• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

SSD in Mac Mini if anybody is interested.....

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
F

FudgeNuggets

Member
Joined
Mar 2, 2006
Location
Gone Racing
So, I've got my OCZ Vertex 60GB drive set to arrive on Monday with a copy of OS-X 10.5 Leopard. (Supposedly I'll get a free upgrade to 10.6 Snow Leopard when it is released.)


I've got benchmarks of my system when it was stock (Yohah 1.66 Core Duo) and I've got benchmarks of it with the Merom Core2Duo 2.0GHZ

I'll be running a benchmark in OS-X 10.5 with the Vertex and one in either XP PRO or Windows 7 via Bootcamp or Parallels (haven't decided yet).

Then I'll be taking a vacation from OCF until Snow Leopard is released.

If there's genuine interest in the benches, then I will post them.
 
Yeah I'd be curious not only in benches but temp readings when you get it all ready.


- Blackstar
 
im interested, not sure why but i fancied the mac mini's. pin mod comes to mind for one of those but i would think it would get toasty in that box....
 
Evilsizer said:
im interested, not sure why but i fancied the mac mini's. pin mod comes to mind for one of those but i would think it would get toasty in that box....
Click to expand...

My [former]Mini's temps would get to 83° C folding 24/7; that made me nervous but Macaholic pointed out that such was normal. When not under 100% load the Mini would run in the low 60s.

Edit: BTW,

Fudge said:
Supposedly I'll get a free upgrade to 10.6 Snow Leopard when it is released.
Click to expand...

Really? Is that in writing?
 
ihrsetrdr said:
My [former]Mini's temps would get to 83° C folding 24/7; that made me nervous but Macaholic pointed out that such was normal. When not under 100% load the Mini would run in the low 60s.
Click to expand...


Jezz... I'd never let it get that hot. :eek: Try smcFanControl: http://www.eidac.de/
Even under max CPU and GPU usage I never get over 55C at 3000 rpm. And I still can barely hear it. ;)

Average usage at 2000 - 2500 rpm I get between 35C-40C. I'd be curious how much the SSD gives off in comparison to the regular HD.


- Blackstar
 
Last edited:
Blackstar said:
ihrsetrdr said:
My [former]Mini's temps would get to 83° C folding 24/7; that made me nervous but Macaholic pointed out that such was normal. When not under 100% load the Mini would run in the low 60s.

Edit: BTW,


Jezz... I'd never let it get that hot. :eek: Try smcFanControl: http://www.eidac.de/
Even under max CPU and GPU usage I never get over 55C at 3000 rpm. And I still can barely hear it. ;)

Average usage at 2000 - 2500 rpm I get between 35C-40C. I'd be curious how much the SSD gives off in comparison to the regular HD.


- Blackstar
Click to expand...

I no longer have the Mini(sold it) but I did use smcFanControl to bump up the fan speeds. With folding 24/7 the temps would stay in the low 80° C range with CPU load @100% for months on end, no problem.
Click to expand...
 
ihrsetrdr said:
Blackstar said:
I no longer have the Mini(sold it) but I did use smcFanControl to bump up the fan speeds. With folding 24/7 the temps would stay in the low 80° C range with CPU load @100% for months on end, no problem.
Click to expand...


Well, not for you. Hehehe. :D Poor little guy.

They are tough, but not indestructible. The CPU in the 1.83 were rated to 100C, the newer 2.0 are 90C.

Anything over 50C though and I start to cringe. But that's just me. ;)


- Blackstar
Click to expand...
 
ihrsetrdr said:
Blackstar said:
I no longer have the Mini(sold it) but I did use smcFanControl to bump up the fan speeds. With folding 24/7 the temps would stay in the low 80° C range with CPU load @100% for months on end, no problem.
Click to expand...

I use that too. With fan RPM at 1750 (still inaudible) it stays around 65-70C. When I start encoding video and using up 100% of both cores it gets up to 100C without adjusting SMC. I set it to about 3k-3.5k and it cools back down to around 75-80C.
Click to expand...
 
Ok, I'm going to break this up into a few posts so that it will be easier to read.

