1
00:00:00,400 --> 00:00:06,080
So guys, this is the moment I am finally

2
00:00:03,600 --> 00:00:11,759
testing my 9260 8i RAID card with a configuration that I

3
00:00:08,960 --> 00:00:15,759
think can make it reach an internal bottleneck.

4
00:00:14,080 --> 00:00:23,519
Holy crow, look at that. So I have taken my

5
00:00:19,600 --> 00:00:26,560
dual Elmrest configuration and I have

6
00:00:23,519 --> 00:00:31,599
doubled it. So, I'm running four Intel

7
00:00:26,560 --> 00:00:36,000
SSD 510 series 120 GB drives in RAID

8
00:00:31,599 --> 00:00:38,079
zero using a 128 kilo kilo bit 128k

9
00:00:36,000 --> 00:00:42,480
stripe and all of the other options that I

10
00:00:40,239 --> 00:00:47,680
found to be quite optimal. So, here you can actually see my results with two

11
00:00:44,399 --> 00:00:49,760
drives and uh these are my preliminary

12
00:00:47,680 --> 00:00:55,680
results with a single drive. Focus please. Come on, focus. You can focus.

13
00:00:53,199 --> 00:01:03,920
There you go. Okay, so I'm running five tests at 4 gigs across my 445 gig drive,

14
00:01:01,840 --> 00:01:07,600
which is those four drives in RAID zero. And it looks like we're looking at about

15
00:01:06,080 --> 00:01:11,280
1.5 megabytes per second read. And that

16
00:01:09,840 --> 00:01:16,799
seems to be hitting an internal bottleneck on the RAID card itself. And

17
00:01:14,320 --> 00:01:21,680
then for rights, you can see the rights are just monstrous. So the rights are

18
00:01:19,439 --> 00:01:28,159
actually still scaling from two drives to four drives almost exactly double. So

19
00:01:25,360 --> 00:01:34,320
we're looking at almost 1 Gigabyte per second sequential write speeds to these

20
00:01:31,360 --> 00:01:36,720
four Intel 510 series drives. So I'm going to just wait for the rest of the

21
00:01:35,520 --> 00:01:42,799
test to finish and then I'll show you guys the final results.

22
00:01:40,320 --> 00:01:48,560
Okay, so I finished running Crystal Disc Mark as well as PC Mark Vantage. So you

23
00:01:46,799 --> 00:01:53,280
can see here the suite has finished running. So on the left I have the four

24
00:01:51,040 --> 00:01:58,159
drive array. You can see that random performance doesn't scale uh especially

25
00:01:55,759 --> 00:02:03,360
on the right side quite as well as the sequential performance. On the right, I

26
00:02:00,479 --> 00:02:07,920
have dual drives on the onboard controller because really the only

27
00:02:05,840 --> 00:02:14,560
reason to run a third-party controller on a P67 motherboard would be

28
00:02:12,000 --> 00:02:19,920
if you want to run more than two 6 GB per second SSD drives in RAID zero. So

29
00:02:17,840 --> 00:02:23,840
based on testing two drives on this controller, the 9260 against two drives

30
00:02:21,920 --> 00:02:28,400
on the onboard, performance was very similar. So, we're looking at whether

31
00:02:26,080 --> 00:02:33,200
you would get any benefit from going to four drives from two drives using a

32
00:02:31,840 --> 00:02:38,239
third party controller, which you would need to do because no P67 motherboard

33
00:02:35,920 --> 00:02:42,959
supports more than two SATA 36 GB per second drives in RAID zero. So, here

34
00:02:41,360 --> 00:02:48,080
you've got the Crystal Disc Mark performance numbers. And here the PC

35
00:02:45,680 --> 00:02:54,319
Mark Vantage numbers are a bit of a mystery. So, I'll just show you the

36
00:02:51,760 --> 00:02:57,920
actual scores here. So hold on, let me just try and remember

37
00:02:56,400 --> 00:03:05,680
which is which. Yeah, here we go. So here's the onboard rate where we have 66

38
00:03:00,959 --> 00:03:10,800
megs per second in as the No, that's the

39
00:03:05,680 --> 00:03:13,120
PC mark score. Yeah, 66,644

40
00:03:10,800 --> 00:03:19,120
hard drive PC mark score. Whereas with the four drive RAID zero, we have 58,537

41
00:03:17,040 --> 00:03:23,519
PC mark score. All right. So, the onboard RAID zero with two drives

42
00:03:20,959 --> 00:03:27,280
actually scores better overall than the four drive RAID zero array. But

43
00:03:25,840 --> 00:03:31,840
we don't really understand why that is until we go a little bit deeper into the

44
00:03:28,879 --> 00:03:35,519
results. So, you can see here, actually, why don't I put these side by

45
00:03:34,000 --> 00:03:40,640
side so you can compare them more easily.

46
00:03:37,440 --> 00:03:42,319
Yeah, that'll do.

47
00:03:40,640 --> 00:03:46,959
Okay. Okay, so at the bottom we have our RAID zero or rather on the left we have

48
00:03:44,560 --> 00:03:50,400
our RAID zero with four drives and on the right we have our RAID zero on the

49
00:03:48,560 --> 00:03:56,640
onboard controller with two drives. So you can see the four drive RAID zero

50
00:03:52,080 --> 00:03:59,040
array cleanly beats the onboard two

51
00:03:56,640 --> 00:04:03,840
drive array which it should so it should in everything but video editing using

52
00:04:00,879 --> 00:04:07,920
Windows Live Movie Maker and Windows Media Center where this one destroys it.

53
00:04:06,159 --> 00:04:11,680
So, this looks like it's actually skewing the results overall because you

54
00:04:10,080 --> 00:04:17,199
can see as you get further down, adding music, uh, application loading, things

55
00:04:14,640 --> 00:04:21,759
like uh, antivirus, these all score dramatically better on the four drive

56
00:04:19,840 --> 00:04:27,840
RAID zero array. So, I'm going to pull out one more result here using my handy

57
00:04:24,639 --> 00:04:29,759
dandy mark reader, and that is going to

58
00:04:27,840 --> 00:04:36,560
be the two drives on the RAID controller

59
00:04:33,840 --> 00:04:40,560
rather than on the onboard. So you can see that the score is actually quite

60
00:04:38,560 --> 00:04:47,600
similar. Look at that. So it looks like our performance doesn't scale that much

61
00:04:44,160 --> 00:04:50,560
with the add-in RAID controller, the LSI

62
00:04:47,600 --> 00:04:54,320
controller by adding more drives. So there you have it. It looks like two

63
00:04:52,320 --> 00:04:57,759
drives is about all that we are able to leverage with PC Mark Vantage. I mean

64
00:04:56,240 --> 00:05:01,360
you can see the scores do improve a little bit. This is four drives. This is

65
00:04:59,759 --> 00:05:07,039
two drives, but that's pretty much it. So, thanks guys for checking out this

66
00:05:02,960 --> 00:05:09,919
video on running four Intel SSD 510 SATA

67
00:05:07,039 --> 00:05:13,039
3 6 GB per second drives against two drives. Don't forget to subscribe to

68
00:05:11,440 --> 00:05:18,039
Linus Tech Tips for more unboxings, reviews, and other computer videos.
