Intel Rapid Storage Technology

[Auric]

Recent WHQL versions (13.6.2.1001 and 13.6.3.1001) are presenting SATA/SCSI devices as ejectable. The last with normal behaviour is 13.6.0.1002. This is on Intel 9-series and AMI BIOS with Hot-Plug enabled. I use HotSwap! for this function and it works better -i.e. can present only desired device types, spindown for safe removal, and scan for for newly attached drives. Whereas presenting all drives under "Safely Remove Hardware and Eject Media" tray icon unnecessarily clutters frequent USB operations.

Have you found the same with your hardware and is there a way with Windows 7 to revert to the previous behavior using such recent drivers?

 

[BonzaiDuck]

Recent WHQL versions (13.6.2.1001 and 13.6.3.1001) are presenting SATA/SCSI devices as ejectable. The last with normal behaviour is 13.6.0.1002. This is on Intel 9-series and AMI BIOS with Hot-Plug enabled. I use HotSwap! for this function and it works better -i.e. can present only desired device types, spindown for safe removal, and scan for for newly attached drives. Whereas presenting all drives under "Safely Remove Hardware and Eject Media" tray icon unnecessarily clutters frequent USB operations.

Have you found the same with your hardware and is there a way with Windows 7 to revert to the previous behavior using such recent drivers?

Funny you just now posted this. I'd made some observations this morning parallel to those issues per a thread about "Samsung RAPID." Also mentioned Hot Swap!.

The discussion had turned to whether or not IRST was necessary for anything. For AHCI mode, this leads to a question as to whether MSAHCI or the Intel AHCI drivers are "better." I'm fine with the MSAHCI drivers, but it's interesting to know about these newer IRST-version features you mention.

Generally, I don't have a problem with any "clutter" per hot-plug/hot-swap SATA and my USB devices per "safely remove." I can see how it might complicate things, but not for me.

 

[Fernando 1]

For AHCI mode, this leads to a question as to whether MSAHCI or the Intel AHCI drivers are "better." I'm fine with the MSAHCI drivers, but it's interesting to know about these newer IRST-version features you mention.

The Intel RST drivers from v11.2.0.1006 up give Intel 6-9 Series Chipset systems running in AHCI mode a much better performance than the MS AHCI drivers.
Recently I have done some intensive benchmark comparison tests with my Z68 and Z97 systems. You can see the results >here<.

 

[BonzaiDuck]

The Intel RST drivers from v11.2.0.1006 up give Intel 6-9 Series Chipset systems running in AHCI mode a much better performance than the MS AHCI drivers.
Recently I have done some intensive benchmark comparison tests with my Z68 and Z97 systems. You can see the results >here<.

For some reason, I couldn't find any summary throughput numbers or graphs in that link.

I'd have to ask whether I need to update the BIOS of my Z68 system to use these newer RST drivers. Then I could reinstall IRST.

I remember this issue comparing MSAHCI to the Intel [forgot-what-they're -called] drivers -- we'd discussed sometime earlier last year.

There are other cases where the MSAHCI driver is the only driver I can use: for instance, with a PCI-E controller that operates in AHCI mode with the native driver, but would require its proprietary driver for RAID. That particular example is one case where I'm definitely not interested in any change.

These two Sandy Bridge workstations -- maybe more flexible. But with SSDs operating in AHCI with both RAM-caching and SSD-cached HDD, I don't have much incentive to swap drivers.

 

[Charlie98]

The Intel RST drivers from v11.2.0.1006 up give Intel 6-9 Series Chipset systems running in AHCI mode a much better performance than the MS AHCI drivers.
Recently I have done some intensive benchmark comparison tests with my Z68 and Z97 systems. You can see the results >here<.

Your stated conclusion in the Z68 results...

Evaluation of the benchmark results:

• Most surprisingly the old "classical" Intel RST driver v11.2.0.1006 gave my Z68 system the best READ, WRITE and OVERALL performance even while runnig in AHCI mode.
• The generic MS AHCI driver seems to be not a good choice for Intel Z68 systems running in AHCI mode.
• The benchmark result differences between the 8 tested Intel RST(e) AHCI drivers are minimal and probably unobservable while working.

The v11.2.0.1006 driver you listed shows up as only being for Vista and XP... and you are running it OK on a W8.1 OS? (assume 64-bit.)

Do you think running the latest driver is a better idea? I think I'm currently running v12.8.0.1016 without issue... but I don't think it's the correct one.