First, I decided to stick with Tiger and send Leopard back just in case the past upgrade model does not hold true. Also makes the speed tests more legit.

It took about an hour to do the swap. 30 minutes to clone my 60GB 5400 RPM internal to the 60GB OCZ Vertex hooked up via an SATA to USB2 converter. About 30 minutes to gut the mini and get the drive to fit properly.

Tips for anybody considering getting a mini. This machines are quite elegant and made so as well. Everything fits in TIGHT. You have to be very careful so as to not break any wires because they are all very taut. Also, I had to fool around with getting the drive aligned correctly to the SATA interface which is built into a tiny vertical daughter card and not cabled like a normal PC.

Now for those interested in temps, you'll be quite pleased. When idle and SMC set to 1750 RPMs before the swap it ran 65-70C. Now it is staying steady at 50C and 1500RPMs so the change is pretty significant. I am going to encode a DVD later and we'll see how the temps do then because that's when it really got cookin before. I'll edit this post later with those temps.

Summary:
Swap is easy but like performing delicate surgery. (Pizza cutter is best tool for opening it).

Temps at idle are 10-15C lower at 1500rpm less.
 
Now my stopwatch app benches.

With 60GB 5400 RPM drive
Boot chime to usable desktop: 34 sec
XP from start to usable desktop in Parallels 34 sec
Neo Office (Mac version of Open Office) my slowest loading app: 24 sec

With 60GB OCZ Vertex
Boot chime to usable desktop 10 sec
XP start to usable desktop in Parallels 18 sec
Neo Office 7 sec

Pretty good increase there. My apps used to bounce 2-5 times on the dock before opening, now I click on them and BAM! they're open.
 
X-Bench scores:

This is the original bone stock system with the CoreDuo 2GB RAM and 5400RPM drive

Results 93.35
System Info
Xbench Version 1.3
System Version 10.4.11 (8S2167)
Physical RAM 2048 MB
Model Macmini1,1
Drive Type FUJITSU MHV2060BHPL
CPU Test 63.96
GCD Loop 217.76 11.48 Mops/sec
Floating Point Basic 75.62 1.80 Gflop/sec
vecLib FFT 36.87 1.22 Gflop/sec
Floating Point Library 56.82 9.89 Mops/sec
Thread Test 163.94
Computation 145.62 2.95 Mops/sec, 4 threads
Lock Contention 187.53 8.07 Mlocks/sec, 4 threads
Memory Test 96.14
System 92.14
Allocate 84.44 310.10 Kalloc/sec
Fill 94.82 4610.15 MB/sec
Copy 98.31 2030.63 MB/sec
Stream 100.51
Copy 82.51 1704.20 MB/sec
Scale 86.48 1786.71 MB/sec
Add 125.26 2668.31 MB/sec
Triad 122.99 2631.14 MB/sec
Quartz Graphics Test 109.92
Line 94.84 6.31 Klines/sec [50% alpha]
Rectangle 107.02 31.95 Krects/sec [50% alpha]
Circle 105.40 8.59 Kcircles/sec [50% alpha]
Bezier 103.37 2.61 Kbeziers/sec [50% alpha]
Text 155.30 9.71 Kchars/sec
OpenGL Graphics Test 187.27
Spinning Squares 187.27 237.56 frames/sec
User Interface Test 260.10
Elements 260.10 1.19 Krefresh/sec
Disk Test 36.72
Sequential 52.23
Uncached Write 31.62 19.41 MB/sec [4K blocks]
Uncached Write 67.66 38.28 MB/sec [256K blocks]
Uncached Read 58.37 17.08 MB/sec [4K blocks]
Uncached Read 76.63 38.52 MB/sec [256K blocks]
Random 28.32
Uncached Write 9.94 1.05 MB/sec [4K blocks]
Uncached Write 72.25 23.13 MB/sec [256K blocks]
Uncached Read 64.85 0.46 MB/sec [4K blocks]
Uncached Read 87.54 16.24 MB/sec [256K blocks]
 
