GTX 880 and 800 Series to be More Powerful But Cheaper than the 700 Series

Status
Not open for further replies.
BroHamBone said:
Nvidias Black Friday deals are for lower end items usually. Last year I think it was thelower end 600 series, 660 ti /760/770 for BF sale...and tegra and shield. Te discounts were probably 10-30 $ off if that

Maybe a good deal on the 750/750ti.

http://slickdeals.net/f/6489494-the...er-monday-video-card-thread-updated-regularly

Oh $10 off a 780.
Click to expand...

I didn't say Black Friday would bring deals on an 800 series of video cards.

I said that I think (and hope) that Nvidia will release the new series in October so that they'll be available during the shopping season. Big difference.

Seriously hoping bc I sold my GTX 780 here on the forums a month or so ago and I'm now using integrated graphics until the new series comes out. I think I can hang in there until October but I can't make it to 2015, no way.
 
tehgray said:
So if the card comes out in October, how long until the sought after Overclocked/quiet version cards come out? I would assume it isn't advised to buy the reference cards right out of the gate.
Click to expand...

I'm fine with the titan cooler. Was really impressed with the titan cooler on my GTX 780. Air flowing out the back works better with my Corsair Air 540 case and H100i cooling anyway.
 
GoldenTiger said:
Nvidia's reference coolers have been great.... I tried an r9 290 reference recently though and had previously owned a 7970, and both were obnoxious in tone, whine, and loudness. In short, nvidia reference = fine, amd reference = oh god no!

:)
Click to expand...

+1, agree
 
mls1995 said:
I didn't say Black Friday would bring deals on an 800 series of video cards.

I said that I think (and hope) that Nvidia will release the new series in October so that they'll be available during the shopping season.
Click to expand...

...yea, I don't think I did either..... :eek:
 
First we gotta keep in mind WCCF Tech is a pretty unreliable source.

But I am expecting based on the Maxwell benchmarks so far that we will see a decent gain with the next generation of GPU.

The cheaper part though ... we need a powerful AMD at both GPUs and CPUs or else we get gouged like hell. The Titan is a testament to this as was the 8800 Ultra. We need AMD to be competitive.
 
LordEC911 said:
What fab is it going to built at? GF?
TSMC's 20nm planar process isn't for high performance parts.
Click to expand...

Source? Because Tsmc is touting 20nm production in earnest now. Guessing source is your speculation, aka none, as I have seen no current info about your claim.
 
Michaelius said:
Seems like nvidia has no need to hurry. New amd cards will be higher clocked Tahiti parts

http://videocardz.com/51021/amd-gcn-update-iceland-tonga-hawaii-xtx

And tonga is tahiti with 256 bit bus not maxwell equvalent.
Click to expand...



I'm starting to think nobody really gets excited about what AMD is doing because few people understand their ridiculous codenames. Lets codename our next generation, then codename each product line, and then codename each product in that line. Maybe it's just me, but they are a little ridiculous where I can't even follow them anymore.

Who misses the old school Rxxx and NVxx method? o/
 
mls1995 said:
Seriously hoping bc I sold my GTX 780 here on the forums a month or so ago and I'm now using integrated graphics until the new series comes out. I think I can hang in there until October but I can't make it to 2015, no way.
Click to expand...

She's treating me well :D
 
Michaelius said:
Seems like nvidia has no need to hurry. New amd cards will be higher clocked Tahiti parts

http://videocardz.com/51021/amd-gcn-update-iceland-tonga-hawaii-xtx

And tonga is tahiti with 256 bit bus not maxwell equvalent.
Click to expand...
I thought they were saying Tonga was supposed to be something totally new, to compete with Maxwell @ mid-range. Now they're calling it a Tahiti refresh. And the XTX is a Hawaii refresh.
...At best.

My next card will be a GTX 880 / Ti but honestly I was expecting that for a long time now. How many Maxwell leaks have we had this year, compared to AMD leaks? The lack of tantalizing information from AMD certainly does not bode well.