 

[Fernando 1]

For some reason, I couldn't find any summary throughput numbers or graphs in that link.

Only registered Forum members can see the pictures, which are directly attached to the related posts.

I'd have to ask whether I need to update the BIOS of my Z68 system to use these newer RST drivers.

There is no BIOS update required, if you want to use newer Intel RST drivers in AHCI mode.
The only restriction is, that the Intel RST drivers v12.9.4.1000 WHQL are the latest ones, which natively do support Intel 6-Series Chipsets running in AHCI mode. Nevertheless it is possible to get the newest Intel RST drivers v13 or even v14 series installed onto Intel 6-Series chipsets, if the required HardwareIDs have been added to the related INF files. I am offering such already modified and nearly universally usable Intel RST drivers within >this< thread.

There are other cases where the MSAHCI driver is the only driver I can use: for instance, with a PCI-E controller that operates in AHCI mode with the native driver, but would require its proprietary driver for RAID. That particular example is one case where I'm definitely not interested in any change.

The Intel SATA Controller has nothing to do with PCIe Controllers. Intel RST drivers can only manage drives, which are connected to the Intel SATA ports and running in AHCI or RAID mode. As long as you have SATA drives connected to Intel SATA ports and PCIe/M.2 ports, it is no problem to run the Intel RST driver (managing the Intel SATA ports) and the MSAHCI driver (managing the PCIe/M.2 slots) simultaneously,

The v11.2.0.1006 driver you listed shows up as only being for Vista and XP...

You obviously misunderstood that. The Intel RST driver v11.2.0.1006 is the last one, which can be used with all Windows Operating Systems from Windows XP up. None of the later released Intel RST drivers has the file named TXTSETUP.OEM, which is required for the XP installation in AHCI or RAID mode.

and you are running it OK on a W8.1 OS? (assume 64-bit.)

Yes, my tests were done with the OS Win8.1 x64. There is no restriction of the usability of Intel's RST drivers with the latest Windows Operating Systems (incl. Win10).

Do you think running the latest driver is a better idea? I think I'm currently running v12.8.0.1016 without issue... but I don't think it's the correct one.

The latest Intel RST drivers, which are running fine and natively do support your Intel 6-Series Chipset system, are the v12.9.4.1000 WHQL ones. You can find the download links >here<.

 

[BonzaiDuck]

Well, here's a choice I would face in reverting to IRST over the native MS drivers. You tell me that I'd now have the same hot-swap answer to "Safe Removal," but through the IRST systray, dialogs and options. With MSAHCI, I don't have the extra systray item; the safe-removal feature coincides with the safe-removal icon also serving USB devices. At that point, the MSAHCI route seems simpler.

So it boils down to just how much performance improvement I can expect, or whether it's worth it.

I DO think I'd need to update the BIOS for my sig-rig, while the twin system is the later-gen Z68 board with the very last BIOS. For the sig-rig, an update in the IRST software actually required a BIOS update. I could easily test the look and feel of the IRST on the second rig.

For the PCI-E cards, the ones I'd bought were promoted with the idea of using the native MSAHCI, and obviously there's not IRST option for a Marvell chip.

So for the 2700K system with the Gen3 motherboard, latest BIOS etc. -- a computer which actually has "daily use" -- it's a matter of convenience or when I want to take time to install the IRST software and make the driver changeover. And that leaves me with the other factors I mentioned about the extra systray item, the dialogs and so forth.

We'll just have to "wait and see." On the sig-rig, I'd already had the IRST installed, and I uninstalled it after switching over to MSAHCI.

 

[Fernando 1]

@ BonzaiDuck:

1. There is no reason to install the IRST software, since you are running your system drive in AHCI mode. You will not have any benefit from it, but the IRST Service, which will run in the background, may drop your performance. All you have to do is to manually "update" the driver of your on-board Intel SATA AHCI Controller from within the "IDE ATA/ATAPI Controllers" section of the Device Manager. So you should download just the desired Intel RST driver and not the complete "Intel RST Driver & "Software" package.
2. If you should not be satisfied with the "Safely Remove Hardware" handling by the Intel RST driver, I recommend to install the small tool named "HotSwap!". You can get it >here<.
3. A BIOS update is not required at all, because it doesn't contain any Intel AHCI ROM or Firmware. Only for Intel RAID users I recommend to update the Intel RAID ROM module, which is within the mainboard BIOS (it is not used at all by a system, which is running in AHCI mode).

