- Joined
- Aug 22, 2010
- Messages
- 34,060
- Reaction score
- 12,912
So here come some rumors about AMD video card plans. I am mildly curious about that Navi 21. Ray tracing and 16 GB memory.
i cant even read this with out getting sexually Excited
Tech Gaming Hardware discussion (& Hardware Sales) thread
- Thread starter Madmick
- Start date
i cant even read this with out getting sexually Excited
- Joined
- Jun 13, 2005
- Messages
- 66,596
- Reaction score
- 38,494
- Joined
- Jul 4, 2009
- Messages
- 69,682
- Reaction score
- 42,746
Ryzen 3000 series is coming out in roughly 2 weeks. They've got a nice alternative.
- Joined
- Jul 5, 2017
- Messages
- 13,139
- Reaction score
- 21,131
Downside to that is I already have a pimptastic Noctua Cooler for a LGA 1151 CPU, for AMD I'd have to get another cooler, which while I spend money like a fat kid eats cake, I can't justify buying another cooler less than a year after my other one hahaha
- Joined
- Jun 13, 2005
- Messages
- 66,596
- Reaction score
- 38,494
I love this. A peek behind the curtain. Honesty. Reality. The difference between build videos and expert builder builds IRL (it's a tangled mess, nobody cares, stop being so hard on yourself):
Now the big (and confusing) news. I skipped this headline three days ago via Tom's Hardware because I didn't pay attention to the named Ryzen processor. Everyone else is circulating it now:
AMD Ryzen 7 3800X Glides Past Intel Core i9-9900K In Leaked Geekbench Numbers
Alternative WCCF Tech coverage
The R7-3800X also appears to be cheating a bit, here. It's clocked +100MHz above the quoted stock frequency. Let's get that out of the way, first. Still, that's a tiny boost.
The reason I find this a bit confusing is because the R7-3800X is an 8-core/16-thread processor exactly the same as the i9-9900K. Normally, in the past, with early leaks for AMD processors, if comparing to Intel competitors in terms of cost, we have had to look at multicore scores for AMD CPUs with more cores or threads. The result is we have only a vague idea what single core per-thread performance is, and once a processor has four cores, even in 2019 for all but a handful of games, this will determine the best gaming processor. There's a reason Intel nerfs the frequencies on its quad core processors, lately. The reason is nobody would be willing to buy their hexacore and octacore processors if the quad core i3's matched them on frequency. Wasted cash.
But given identical cores and threads it doesn't at first make sense why the R7-3800X wins one, but not both. The extra cache might pay dividends in the multicore score, but my suspicion is the lower peak turbo boosts on a single core for the R7-3800X are the culprit.
i9-9900K
Ryzen 3000 Line-up
The i9-9900K can hit a 5.0 GHz peak turbo on a single core, and in fact will briefly overcharge to 5.1 GHz via "Turbo 3.0". Meanwhile, the R7-3800X will only hit 4.5 GHz on a single core.
The good news for Ryzen 3000 is that it obviously turbos very well at factory clocks across all cores; better than the Intel which has to drop to 4.7 GHz. The bad news is this won't help it win most game benchmarks.
Remember that previous leaks analyzed by Redditors suggested a 7% advantage to Ryzen 3000's IPC against Kaby Lake while the Coffee Lake Refresh holds a roughly 2.3% advantage over Kaby Lake in terms of IPC:
https://cpugrade.com/articles/cinebench-r15-ipc-comparison-graphs/
However, at least in the Cinebench charts above, which aren't normalized to a certain frequency, the R7-3800X is only beating its predecessor the R7-2700X by 11.2% despite that it has a 4.5 GHz vs. 4.3 GHz turbo advantage (+4.7% frequency advantage). So my skepticism about the separate 15% IPC gain claims over Ryzen 2000 appears well-founded. That's gotta be vapor.
The R7-3800X is the one losing by 15% to the 9900K in single core scores, here, and even the 4% higher turbo on the R9-3950X can't be expected to make that up.
