[Guru3D] AMD ThreadRipper (1) MSRPs lowered!

[VirtualLarry]

Don't know if AT covered this yet, but apparently, there have been some price cuts.

Yes, TR2 is coming VERY soon, so why not lower prices on TR1? BTW, AFAIK, TR1 is still faster for gaming, too.

https://www.guru3d.com/news_story/threadripper_gen1_processor_prices_have_been_lowered.html

 

[Markfw]

The only problem is that Amazon and newegg have not seen that yet. At least my the prices as of right now.

 

[VirtualLarry]

I'm looking at an X399 ASRock Fatality Pro, and a $399 TR1 1920X CPU. I already have four RX 570 GPUs to go in it. Then, I just need a good case. Thinking of getting another Rosewill Blackhawk.

Edit: Still in the speculative phase, thinking of "if I could afford a ThreadRipper". Maybe if I sell a few lesser PCs. I could consolidate all of my mining rigs, pretty-much, in one rig.

 

[scannall]

I'm looking at an X399 ASRock Fatality Pro, and a $399 TR1 1920X CPU. I already have four RX 570 GPUs to go in it. Then, I just need a good case. Thinking of getting another Rosewill Blackhawk.

Edit: Still in the speculative phase, thinking of "if I could afford a ThreadRipper". Maybe if I sell a few lesser PCs. I could consolidate all of my mining rigs, pretty-much, in one rig.

Just do it Larry. Run with the big dogs for a bit. I love and appreciate your low end builds, but take a break for a bit.

 

[VirtualLarry]

... slowly moving UpTown. Bought three Ryzen 5 1600 rigs almost two years ago, when they came out. I was a semi-early-adopter. I saw what those CPUs could do, and thought that, although as expensive as a top-end Intel consumer chip at the time, they seemed to offer even greater value than Intel, even if they were 8% slower at 1080P gaming. Big deal. (I don't game much anymore.)

Now that ThreadRipper is proven tech, and the boards are stable, I'm probably going to move up to one "monster rig" (primarily for DC and mining, maybe some casual gaming), and move all of my other PC "stations" here to my mini-PCs. (My ASRock DeskMini units are still kickin', even with a "lowly" G4560 and G4600 CPU in them. Plenty fast for web browsing.)

Edit: This desire to move up to TR1, 12C/24T for $399, was the realization, that I was willing to move to Ryzen, 6C/12T, for $200-220,, and that this TR1 1920X deal (once it hits Newegg), made TR1 just as affordable as Ryzen was originally, and then, I can get a board with four PCI-E x16 slots (at full lanes!), along with 3x M.2 PCI-E 3.0 x4 (RAID), AND 10GbE, which I really want on my boards, moving forwards.

I realize that TR, in general, is probably worse for gaming, than Ryzen or Coffee Lake, but I primarily do DC / mining right now, and very little gaming, so that doesn't bother me a bit. Bring on the MOAR COARS baby!

 

[Topweasel]

TR1's pricing is becoming very tempting. The 1950 is more my cup of tea and not worth it yet in pricing to forgo waiting for the highest Zen2 2 chip solution (not in so dire need of cores to go with such a crippled memory solution). But it's getting close. Think I will wait it out till Q1/q2 next year, see what Zen 2 is shaping up to be and if it's not much get a 1950 at what will be bargain pricing or wait it out and get a 3950.

 

[kawi6rr]

I was just looking at the 1950 on NewEgg and the price is tempting but I really have no need for a TR but it does get me thinking about it.

 

[Markfw]

I see that microcenter has the 1950x for $599,99, but in store only. But I do not see the new MSRP anywhere to be found. Not at newegg or Amazon for sure.

 

[Justinbaileyman]

Because I bet they know they will sell like hot cakes soon as they drop the prices. So I bet they are trying to sit on them as long as possible to make as much moola as possible. Both Newegg and Amazon love to price gouge the crap out of us thats for sure!!

 

[SPBHM]

the platform cost is still way higher than the AM4 cost, considering entry price for motherboard and ram cost, so even if the CPU cost the same it's not a clear win, unless you really have some heavy work for those cores.

 

[EXCellR8]

Getting there... 1950X for around $500 sounds better. Maybe by the time Castle Peak shows up.

 

[Shmee]

Now that there are additional TR options, and prices are coming down, which chip would be best for gaming, if any?

 

[ZGR]

1950X or 2950X imo. But the 1900X is fine for gaming too, but I wouldn't do 8 cores on TR.

 

[Markfw]

1950X or 2950X imo. But the 1900X is fine for gaming too, but I wouldn't do 8 cores on TR.

Where the 8 core TR comes in, is for the quad channel memory, and the highest possible OC, which would be great for gaming. Not sure it could beat the 9900k when it comes out though, but then you would have the platform for future TR's that might beat it,

 

[lightmanek]

It's here!

Soon I will join Threadripper owners club, just need half a day of time to put this together!
Couldn't resist this new price! Only £348 for TR1920X To offset my happines, memory was expensive, at £455 for 4x8GB Tidents 3600 CL16.

PS. I also bought Enermax 360 AIO ... anyone know how to differenciate REV 1 from REV 2? If I got old one I will have to RMA it at some point in near future

 

[krumme]

Cool. You can always upgrade that light websurfing machine to a lean 32c machine if you need a proper office rig.
Not sure it will be 348 pounds though...

 

[cbn]

I'm looking at an X399 ASRock Fatality Pro, and a $399 TR1 1920X CPU. I already have four RX 570 GPUs to go in it. Then, I just need a good case. Thinking of getting another Rosewill Blackhawk.

I know you mentioned DC and Mining, but that kind of set-up would make for a very good Blender Workstation also (though technically the high number of PCIe lanes of X399 platform aren't really needed for GPU rendering).

P.S. I find it very interesting that Blender is being used for the creation of machine learning synthetic data sets.

https://ai.intel.com/fake-it-till-y...machine-learning-in-data-scarce-environments/

Image classification is one of the most common uses of neural networks today. By training deep neural networks on thousands of 2-dimensional images that can be easily found online, researchers and companies alike have been able to build strikingly accurate image classifiers. One remaining challenge is to create object identifiers that are both scale and rotation invariant – meaning that they are able to identify an object regardless of its size and position relative to the camera sensor.

By training a system on 3-dimensional objects could potentially yield a very accurate object identification algorithm, but the process of creating large-scale 3D datasets can be prohibitively time consuming. Capturing every conceivable angle and position of say, a sofa in a living room could take dozens of hours – and that’s merely one of the many objects that one may want to build up a dataset for. To solve this problem, Intel researchers have again turned to synthetic datasets.

To solve this problem using synthetic data, Intel researchers recreated 3D environments of rooms in various sizes and dimensions, even going to far as to realistically lighting and texturing the scenes using the popular 3D development tool Blender* 3D. Once base models were created, our researchers could generate thousands of image pairs (two sets of images of the exact same scene, but with the virtual cameras spaced a few inches apart). With this flexible model, it is possible to created image pairs from virtually any angle, lighting condition and baseline (the space between the two cameras).

(So application for Blender increasing. This in addition to its future potential as an engineering tool.)