 

[john3850]

In the past I had to use the bloated IRST software to just install the intel drivers in AHCI mode.
Thanks for the quick modded inf Fernando 1.
727 as-ssd score dirty drive msahic
967 as-ssd score dirty drive iaStorA

 

[vailr]

@ BonzaiDuck:

1. There is no reason to install the IRST software, since you are running your system drive in AHCI mode. You will not have any benefit from it, but the IRST Service, which will run in the background, may drop your performance. All you have to do is to manually "update" the driver of your on-board Intel SATA AHCI Controller from within the "IDE ATA/ATAPI Controllers" section of the Device Manager. So you should download just the desired Intel RST driver and not the complete "Intel RST Driver & "Software" package.
2. If you should not be satisfied with the "Safely Remove Hardware" handling by the Intel RST driver, I recommend to install the small tool named "HotSwap!". You can get it >here<.
3. A BIOS update is not required at all, because it doesn't contain any Intel AHCI ROM or Firmware. Only for Intel RAID users I recommend to update the Intel RAID ROM module, which is within the mainboard BIOS (it is not used at all by a system, which is running in AHCI mode).

I could be wrong, but weren't the Intel Raid ROM and Intel AHCI ROM combined into one single firmware within the motherboard bios, at least in Z68 and older motherboard bioses? Maybe you could elucidate the finer points in the history of various Intel Raid and AHCI versions: which chipsets included specific ROM versions at launch, and which are capable of being updated to newer ROM versions, even though the motherboard makers have let such updates mostly relegated to being abandoned (and not officially updated).

 

[Fernando 1]

weren't the Intel Raid ROM and Intel AHCI ROM combined into one single firmware within the motherboard bios, at least in Z68 and older motherboard bioses?

No, they have never been combined into one single Firmware.
Old Intel Chipset (up to P45/X58) mainboard BIOSes usually have an Intel RAID ROM (only needed and used in "RAID" mode) and additionally an Intel AHCI ROM (latest version: v1.20E, only usable, when the Intel SATA Controller has been set to "AHCI"), but non of the newer Intel Chipset (from 5-Series up) mainboard BIOSes contain an Intel AHCI ROM. It is not required, because the modern Intel SATA AHCI Controllers natively do support the fast boot in AHCI mode.

 

[BonzaiDuck]

@ BonzaiDuck:If you should not be satisfied with the "Safely Remove Hardware" handling by the Intel RST driver, I recommend to install the small tool named "HotSwap!". You can get it >here<.

And . . . that's sort of "back to square one," since I'd been compelled to use HotSwap! before I . . .ah . . . made the "swap" to MSAHCI.

I think the latest version of the freeware was version 6.xx. I probably have it still in my server's "software" directory.

I'm just not so sure to be eager to make a change. I suppose I could simply be sure to establish a checkpoint, and drop back to it as the easiest way to revert to "the way things are."

It's always true what my old friend the Swiss chef-turned-psycho-pharmacologist told me about gourmet preparations: "Simple is best." There are just many flavors of "simple." ["Flavors." That's funny . . ]

 

[BonzaiDuck]

I'm going to post this as a "double" or sequential post to the one that precedes, instead of simply revising the latter.

While I'm not planning to go through the hoops Fernando says are necessary to review the data on the driver comparisons, I've reviewed the old thread from last year in which such comparisons were posted.

In that thread, it did not appear that the differences -- measurable as they are -- between the MSAHCI driver and the Intel IASTOR AHCI driver -- are significant enough to get your panties in a bunch over a driver-swap.

For instance, a 20 MB/s difference for a spec-rated 500 MB/s SSD is insignificant. It would be more significant for a high-end HDD with performance somewhere between 150 and 300 MB/s; it would be even more significant for an SATA-II drive operating at <= 150 MB/s.

However. For such HDDs that are either cached to SSD or to RAM or even both -- it is still piddly-squat. RAID doesn't enter the picture here, because we're essentially comparing one AHCI driver to another.

I think this is one of those "gray-area" issues. Based on the raw numbers, users can choose the best-performing driver. Based on convenience or any number of other parameters, the native driver is probably just A-OK and fine.

 

[ignatzatsonic]

I was bored and inquisitive enough to play around with this a bit yesterday.