Here is the same with the Core2Duo T7200 2.0GHZ swap

Results 112.60
System Info
Xbench Version 1.3
System Version 10.4.11 (8S2167)
Physical RAM 2048 MB
Model Macmini1,1
Drive Type FUJITSU MHV2060BHPL
CPU Test 105.12
GCD Loop 236.07 12.44 Mops/sec
Floating Point Basic 119.14 2.83 Gflop/sec
vecLib FFT 83.74 2.76 Gflop/sec
Floating Point Library 74.18 12.92 Mops/sec
Thread Test 199.72
Computation 182.22 3.69 Mops/sec, 4 threads
Lock Contention 220.93 9.50 Mlocks/sec, 4 threads
Memory Test 114.94
System 113.60
Allocate 103.16 378.85 Kalloc/sec
Fill 115.76 5628.33 MB/sec
Copy 123.83 2557.57 MB/sec
Stream 116.32
Copy 110.37 2279.71 MB/sec
Scale 112.98 2334.22 MB/sec
Add 123.37 2627.98 MB/sec
Triad 119.45 2555.23 MB/sec
Quartz Graphics Test 143.74
Line 122.29 8.14 Klines/sec [50% alpha]
Rectangle 147.50 44.04 Krects/sec [50% alpha]
Circle 142.87 11.65 Kcircles/sec [50% alpha]
Bezier 132.18 3.33 Kbeziers/sec [50% alpha]
Text 190.00 11.89 Kchars/sec
OpenGL Graphics Test 248.35
Spinning Squares 248.35 315.05 frames/sec
User Interface Test 328.65
Elements 328.65 1.51 Krefresh/sec
Disk Test 40.13
Sequential 55.36
Uncached Write 54.46 33.44 MB/sec [4K blocks]
Uncached Write 47.00 26.59 MB/sec [256K blocks]
Uncached Read 61.49 18.00 MB/sec [4K blocks]
Uncached Read 61.14 30.73 MB/sec [256K blocks]
Random 31.47
Uncached Write 11.80 1.25 MB/sec [4K blocks]
Uncached Write 64.55 20.67 MB/sec [256K blocks]
Uncached Read 65.75 0.47 MB/sec [4K blocks]
Uncached Read 85.88 15.94 MB/sec [256K blocks]
 
and finally, with the 60GB OCZ Vertex

Results 130.28
System Info
Xbench Version 1.3
System Version 10.4.11 (8S2167)
Physical RAM 2048 MB
Model Macmini1,1
Drive Type OCZ-VERTEX 1275
CPU Test 103.34
GCD Loop 237.07 12.50 Mops/sec
Floating Point Basic 119.15 2.83 Gflop/sec
vecLib FFT 79.41 2.62 Gflop/sec
Floating Point Library 74.06 12.90 Mops/sec
Thread Test 199.58
Computation 180.20 3.65 Mops/sec, 4 threads
Lock Contention 223.64 9.62 Mlocks/sec, 4 threads
Memory Test 119.90
System 116.41
Allocate 110.69 406.50 Kalloc/sec
Fill 115.71 5626.14 MB/sec
Copy 123.55 2551.79 MB/sec
Stream 123.61
Copy 113.00 2333.90 MB/sec
Scale 114.65 2368.62 MB/sec
Add 135.81 2893.12 MB/sec
Triad 134.69 2881.32 MB/sec
Quartz Graphics Test 145.21
Line 121.03 8.06 Klines/sec [50% alpha]
Rectangle 146.67 43.79 Krects/sec [50% alpha]
Circle 142.44 11.61 Kcircles/sec [50% alpha]
Bezier 130.33 3.29 Kbeziers/sec [50% alpha]
Text 214.64 13.43 Kchars/sec
OpenGL Graphics Test 245.81
Spinning Squares 245.81 311.83 frames/sec
User Interface Test 379.88
Elements 379.88 1.74 Krefresh/sec
Disk Test 211.57
Sequential 189.13
Uncached Write 215.80 132.50 MB/sec [4K blocks]
Uncached Write 212.55 120.26 MB/sec [256K blocks]
Uncached Read 128.04 37.47 MB/sec [4K blocks]
Uncached Read 250.00 125.65 MB/sec [256K blocks]
Random 240.06
Uncached Write 83.12 8.80 MB/sec [4K blocks]
Uncached Write 379.39 121.46 MB/sec [256K blocks]
Uncached Read 2558.97 18.13 MB/sec [4K blocks]
Uncached Read 622.81 115.57 MB/sec [256K blocks]
 
The scores don't lie. Really after this swap, this machine feels like the fastest machine I've ever used even though the specs are pretty modest.
 
