表于 2010-8-29 14:39

    请你下结论NV的下一代旗舰什么时候出，就能算出来了。

今年低HD6系列肯定出，然后性能超480没有悬念。

goodayoo

等啥时候A卡的单核心能追上N卡的单核心性能，DX11不是骗人的残废性能再来说吧

btw, 三年前 ...
jhj9 发表于 2010-8-28 13:15

    这个嘛，当5870出来时，不是有大半年是单核心王者吗？然后年底等HD6系列出来时，肯定又能当很久单核心王者，GTX580？我连个影子都没看到。

jhj9

哈哈 你这是抬杠嘛  同功耗下480比5970怎么样？

    到底是谁花样别出
380 发表于 2010-8-28 21:02

“同功耗下”这样的假设根本就不该存在
要我说几遍你才能明白？

winson_surewin

win7 64位系统，之前10.7驱动跟植物大战僵尸有冲突，换回10.6就好了

不知10.8有这问题没~

jhj9

这些不是我的yy，是在硬件和本本论坛的见闻。“经常”这两个字本来就不是一个量化的表达，所以我当然 ...
yokuz 发表于 2010-8-28 20:29

不管是在哪里的见闻，说花了的人数比起论坛上的所有人来说有几成？
你完全没注意吧？
所以我才要告诉你，这里这个词用得一点意义都没有

jhj9

你就不要胡搅蛮缠了。统计故障率显然不是普通人能做到的，另一方面也没有义务去帮nv做这个事情。但显 ...
yokuz 发表于 2010-8-28 19:19

既然你不知道故障率到底有多高，那就请不要随意的YY什么“经常花”
总销量那么大，就算有10%花了，你也能随便就看到一大堆，但这不等于“经常花”
故障率偏高是没错，但你在没有客观数据的情况，随便夸大就是不对了。

cv-79

那3D MARK最新版出来AMD5系不是被秒翻了。

kyrome

也用460SLI跑了一下，不超频的话基本上与480相当，超到850则强那么一点
Tessellation: insane

jhj9

不用啥时候了 年底nv就难堪了
同功耗下nv现在性能就是输 300w下 480完败5970  
5870超到300w 4 ...
380 发表于 2010-8-28 13:43

同功耗下相比？你怎么不把QQ和布加迪威龙放在一起比比同功耗下谁跑得更快呢？
这不是搞笑吗？


为了吹嘘A，你真是花样别出，让人匪夷所思啊

wangjianku_0312

没有玩过  学习了。

erhang

回复 28# 李主任


    打平手还不满足吗？
5850和460价钱差多少？