I have a Gigabyte H67 board (GA-H67A-UD3H-B3) and a Crucial MX100 SSD (128 GB) on Windows 7 Home Premium 64 bit, using AHCI.

I've been using Intel RST driver version 10.0.0.08 for several years with no thought of updating. I've got an AS SSD benchmark of that.

Yesterday, I removed the Intel RST and tried the Microsoft AHCI drivers and ran a new benchmark.

Then I went to Fernando's site and saw that his favorite RST for my chipset was version 12.9.4.1000. So I changed to that and ran another benchmark.

AS SSD scoring:

Microsoft AHCI: 372 read, 197 write, 751 overall
RST 10.0.0.08: 388 read, 216 write, 795 overall
RST 12.9.4.1000: 387 read, 212 write, 789 overall


For what it&#8217;s worth.

 

[Fernando 1]

I think this is one of those "gray-area" issues. Based on the raw numbers, users can choose the best-performing driver. Based on convenience or any number of other parameters, the native driver is probably just A-OK and fine.

What do you mean with the "native" driver?
The generic Win7/8/8.1 AHCI drivers named MSAHCI.SYS resp. STORAHCI.SYS are universally usable AHCI drivers, which support the SATA AHCI Controllers of various chipset manufacturers (Intel, AMD, Marvell, ASMedia etc.), whereas the Intel RST drivers have been especially developed and designed for being used with Intel SATA AHCI Controllers.
That is the simple reason why I recommend to use the "specialist" and not the "allrounder" under the available AHCI drivers, especially for Intel chipset systems.

 

[BonzaiDuck]

What do you mean with the "native" driver?
The generic Win7/8/8.1 AHCI drivers named MSAHCI.SYS resp. STORAHCI.SYS are universally usable AHCI drivers, which support the SATA AHCI Controllers of various chipset manufacturers (Intel, AMD, Marvell, ASMedia etc.), whereas the Intel RST drivers have been especially developed and designed for being used with Intel SATA AHCI Controllers.
That is the simple reason why I recommend to use the "specialist" and not the "allrounder" under the available AHCI drivers, especially for Intel chipset systems.

Yes -- MSAHCI versus IAStor was the comparison from last year's thread.

I had even seen elsewhere the assertion that the MSAHCI driver had actually been an Intel product, but cannot confirm.

 

[Fernando 1]

Yes -- MSAHCI versus IAStor was the comparison from last year's thread.

Yes, and all tested Intel RST AHCI driver versions were better performant than the "universally usable" Win7/8 in-box Microsoft AHCI driver.

I had even seen elsewhere the assertion that the MSAHCI driver had actually been an Intel product, but cannot confirm.

As the name of the MSAHCI driver already indicates, it has been developed and released by Microsoft. Intel has nothing to do with the Win7/8/10 in-box AHCI driver (contrary to the Win7/8/10 in-box Intel RAID driver named iaStorV.sys resp. iaStorAV.sys, which has been made by Intel).

 

[BonzaiDuck]

Yes, and all tested Intel RST AHCI driver versions were better performant than the "universally usable" Win7/8 in-box Microsoft AHCI driver.
As the name of the MSAHCI driver already indicates, it has been developed and released by Microsoft. Intel has nothing to do with the Win7/8/10 in-box AHCI driver (contrary to the Win7/8/10 in-box Intel RAID driver named iaStorV.sys resp. iaStorAV.sys, which has been made by Intel).

So? I haven't even questioned your assertions about the Intel drivers. I assume they are true. But, like I said, the comparisons were posted in the forums last year. The really significant differences -- Intel over MSAHCI -- were about 20+ MB/s in the sequential tests. It couldn't have made a hill-a-beans difference in day-to-day operation. Nor has there ever been a stability problem, problems with the disks, etc.

AND! I'd used the Intel AHCI drivers, but annoyed at the problem we discussed, the extra bloatware, the need for HotSwap! freeware, etc. So why so insistent that I NEED to change MY drivers?!

Nobody is contesting your point, nor your many tests on the many versions of IRST.

Just to be fair -- not wanting to disable my RAPID-mode on my Sammy 840-Pro boot disk today -- I've run the Magician benchmark on the Crucial MX100 connected to the other SATA-III port on my sig motherboard.

Not going to take the time with a 'screenie-capture', but no reason to fudge the results either:

556 MB/s seq read
507 MB/s seq write
65,338 IOPS random read
67,940 IOPS random write

What DIFFERENCE does it make to me, to swap the Intel driver back into the mix? Because -- I'll never surpass 600 MB/s on any of those benchmarks. It's not in the cards; it's not in the hardware to do it. Tell me -- I want to know! Curious minds want to know!