The only thing I can guess is that AMD is expecting their refreshed parts to compete with Maxwell especially if Nvidia is really launching an 860, 870, 880, and 880 Ti.
I hope they know what they're doing. It sounds bad on paper.
 
I guess Nvidia was working on Maxwell while AMD was busy with Hawaii and they're not ready to move on yet.
 
DASHlT said:
Funny thats the same nvidia said for the 400/500 series.
Click to expand...
Point was AMD tends towards that configuration far more often. Every single one of AMD's current high-end parts lose pretty spectacularly when graphed based on performance-per-watt.
 
Unknown-One said:
Point was AMD tends towards that configuration far more often. Every single one of AMD's current high-end parts lose pretty spectacularly when graphed based on performance-per-watt.
Click to expand...
If I had to pick one feature to sacrifice in order to save $150 over Nvidia's offering, it would be power.
Also, drivers apparently.

Last rumors seemed to indicate AMD was pushing to be more competitive with Nvidia on power consumption, but... I doubt AMD can do anything with Tahiti/Hawaii to compete with Maxwell.
 
TaintedSquirrel said:
If I had to pick one feature to sacrifice in order to save $150 over Nvidia's offering, it would be power.
Click to expand...
And that would be the last thing I'd cut. More power = more heat = more noise trying to keep the darn thing cool. One of the primary complaints with AMD's latest cards is due to noise, which is directly related to efficiency.

The one thing I CANNOT stand is a noisy system. Cannot tell you the number of time's I've RMAed components due to noise issues.

TaintedSquirrel said:
Also, drivers apparently.
Click to expand...
Yeah... not a fan of AMD's drivers either. That's a topic for another thread, though.
 
Last edited:
Unknown-One said:
And that would be the last thing I'd cut. More power = more heat = more noise trying to keep the darn thing cool.
Click to expand...
The Tri-X is better than anything Nvidia's partners offer and has been as low as $350 on the 290, $500 on the 290x.
Heat and noise are easily dealt with especially when AMD is working with such a sizable price gap at the moment.

I value noise more than anything in my rig. If the new Hawaii card(s) can compete with high-end Maxwell cards, slap a Tri-X on there, and be $150 cheaper (again), I'll buy AMD in a heartbeat. Couldn't care less about power.
Quieter, cooler, and cheaper? Yes please.
 
TaintedSquirrel said:
The Tri-X is better than anything Nvidia's partners offer and has been as low as $350 on the 290, $500 on the 290x.
Click to expand...
Yeah, dealt with the likes of the Tri-X before. It's most certainly not as good as the Nvidia reference GTX780/Titan cooler for noise levels or your system as a whole.

Doesn't evacuate heat from the case, which means in a normal PC case, you now have to spin up all your other fans to keep temps in-check. Every time I've tried a cooler like this, it leads to an overall WORSE noise profile than just using the Nvidia reference blower + low fan speeds everywhere else.

Only two solutions that seem to be viable for low noise levels inside a PC case are blower coolers and AIO water cooling. I'm currently attempting the latter, but as I feared, pump-noise is once-again making the reference blower the superior option...
 
Michaelius said:
Seems like nvidia has no need to hurry. New amd cards will be higher clocked Tahiti parts

http://videocardz.com/51021/amd-gcn-update-iceland-tonga-hawaii-xtx

And tonga is tahiti with 256 bit bus not maxwell equvalent.
Click to expand...
Let's just say that WhyCry has no idea what is going on in either camp.
That is about as nice as I can put it.

GoldenTiger said:
Source? Because Tsmc is touting 20nm production in earnest now. Guessing source is your speculation, aka none, as I have seen no current info about your claim.
Click to expand...
Source? Look at their roadmaps for 20nm.
They cancelled 20G, and moved 20SOC up to the same spot as 20HPM, on paper.
Any 20nm designs that were in the works by AMD or Nvidia would be much better off waiting 6 months for 16FinFet.
This was discussed over a year ago along with capacity issues... I guess people have short-term memories.