FudgeNuggets said:
(Pizza cutter is best tool for opening it).
Click to expand...

Really? Did the pizza cutter scratch up the bottom plastics, or tear into the metal shielding? I sanded down one edge of a 2" putty knife and used that, it worked very well.

Those are some impressive jumps in the disk benchmarks, and a ten second boot to OS X, looks like it was a great upgrade. Those minis make for fine HTPCs.
 
benbaked said:
Really? Did the pizza cutter scratch up the bottom plastics, or tear into the metal shielding? I sanded down one edge of a 2" putty knife and used that, it worked very well.
Click to expand...


Yeah I use a putty knife on all of my minis the same way. Works great.

30 minutes to swap a drive Fudge?!?! o_O That's horrible! :p 10 minutes from crack to stack. ;) You're slipping. :D



Do you have a link to the Xbench scores web page? I'd love to get a comparison page of my old mini's up. I have to say I'm surprised you kept with 10.4. I find 10.5 far faster in the Intel mini's and 10.4 better on the PPC Mini's. Are you going to try 10.5 at some point?

That Temp drop is great. Since it's an older one, does the SSD drive have an internal sensor that can be read by iStats, or are you going by the ambient and CPU temps?

Any difference in the Geekbench scores? They say it just does CPU and Ram, but I'd be curious if the results were changed at all now.



- Blackstar
 
benbaked said:
Really? Did the pizza cutter scratch up the bottom plastics, or tear into the metal shielding? I sanded down one edge of a 2" putty knife and used that, it worked very well.

Those are some impressive jumps in the disk benchmarks, and a ten second boot to OS X, looks like it was a great upgrade. Those minis make for fine HTPCs.
Click to expand...

No so much as a single scratch.
 
Blackstar said:
Yeah I use a putty knife on all of my minis the same way. Works great.

30 minutes to swap a drive Fudge?!?! o_O That's horrible! :p 10 minutes from crack to stack. ;) You're slipping. :D



Do you have a link to the Xbench scores web page? I'd love to get a comparison page of my old mini's up. I have to say I'm surprised you kept with 10.4. I find 10.5 far faster in the Intel mini's and 10.4 better on the PPC Mini's. Are you going to try 10.5 at some point?

That Temp drop is great. Since it's an older one, does the SSD drive have an internal sensor that can be read by iStats, or are you going by the ambient and CPU temps?

Any difference in the Geekbench scores? They say it just does CPU and Ram, but I'd be curious if the results were changed at all now.



- Blackstar
Click to expand...

I think I'm going to skip 10.5 and get 10.6 as soon as it comes out. I was going by ambient and CPU temps. I uninstalled iStat Pro, it annoyed me.

Here's the link to the Xbench scores, be forewarned though there is so much cheating going on there it is pathetic. People with Hackintoches running their Quad-Core machines with the Mac Mini EFI, people making Ramdisks etc....

Still my machine managed to make the front page in the Mac Mini section even without cheating.

http://db.xbench.com/compareindex.xhtml

BTW.... I'm not slipping, I'm just meticulous when it comes to my Mac. I can swap a PC drive in under a minute. :beer:

I didn't run geekbench before the swap, just xbench.

edit: I just threw iStat Pro up just for you, Blackstar and it looks like it has no temp sensor as it reads 0C

Edit2: Most of the heat must come from the CPU not the HDD because shortly after handbrake maxed out the CPUs at 100%. temps jumped up into the 90s.
 
Last edited:
You must log in or register to reply here.

Similar threads

N
Multiplier vs BCLK - which is better?
Replies
3
Views
11K
UltraTaco
UltraTaco
Fr3@k3r
New MS Server geared towards families and homes.
Replies
15
Views
1K
aftermath
aftermath
Back