Now if need be, I can probably run the ATTO or other benchie, but this should be enough! Enough, I say!

 

[Fernando 1]

But, like I said, the comparisons were posted in the forums last year. The really significant differences -- Intel over MSAHCI -- were about 20+ MB/s in the sequential tests.

The sequential read/write scores are much less interesting for the "normal" user during his daily work than the 4K read/write scores.

AND! I'd used the Intel AHCI drivers, but annoyed at the problem we discussed, the extra bloatware, the need for HotSwap! freeware, etc. So why so insistent that I NEED to change MY drivers?!

Nobody forces you to switch to the Intel AHCI driver.
By the way: When I am running my system in AHCI mode, I install just Intel's AHCI driver and nothing else (no Intel RST Software and no additional tool like HotSwap!).All my drives except the system drive are automaticly listed as "Safely Removable".

 

[BonzaiDuck]

The sequential read/write scores are much less interesting for the "normal" user during his daily work than the 4K read/write scores.
Nobody forces you to switch to the Intel AHCI driver.
By the way: When I am running my system in AHCI mode, I install just Intel's AHCI driver and nothing else (no Intel RST Software and no additional tool like HotSwap!).All my drives except the system drive are automaticly listed as "Safely Removable".

I'm still "toying" with the driver-swap idea, and now I'm going to run another test to see about the random reads and writes. I just don't think it's essential to performance of these systems. What I might lose for a driver choice is made up with the "gimmicks" -- RAM caching, HDD-to-SSD caching -- even RAM-caching for the latter as "whole enchilada."

LATER: just ran AS SSD, and revisited last year's thread. 4K random tests, not so good. 4k 64Thrd -- more acceptable. But that's on the uncached Crucial MX100. Looking more carefully at this . . .

 

[BonzaiDuck]

OK. O . . . KAY.

On my 2700K system, I disabled RAPID and disabled SSD-caching of my HDD with Primo. [Eventually, I'll simply replace RAPID with comprehensive Primo-cache for both RAM caching SSDs and an SSD caching my HDD. That's no issue here; we're more interested in raw performance on disks.]

There's no reliable basis for comparison with the test I'd run earlier on the 2600K's Crucial MX100. But the Crucial's 4K test under AS SSD came in at around 26 MB/s random read rate -- using the MSAHCI driver. The 2700's test for the 840 EVO boot-disk under Crystal Disk Mark showed about 35 MB/s with the Intel 12.9.4.1000 AHCI driver.

Now it's probably true -- the smaller differences in benchmarks on the SAME disk, same system and SAME TEST programs between the two drivers might mean noticeably better performance for the Intel driver and the situations Fernando quoted as "mainstream user profile."

My results -- with two different test programs and SSDs of different manufacture are only CONSISTENT with the basic premise of this thread. If I were to accept the 9 MB/s as representative of the stringent conditions I posed in the last paragraph, I could say that it doesn't SEEM like a very big performance difference. YET it has been argued that these smaller differences for 4K tests are indicative of significantly better performance.

While nothing I've produced here is conclusive of anything, it's vaguely consistent with the premise. And swapping drivers is pretty simple. I'll still have to test the hot-swap functionality though. Not a major priority today, but nothing that would take much time.

I promise not to get into a pissy-fit if I'm left displeased for hot-swapping a caddy into my bay.

 

[Towermax]

The Intel RST drivers from v11.2.0.1006 up give Intel 6-9 Series Chipset systems running in AHCI mode a much better performance than the MS AHCI drivers.

I just installed the v11.2.0.1006 drivers on my Z68 board (had an earlier version installed, not MSAHCI), and got about a 4% performance boost on my primary SSD. Thanks for the info.

 

[BonzaiDuck]

OK backwards . . .

I AM sort of in a pissy-fit now.

First, in order to swap out a caddy from my bay, I'll have to install the Hot Swap! freeware. Pissy-fit #1.

Second, IF I IMPLEMENT MY Primo-Cache SSD-caching, the benchies don't show the usual scores. Primo SHOWS the hit-rate for the drive -- OK. But the benchies, if they reflect anything, show LOWER than expected scores just for the standalone HDD. They actually might be HIGHER if I disable the cache.