*CLIFFS*
Intel Coffee Lake refresh chips will maintain real-world dominance as gaming CPUs even against Ryzen 3000. Now the question is how interested the market will be in better overall horsepower value that is theoretical. Greg Salazar nailed it at Computex. Ahead of the game. All the YouTube & Reddit commenters who hated on him can eat shit.
Now the big (and confusing) news. I skipped this headline three days ago via Tom's Hardware because I didn't pay attention to the named Ryzen processor. Everyone else is circulating it now:
AMD Ryzen 7 3800X Glides Past Intel Core i9-9900K In Leaked Geekbench Numbers
Alternative WCCF Tech coverage
The R7-3800X also appears to be cheating a bit, here. It's clocked +100MHz above the quoted stock frequency. Let's get that out of the way, first. Still, that's a tiny boost.
The reason I find this a bit confusing is because the R7-3800X is an 8-core/16-thread processor exactly the same as the i9-9900K. Normally, in the past, with early leaks for AMD processors, if comparing to Intel competitors in terms of cost, we have had to look at multicore scores for AMD CPUs with more cores or threads. The result is we have only a vague idea what single core per-thread performance is, and once a processor has four cores, even in 2019 for all but a handful of games, this will determine the best gaming processor. There's a reason Intel nerfs the frequencies on its quad core processors, lately. The reason is nobody would be willing to buy their hexacore and octacore processors if the quad core i3's matched them on frequency. Wasted cash.
But given identical cores and threads it doesn't at first make sense why the R7-3800X wins one, but not both. The extra cache might pay dividends in the multicore score, but my suspicion is the lower peak turbo boosts on a single core for the R7-3800X are the culprit.
i9-9900K
Ryzen 3000 Line-up
The i9-9900K can hit a 5.0 GHz peak turbo on a single core, and in fact will briefly overcharge to 5.1 GHz via "Turbo 3.0". Meanwhile, the R7-3800X will only hit 4.5 GHz on a single core.
The good news for Ryzen 3000 is that it obviously turbos very well at factory clocks across all cores; better than the Intel which has to drop to 4.7 GHz. The bad news is this won't help it win most game benchmarks.
Remember that previous leaks analyzed by Redditors suggested a 7% advantage to Ryzen 3000's IPC against Kaby Lake while the Coffee Lake Refresh holds a roughly 2.3% advantage over Kaby Lake in terms of IPC:
https://cpugrade.com/articles/cinebench-r15-ipc-comparison-graphs/
However, at least in the Cinebench charts above, which aren't normalized to a certain frequency, the R7-3800X is only beating its predecessor the R7-2700X by 11.2% despite that it has a 4.5 GHz vs. 4.3 GHz turbo advantage (+4.7% frequency advantage). So my skepticism about the separate 15% IPC gain claims over Ryzen 2000 appears well-founded. That's gotta be vapor.
The R7-3800X is the one losing by 15% to the 9900K in single core scores, here, and even the 4% higher turbo on the R9-3950X can't be expected to make that up.
*CLIFFS*
Intel Coffee Lake refresh chips will maintain real-world dominance as gaming CPUs even against Ryzen 3000. Now the question is how interested the market will be in better overall horsepower value that is theoretical. Greg Salazar nailed it at Computex. Ahead of the game. All the YouTube & Reddit commenters who hated on him can eat shit.
Last edited:
- Joined
- Jul 23, 2011
- Messages
- 34,956
- Reaction score
- 26,196
- Joined
- Jul 5, 2017
- Messages
- 13,139
- Reaction score
- 21,131
Depends if they honour that outside of the US
- Joined
- Jul 4, 2009
- Messages
- 69,682
- Reaction score
- 42,746
Sell it. The higher end Ryzens come with AMD's Wraith Prism cooler.
https://www.amd.com/en/technologies/cpu-cooler-solution
- Joined
- Jul 5, 2017
- Messages
- 13,139
- Reaction score
- 21,131
Sell it. The higher end Ryzens come with AMD's Wraith Prism cooler.
https://www.amd.com/en/technologies/cpu-cooler-solution
The Fanless Noctua was better than the Wraith with a fan... Where I live good cooling is a must (Central East Coast Australia) we get 3 months of cold weather and then it's back up to 30degree every day with 35degree room temps
- Joined
- Mar 17, 2007
- Messages
- 20,866
- Reaction score
- 12,670
Need help Sherbros.