Liger88 said:
I'm starting to think nobody really gets excited about what AMD is doing because few people understand their ridiculous codenames. Lets codename our next generation, then codename each product line, and then codename each product in that line. Maybe it's just me, but they are a little ridiculous where I can't even follow them anymore.

Who misses the old school Rxxx and NVxx method? o/
Click to expand...
That's the point. Hard to catch your opponent off guard when they know exactly what is coming +6months in advance.

TaintedSquirrel said:
I guess Nvidia was working on Maxwell while AMD was busy with Hawaii and they're not ready to move on yet.
Click to expand...
Yeah... because there are no other reasons as to why nothing else has been released yet or will be released in the near future. :rolleyes:
 
CUDA version 6.5(which adds ARMv8 support to original CUDA 6.0) was added to Nvidia Quadro by the driver on July 18th.

It means that ARM cpu has a driver implementation already. Similar to when Geforce driver included gm200,ect, codenames a few months ago.

Now, Maxwell will support cuda ver.6.5 for arm cpus.
 
SinOfLiberty said:
CUDA version 6.5(which adds ARMv8 support to original CUDA 6.0) was added to Nvidia Quadro by the driver on July 18th.

It means that ARM cpu has a driver implementation already. Similar to when Geforce driver included gm200,ect, codenames a few months ago.

Now, Maxwell will support cuda ver.6.5 for arm cpus.
Click to expand...

Wouldn't this just be for the Tegra K1 (the dual-core ARMv8 one)?
 
SinOfLiberty said:
CUDA version 6.5(which adds ARMv8 support to original CUDA 6.0) was added to Nvidia Quadro by the driver on July 18th.

It means that ARM cpu has a driver implementation already. Similar to when Geforce driver included gm200,ect, codenames a few months ago.

Now, Maxwell will support cuda ver.6.5 for arm cpus.
Click to expand...

Have you figured out what single precision is used for yet?
 
SinOfLiberty said:
GTC JAPAN 2014!!
Click to expand...

NVIDIA at 2014 GPU Technology Conference Japan (2014/7/16)..

QqVt9IC.jpg

"Donut that is now available from the venue two years ago has increased the number with each passing year" (translated, think they're talking about CUDA power! ..or maybe GTCJ attendance figures :D)

http://pc.watch.impress.co.jp/docs/news/20140716_658255.html


Summary: CUDA 6.5 w/ARM support, Tesla HPC w/64bit ARM cores, Android devices w/64bit Tegra K1, GRID virtualization, more automotive crap :rolleyes:

Maxwell (DX12) still slated for 2014, and Pascal in 2016. No 800-series announcement (as SoL later mentioned).
 
Hehe.

1)arm support is for compute purposes, does not need GeForce driver yet.

2) GTC Japan. Most importantly, NV has shipped its Kepler SOC! As was shown on the pc watch diagram. Now, with SOC and mobile shipped, next up is a desktop launch! As this time NV launches in power efficient areas first..

3)TSMC has released its June reports on the same day as gtc japan! and guess what! TSMC has started volume shipment of 20nm Wafers in JUNE! The same time(June 11) when 1st GM200 LoadBoard(ES) has been built and shipped for testing!! Followed by multiple integrated circuits!

http://seekingalpha.com/article/231...-results-earnings-call-transcript?part=single

After two years of meticulous preparation, we began volume shipments of our 20 nanometer wafers in June. The steepness of our 20 nanometer ramp sets a record. We expect 20 nanometer to generate about 10% of our wafer revenue in the third quarter and more than 20% of our wafer revenue in the fourth quarter and we expect the demand for 20 nanometer will remain strong and will continue to contribute more than 20% of our wafer revenue in 2015.
Click to expand...

4) I supposed GTC Japan would mention some of what is stated above. And I think they did; behind closed doors, similar to Computex.
 
October/November works for me, although sooner would be nice :)
I definitely need something this year and if I buy the current top end stuff I know I'll regret it when these arrive.
 
Not that I'm surprised, I will just end up passing this gen if it's still 28nm. Power consumption should be intriguing, if the flagship cards show similar performance per watt as the 750ti.
 
Status
Not open for further replies.
Back
Top