So I'm still on the FENCE as I was before, between the MS driver and the Intel driver. And I THINK that the caching program may actually bridge any shortfall with the MS driver.

You could argue "don't use that caching program!" I'd only follow that advice if I weren't going to use any electro-mechanical drive in my box. The caching program provides the same benefit that ISRT provides in its storage-mode-limited way, without driving me toward RAID-mode in BIOS.

I can GUESS that it's working properly, and it seems to be -- according to the caching program with its presentation of the cache hit-rate. But while Magician shows the expected results for RAPID, CrystalDiskMark does NOT for the Primo program, the cache-SSD and the accelerated HDD.

All of these things are annoyances for my preferred configuration. So I have to ask if a possible 10 MB/s improvement in the 4K test is worth it, since the ACTUAL improvement with the caching program is much better -- with or without the Intel driver. AND -- I can't see the BENCHMARKED performance with that driver, even if I KNOW that it's there! I can only "SEE" it in the caching-program's hit-rate presentation.

ADDENDUM: At least a couple people had steered toward at least trying the Primo-Cache program -- indicated in posts on the thread about RAPID (which see), or in PM's to me. I've only suggested that people can try it, and if they don't like it, they have that choice too.

So I only SUSPECT that, while Primo works properly with Intel's AHCI driver, it's "working" shows up properly in benchmarks with the MS driver -- no less than RAPID performance shows up in Samsung benchmarks. Of course caching "works," because the user has automatically moved data from a slower device to a faster one, for much faster subsequent access. But the benchies don't show it for Primo combined with Intel AHCI.

But what would you expect? Two different big-hitter hardware manufacturers require you to use either RAID or AHCI but not "either" in their bundled caching solutions. A third party comes up with a universal software solution that is a sort of Swiss Army knife of caching. It works with the MS driver; it works with the Intel driver. But with the latter, it doesn't "show itself" in other independent benchmark programs, even though it's own hit-rate reporting shows it working.

I'm willing to bet that the caching solutions make the difference in drivers a "wash." You could only measure it with benchmarks, though. And caching gets attacked with charges that "it only gets better benchmarks." But that doesn't really make sense, because we know what caching does -- utilizing different hardware layers.

That's all I have to say in the matter. If you want to understand why certain individuals in certain political parties get elected President for more terms or term-years than the field of congressional districts would suggest for their respective home-states -- you look at the concentrated industries with an imposing presence in those states, and it will explain to you why someone would say "bring democracy to Iraq" while pulling a fast one with Halliburton's sole source contract, the "WMDs story" and the world's third largest oil reserve. If you want to understand how or why certain software anomalies occur in the situations we've discussed -- look at the market power for dominant firms who produce the controllers, the drives and their own software solutions.

Oddly, in all this -- you've got another "dominant firm" in the software market: Microsoft. And -- really -- they've got to be "software agnostic" and hardware-impartial. They provide an OS and software platform for other program solutions, and they're caught in the middle.

I'm probably going to stick with the MSAHCI program until I stop using caching solutions like Primo.

 

[BonzaiDuck]

RETRACTION:

Apparently I made premature conclusions.

My problem with Primo isn't the Intel Driver.

It is either Crystal Disk Mark, or it is Primo.

Whether MSAHCI or the Intel AHCI driver, SSD-cached HDD scores -- just for the sequential read/write test -- are in the 118 to 120 MB/s range, for a WD Blue drive connected to an SATA-II port. This would've shown up with a similar configuration using ISRT in benchies showing 80% of the SSD standalone speed.

And it is interesting. With HDD RAM-CACHING under Primo, the benchies rise as expected with the caching, the more RAM you use for it, and the scores were taken under CrystalDiskMark.

Hmmm. . . . WELL!! I'd rather not add the HotSwap! program just because I get a bump in 4K speeds with the Intel driver. And I'm still sitting on the fence with this.

Sorry to trouble anyone with my screed in the last post, though. Thought I was "on to something."

Actually, it occurs to me that the "L2" feature of HDD-to-SSD-caching was a sort of ancillary feature of Primo -- promoted more for its RAM-caching. So that might explain some shortcoming in making the SSD/HDD combination appear as a single drive to the benchmarks. Interesting . . .

 

[mikeymikec]

Hopefully not too OT a question and quickly answered - what's the latest IRST driver that works on Win7? I've tried to use the latest driver on a few hardware platforms running Win7 and it always tells me that the platform is not supported.