I have a closed loop water cooling system. I've noticed my pump is losing some coolant. I'd say when I originally got it, the pump reservoir looked like it was 90% full. Now it's looking more like 60%. This is after I'd say about 2.5+ years of use (I got it I think in October 2016).
I don't think it's actually leaking anywhere, but if it is, it must be at minuscule levels.
Question is would you happen to know how empty the reservoir needs to get before it's an issue for the system? If I looked to fill it up myself, any advice on how to go about? What kind of coolant should I buy? Can I mix the current coolant (I don't know what it is) with something else? Anybody know where to buy and know how to actually do this? As a novice, am I maybe better off taking it to a pro?
I have a closed loop water cooling system. I've noticed my pump is losing some coolant. I'd say when I originally got it, the pump reservoir looked like it was 90% full. Now it's looking more like 60%. This is after I'd say about 2.5+ years of use (I got it I think in October 2016).
I don't think it's actually leaking anywhere, but if it is, it must be at minuscule levels.
Question is would you happen to know how empty the reservoir needs to get before it's an issue for the system? If I looked to fill it up myself, any advice on how to go about? What kind of coolant should I buy? Can I mix the current coolant (I don't know what it is) with something else? Anybody know where to buy and know how to actually do this? As a novice, am I maybe better off taking it to a pro?
- Joined
- Aug 6, 2004
- Messages
- 19,055
- Reaction score
- 9,809
more like... what possible reasons are there to buy a 9900k right now?
- Joined
- Jul 5, 2017
- Messages
- 13,139
- Reaction score
- 21,131
I do a lot of photoshop and rendering so it trumps my 7700k in that department. I do a lot of gaming and I can use my mobo and i7 to upgrade my girlfriends PC.
- Joined
- Dec 21, 2009
- Messages
- 21,063
- Reaction score
- 5,236
Ryzens come with good coolers. Won’t need to buy one
- Joined
- Aug 6, 2004
- Messages
- 19,055
- Reaction score
- 9,809
ok, but i mean... ryzen gen3 comes out on 7/7. and if you're getting a new mobo, anyway...
- Joined
- Jul 23, 2011
- Messages
- 34,956
- Reaction score
- 26,196
Quick Sync
- Joined
- Jul 5, 2017
- Messages
- 13,139
- Reaction score
- 21,131
Right now the 9900k is $200 off at a local store, Ryzen Gen 3 doesn't seem to offer anything that in real world usage actually beats the 9900k other than the fact that for once in their pathetic existence AMD is being competitive.
Though I may wait for the supposed Intel price cuts and see what price they end up at.
Last edited:
- Joined
- Jan 20, 2004
- Messages
- 34,822
- Reaction score
- 27,883
I love this. A peek behind the curtain. Honesty. Reality. The difference between build videos and expert builder builds IRL (it's a tangled mess, nobody cares, stop being so hard on yourself):
Now the big (and confusing) news. I skipped this headline three days ago via Tom's Hardware because I didn't pay attention to the named Ryzen processor. Everyone else is circulating it now:
AMD Ryzen 7 3800X Glides Past Intel Core i9-9900K In Leaked Geekbench Numbers
Alternative WCCF Tech coverage
The R7-3800X also appears to be cheating a bit, here. It's clocked +100MHz above the quoted stock frequency. Let's get that out of the way, first. Still, that's a tiny boost.
