The TR Podcast 111: Spandex, SLI, and a snap-together tablet

The Tech Report Podcast Date: May 6, 2012

Time: 1:10:10

Hosted by Jordan Drake

Co-Hosts: Scott Wasson, Geoff Gasior, and Cyril Kowaliski

Download:
MP3 (50.5MB) | M4A (68.8MB)

Subscribe:
RSS (MP3) | RSS (M4A)
iTunes (MP3) | iTunes (M4A)

Listen now:
 

Show notes

We’re back with a short but sweet episode that kicks off with listener mail. After answering questions about GPU upgrades and CPU/RAM pairings, Jordan polls the panel on their expectations for two upcoming (or just released) summer geek action films: The Avengers and Prometheus. Then it’s on to the award-winning Asus Transformer Pad 300, followed by the ceiling-shattering Nvidia GeForce GTX 690.

Send in listener mail, and we’ll answer on the podcast. – jdrake@techreport.com

Follow us on Twitter – ScottJordanGeoffCyrilThe Tech Report

Listener mail/tweets:


GTX 680?(0:01:29) – from Jordan:

“I’m a huge fan of the podcast. I started listening at episode 75(ish) and I enjoyed it a lot. So much so that I listened to all the previous episodes even though it wasn’t relevant anymore. Anyways, back on track, I am going to get a GTX 680 for my 16th birthday (April 30) and I was wondering if I should wait for a price drop before I get one? My parents were going to pay for half and I was wondering if the price drop worth be waiting for or should I get one now if I can find one? Thanks again for the great website and podcast. From the Great White North, Jordan”

CPU/RAM combo question (0:06:54) – from JMC:

“I WAS planning on overclocking my Corsair ram with my new Asus Sabertooth X79/3930K but after reading this I am not so sure. Would really love to hear your thoughts on the comments below.

(from Newegg product comments):

***
Pros: NOTE: Contact Intel if you’d like, this CPU should be coupled with 1600 @ 1.5v RAM and NO higher to ensure stability and CPU life.

***
Cons: Not really a con. However I just got off of the phone with Intel as I needed to RMA this product. A lot of questions were asked and answered. I found out that this CPU will NOT work, at least for very long, if you use ram that runs over 1600 MHZ and 1.5 volts. I WAS running GSkill 2133 at 1.65 volts. This, based on tech support probably killed the cpu. The ram controller is built into the CPU and will not work properly over 1600 MHZ. Also I run a system with an intel 990x. The ram required for this CPU is 1066. At $1000 you can bet I will comply no matter what my Asus mobo says I can use

Tech discussion:

    Summer movie expectations – (0:11:08)

    Asus’ Transformer Pad 300 tablet – (0:23:52) – Read more

    Nvidia’s GTX 690 – (0:41:10)

That’s all, folks! We’ll see you on the next episode.

Comments closed
    • dashbarron
    • 7 years ago

    The decision with the 300 comes down to rather the $100(ish) difference between it and the Infinity is worth the purchase, plus the extra wait time. — I really enjoyed your input on the subject Scott.

    Also, agreed Dark Knight style is better, but you won’t see Avengers? Geesh Geoff, tough cookie.

    • ludi
    • 7 years ago

    RE: Asus/Hasbro (@30:30): The suit was, in fact, thrown out entirely:

    [url<]http://www.ubergizmo.com/2012/03/asus-beats-hasbros-transformer-prime-lawsuit/[/url<]

      • Damage
      • 7 years ago

      Bad reporting. One of us, ahem, fell for it, too. See the update here:

      [url<]https://techreport.com/discussions.x/22706[/url<]

    • cynan
    • 7 years ago

    Regarding the response to the intro question about overclocking memory on the X79 platform. I’m not sure that the DIMM voltage, as discussed in the response, is entirely relevant. Doesn’t the DIMM voltage specify the voltage at which the ram DIMMs are powered at? If the ram itself is resistant to clock speeds, then one can try increasing VDIMM up to 1.65V, etc, but this has little to do with the actual memory controller.

    For the on-CPU memory controller, isn’t VTT and VCCSA more relevant as these (to my limited understanding) specify what voltages the actual memory controller is run at? There has been some conflicting information floating around, but I think Intel recommends keeping VTT at 1.2V and under and VCCSA at 1.1V and under.

    [i<]Edit:[/i<]. So you guys were right: VDIMM does impact the actual memory controller somehow... Just not sure how VTT, VCCSA and VDIMM all interact to create an optimal operating environment for the memory controller. Does Intel have docs that explain this in plain English? I know that with my X79 and 3930K, when I have more than one DIMM per channel (ie, more than 4 DIMMs) installed, running much above 1600Mhz is a no-go (won't even boot Windows 7). I can cheat a bit up to 1700 MHz max by increasing the BCLK (FSB) to 106, though I keep it at 104 and lower for best stability. This is very much a limitation of the memory controller in the 3930k (and perhaps a little to do with the power delivery on the motherboard as well). Relaxing timings from 8-8-8-24 to 11-11-11-28 and increasing VDIMM from 1.4V to 1.6V doesn't do a thing. Yet with only 4 sticks of the same ram (vs 8), I can boot into Windows at 2133MHz. Anyway, just wanted to add that at least my X79 3930K setup doesn't like going past DDR-1600 when I have all DIMM slots filled.

    • odizzido
    • 7 years ago

    One thing about the delayed frames on SLI…….if you do delay a frame in the buffer to make delivery more consistent, all you are going to do is cause inconsistent “time between what actually happens in the game and what shows up on the screen” times.

    For example carmak was talking about how games are made to be Xms games, and rage was designed to be 16.6 or whatever. With SLI frame delay those lag times will keep changing from like 20ms to 28, 20, 28, 20, 28. Not only that, but the animation will speed up and slow down in exactly the same manner.

    Whether it is best to do that or not probably depends on the type of person you are. Some people don’t use Vsync and some do.

    • Hirokuzu
    • 7 years ago

    From a quick look at the AMD FAQ on Eyefinity…
    [url<]http://www.amd.com/us/products/technologies/amd-eyefinity-technology/how-to/Pages/faqs.aspx[/url<] It seems that the Eyefinity configurations need to use DisplayPort, which may explain why there were no issues with setup during the testing. Maybe it's simply an issue of nVidia and AMD deciding on different "must have" output cables for their multi-screen setups.

      • Damage
      • 7 years ago

      The Eyefinity config was happy to mix DVI and DP–and to have all three displays connected to a single card, even with a second one installed. Given the connector payload on the cards, it “just worked” for me. The GeForce was more particular, although “just use DVI” is a very helpful rule.

    • swampfox
    • 7 years ago

    Download links are currently to #110, not #111. 🙁

      • jdrake
      • 7 years ago

      Fixed! Thanks

    • jensend
    • 7 years ago

    A prybar and a crate?

    I didn’t imagine a video card would ever fail the Old Man Murray Start To Crate review system…

      • d0g_p00p
      • 7 years ago

      You Sir get a thumbs up for this post. Old Man Murray was one of the best sites ever.

Pin It on Pinterest

Share This