Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Matt]

Who will post the first benchmark results running a new Sandy Bridge CPU?

If you get one, let us know how it does in the Media Center benchmark.  I'm curious.

Thanks.

[fitbrit]

I was planning to upgrade one of my HTPCs to SandyBridge, but an article I read stated that there's still an incompatibility between MediaPlayerClassic and the CPUs, and possibly bitstreaming audio via ffdshow as well.
Since I don't need all that power, I think I'm going to just build with another i3 for now - hopefully cheaper now that SandyBridge is here. I'd be curious to see the benchmarks though.

[Matt]

I was bad.

i7 920 @ 3.3:

Code: [Select]

=== Running Benchmarks (please do not interrupt) ===

Running 'Math' benchmark...
    Single-threaded integer math... 3.688 seconds
    Single-threaded floating point math... 2.076 seconds
    Multi-threaded integer math... 1.122 seconds
    Multi-threaded mixed math... 0.774 seconds
Score: 2481

Running 'Image' benchmark...
    Image creation / destruction... 0.814 seconds
    Flood filling... 0.501 seconds
    Direct copying... 0.590 seconds
    Small renders... 2.259 seconds
    Bilinear rendering... 1.259 seconds
    Bicubic rendering... 1.128 seconds
Score: 3359

Running 'Database' benchmark...
    Create database... 0.516 seconds
    Populate database... 1.442 seconds
    Save database... 0.131 seconds
    Reload database... 0.038 seconds
    Search database... 1.296 seconds
    Sort database... 1.018 seconds
    Group database... 1.014 seconds
Score: 3941

JRMark (version 16.0.6): 3260

Sandy Bridge 2600K @ 4.8:

Code: [Select]

=== Running Benchmarks (please do not interrupt) ===

Running 'Math' benchmark...
    Single-threaded integer math... 2.838 seconds
    Single-threaded floating point math... 1.876 seconds
    Multi-threaded integer math... 0.901 seconds
    Multi-threaded mixed math... 0.699 seconds
Score: 3009

Running 'Image' benchmark...
    Image creation / destruction... 0.608 seconds
    Flood filling... 0.547 seconds
    Direct copying... 1.113 seconds
    Small renders... 1.592 seconds
    Bilinear rendering... 0.834 seconds
    Bicubic rendering... 0.751 seconds
Score: 4040

Running 'Database' benchmark...
    Create database... 0.347 seconds
    Populate database... 0.931 seconds
    Save database... 0.090 seconds
    Reload database... 0.025 seconds
    Search database... 0.789 seconds
    Sort database... 0.718 seconds
    Group database... 0.637 seconds
Score: 6080

JRMark (version xx.x.xx): 4376

The Sandy Bridge also draws about 30 watts less at idle with all the same components (measured at the wall).

[newsposter]

the expanded matrix math functions in SB could in theory be useful for transcoding but the s/w needs to be written to take advantage of them.

[SiGuy]

I've built a new HTPC using Sandy Bridge, 2600K, and an ASUS P8H67M-EVO motherboard.

JRMark is 3360 with no CPU overclocking (Stock=3.4/3.8GHz), DDR3 1333, and a SATA-II SSD.

I can't seem to get Image playback (photos from my network) to playback in an MC15 slideshow.  I see the image title, and the titles change at the 5second image interval I use.  But other than the image title, my screen is black.  I've tried manipulating the GPU memory buffer size (from 32MB to 128MB), and over/under clocking the GPU, but none of those work. 

Other programs will show slideshows of the same images, including Windows7 Explorer & MC14.0.165.  So this problem seems to be unique to MC15 and appears to be an MC15 bug.

I've tried MC15.0.172 & .173, both with the same result.  The OS & Apps installation I'm using is an in-place upgrade from a working Core2/Quad system using an NVidia graphics card which did not show this issue.

With MC15, I've tried two types of display, VGA LCD monitor, and an HDTV via HDMI.  Both produce the same black-screen results. 

MC15 is operating, escape exits full-screen mode.  Music playback is fine.  Internet browsing is fine.  DeviceManager shows no unknown hardware.

My full system configuration is below.  You can see that MC is recognizing my Core/i7 2600K as a Pentium-III.  That's a simple bug and hopefully not significant:


Media Center 15.0.173 Registered -- C:\Program Files (x86)\J River\Media Center 15\

6.1  (Build 7600)
Intel Pentium III 3402 MHz MMX / Memory: Total - 4194 MB, Free - 3145 MB

Internet Explorer: 8.0.7600.16385 / ComCtl32.dll: 5.82.7600 / Shlwapi.dll: 6.1.7600 / Shell32.dll: 6.1.7600 / wnaspi32.dll: N/A
Ripping /   Drive E: _NEC    DVD_RW ND-3520A   Mode:ModeSecure  Type:Auto  Speed:Max
  Drive F:   Mode:ModeSecure  Type:Auto  Speed:Max
  Digital playback: Yes /  Get cover art: Yes /  Calc replay gain: Yes /  Copy volume: 100
  Eject after ripping: No /  Play sound after ripping: No 

Burning /  Drive E: HL-DT-ST BD-RE  UH10LS20    Addr: 0:0:0  Speed:48  MaxSpeed:48  BurnProof:Yes
  Test mode: No /  Eject after writing: Yes /  Direct decoding: Yes /  Write CD-Text: Yes
  Use playback settings: No /

Portable Device Info
  Removed devices: 4G_CF#1_GA,Generic STORAGE DEVICE (I:)


Interface Plugins:
  TiVo Server (Active)

[JimH]

I believe we saw this problem about three weeks ago and it turned out to be a video driver problem.  Removing the driver and letting Windows re-install it fixed it.

[Scolex]

This JRMark you speak of where can I find it?

[DarkPenguin]

help->benchmark, I think.

[SiGuy]

Thanks, Matt.

Re-installing the Intel Graphics driver did resolve this problem.  It does seem odd to me that only MC15 (not MC14 or other apps) showed this problem.  But I appreciate the quick response!

Scolex, see Help-->Benchmarks from MC15.  You can find the JRMark there.

[Scolex]

Thanks

I was curious how my 3.5 year old Core 2 DUO E6600 2.4@3.0, Asus P5N32E-Sli, BFG 8800GT-OC2 and 8GB of OCZ 1066 DDR2 stacks up against an i7 920 and it didn't do too bad.

Code: [Select]

=== Running Benchmarks (please do not interrupt) ===

Running 'Math' benchmark...
    Single-threaded integer math... 3.977 seconds
    Single-threaded floating point math... 2.477 seconds
    Multi-threaded integer math... 3.781 seconds
    Multi-threaded mixed math... 2.363 seconds
Score: 1508

Running 'Image' benchmark...
    Image creation / destruction... 2.613 seconds
    Flood filling... 1.322 seconds
    Direct copying... 3.072 seconds
    Small renders... 3.086 seconds
    Bilinear rendering... 3.204 seconds
    Bicubic rendering... 2.891 seconds
Score: 1359

Running 'Database' benchmark...
    Create database... 0.811 seconds
    Populate database... 2.155 seconds
    Save database... 0.219 seconds
    Reload database... 0.060 seconds
    Search database... 2.111 seconds
    Sort database... 1.546 seconds
    Group database... 1.545 seconds
Score: 2545

JRMark (version 15.0.172): 1804

[glynor]

Uh-oh... Hopefully this isn't the Pentium Pro issue all over again!  Intel has found a flaw in 6-series chipsets, and has halted shipments.

Intel explains, "In some cases, the Serial-ATA (SATA) ports within the chipsets may degrade over time, potentially impacting the performance or functionality of SATA-linked devices such as hard disk drives and DVD-drives."

For folks who have already crossed the Sandy Bridge, Intel adds that it will "work with its OEM partners to accept the return of the affected chipsets," and it plans to "support modifications or replacements needed on motherboards or systems."

Ack!  Looks like you might have to send that shiny new Core 2600k system's motherboard back, Matt.

[Matt]

Ack!  Looks like you might have to send that shiny new Core 2600k system's motherboard back, Matt.

I guess it wouldn't be the bleeding edge if there wasn't bleeding.

I'm already on my second motherboard, as the first had two dead RAM slots.

[glynor]

It is better news (for Intel) than many of the headlines out there seem to suggest, though.  The issue is supposedly in the chipset (the SATA subsystem, it seems) rather than in the CPU itself.  This is probably the case, since the SATA controller is on the P67 chipset chip rather than on-die on the CPU.

So, "Sandy Bridge" doesn't really have a flaw.  The motherboards that support Sandy Bridge are flawed, though.  Conveniently for Intel, the chipset is far less "advanced" tech (and therefore cheaper to manufacture) than the CPU.  That should hold the losses from the recall down substantially, compared to what they would be if the Sandy Bridge silicon itself was bad.

From an end-user perspective, however, a CPU swap-out would be far easier and less painful than a whole motherboard swap out.  On the other hand, this probably really only applies to OEMs and enthusiasts though, as most end-users wouldn't be capable of replacing a CPU if Intel was able to drop-ship a fixed one direct.  To everyone else, you'd still have to send the whole thing back to Dell so where the problem actually lies is largely irrelevant.

Not good PR for Intel in either case.

[glynor]

Update from Intel with some additional interesting details:

The issue is a circuit design problem resulting in a gradual degradation over time of SATA connectivity on the affected ports, manifesting itself as high bit-error rates on those ports and eventually as total device disconnects.

That's a serious issue, but it's limited in scope. Intel says storage devices connected to those ports should not be damaged, and data on the devices should be intact and readable on another system.

The ports potentially affected, interestingly enough, are the four 3Gbps SATA ports on the chipset. The two 6Gbps SATA ports aren't at risk.

Same link as above.

[glynor]

I gotta say, if everything goes according to plan and other OEMs follow suit, this is the way to do a recall in a bad situation.  That "deal with your seller" thing makes me worry a bit though (if the seller isn't NewEgg).

They where honest and up-front about the issue when it was discovered (and it appears they released the news quite quickly).  They clearly explained the issue to the press.  And, they're now communicating a plan for a resolution, and the plan seems quite fair.  You can essentially return the product for a full refund, or wait and swap it out for a replacement board when they become available.

Good job, Intel.  It is too bad this happened.

[Daydream]

You can essentially return the product for a full refund, or wait and swap it out for a replacement board when they become available.

What's the word out for the B3 stepping? April? Meanwhile one doesn't have in what to install a Sandy Bridge CPU. Chipset aside, Intel can't be too happy with the CPU sales in these conditions.