The reason I find this a bit confusing is because the R7-3800X is an 8-core/16-thread processor exactly the same as the i9-9900K. Normally, in the past, with early leaks for AMD processors, if comparing to Intel competitors in terms of cost, we have had to look at multicore scores for AMD CPUs with more cores or threads. The result is we have only a vague idea what single core per-thread performance is, and once a processor has four cores, even in 2019 for all but a handful of games, this will determine the best gaming processor. There's a reason Intel nerfs the frequencies on its quad core processors, lately. The reason is nobody would be willing to buy their hexacore and octacore processors if the quad core i3's matched them on frequency. Wasted cash.
But given identical cores and threads it doesn't at first make sense why the R7-3800X wins one, but not both. The extra cache might pay dividends in the multicore score, but my suspicion is the lower peak turbo boosts on a single core for the R7-3800X are the culprit.
i9-9900K
Ryzen 3000 Line-up
The i9-9900K can hit a 5.0 GHz peak turbo on a single core, and in fact will briefly overcharge to 5.1 GHz via "Turbo 3.0". Meanwhile, the R7-3800X will only hit 4.5 GHz on a single core.
The good news for Ryzen 3000 is that it obviously turbos very well at factory clocks across all cores; better than the Intel which has to drop to 4.7 GHz. The bad news is this won't help it win most game benchmarks.
Remember that previous leaks analyzed by Redditors suggested a 7% advantage to Ryzen 3000's IPC against Kaby Lake while the Coffee Lake Refresh holds a roughly 2.3% advantage over Kaby Lake in terms of IPC:
https://cpugrade.com/articles/cinebench-r15-ipc-comparison-graphs/
However, at least in the Geekbench charts above, which aren't normalized to a certain frequency, the R7-3800X is only beating the R7-2700X by 11.2% despite that it has a 4.5 GHz vs. 4.3 GHz turbo advantage. So my skepticism about the 15% IPC gain leaks/rumors/claims appears well-founded. That's gotta be vapor.
The Ryzen is the one losing by 15% in single core scores, here, and even that 4% higher turbo on the R9-3950X can't be expected to make that up.
CLIFFS:
Intel Coffee Lake refresh chips will maintain real-world dominance as gaming CPUs even against Ryzen 3000. Now the question is how interested the market will be in better overall horsepower value that is theoretical. Greg Salazar nailed it at Computex. Ahead of the game. All the commenters who hated on him can eat shit.
I hear from people who looked at the leaked data say these benchmarks are early non optimized OS an software tuned benchmarks. Microsoft plans on releasing new OS updates that is better optimized for IPC improvements an the other is the Ryzen supports faster DRAM an that could aid Ryzen. Microsoft has been dragging their feet releasing optimized OS updates but now it's happening. Still looks like a slight advantage for Intel but not for the prices they are charging.
- Joined
- Jun 13, 2005
- Messages
- 66,596
- Reaction score
- 38,494
I did notice the RAM was slower on the AMD side, but that isn't creating a 15% deficit. Besides, previous leaks from which those IPC estimates were also drawn via Geekbench, and would also have been using less optimized drivers.
Sure. Wait for Ice Lake. Intel already unveiled the low-powered mobile roadmap in May. This isn't an architectural refinement . This will be Intel's brand new architecture, and a response we should expect Intel is scrambling to prepare for Ryzen 3000 sometime this year. The 9900K released in October 2018, and the 8700K released October 2017.
This makes more sense for you because you've been patient so far considering your upgrade tendencies. You're still on the 7700K and Z170 Asus Sabretooth S, right? If so, you can't slot the 9900K into your motherboard in the first place. Incompatible. So you're going to have to buy a brand new motherboard. Might as well be whatever Ice Lake demands instead of a Z370 which is a glorified Z170 with vamped up VRMs, and a couple other upgrades. Like the X570 for Ryzen 3000 the new Intel motherboards will support PCIe 4.0.
So unless you have a broken game experience that is ravaging your day-to-day driving, I would hold off to see what they offer. Of course, this could mean you run into the same issue with your Noctua and the need for an adapter bracket. So if the sale is that killer, and you see a great price on a Z370 motherboard you want, the 9900K still might make sense.
Last edited:
