FSP Hydro PTM Pro (1200W PSU) review
ASUS ROG Radeon RX 6750 XT STRIX review
AMD FidelityFX Super Resolution 2.0 - preview
Sapphire Radeon RX 6650 XT Nitro+ review
Sapphire Radeon RX 6950 XT Sapphire Nitro+ Pure review
Sapphire Radeon RX 6750 XT Nitro+ review
MSI Radeon RX 6950 XT Gaming X TRIO review
MSI Radeon RX 6750 XT Gaming X TRIO review
MSI Radeon RX 6650 XT Gaming X review
Deepcool AS500 PLUS CPU Cooler Review
Review: Core i9 12900K DDR4 3600 vs DDR5 5200 performance
Recently Intel released Alder Lake-based processor, the Core i9 12900K, and the platform is poised for the future with its incredibly fast performance paired with DDR5 memory, however works with DDR4 as well, really well actually.
Read the review right here.
Review: Corsair H150i Elite LCD - 11/23/2021 11:01 AM
Corsair has a refresh of liquid coolers, this round the Corsair H150i Elite LCD (IPS) edition is put on our test bench where we can check out the behavior of the triple-fan 360mm model. Corsair has pa...
Review: Fractal Design Ion+ 2 Platinum 860W PSU - 11/16/2021 04:47 PM
Highly flexible modular cables and a product that looks fantastic, meet the Fractal Design Ion+ 2 Platinum 870 which provides excellent performance, silent operation, ease of installation, and all at ...
Review: Crucial P5 Plus PCIe 4.0 M.2 NVMe - 11/12/2021 11:35 AM
The Crucial P5 has proven to be popular because to its value and performance, and the P5 Plus appears to be aiming to occupy a similar niche. We take a look at the 1TB model. Read the review here.
Review: Fractal Design Lumen S28 RGB (LCS) - 11/11/2021 01:01 PM
It's been a while since we've put a Fractal Design cooling unit through its paces. The most recent items, on the other hand, feature enhanced cooling, a pump that is integrated into the radiator, an...
Review: WD Red SN500 NVMe SSD - 11/09/2021 03:01 PM
WD's latest NVMe SSD is intended to broaden out the company's product lineup, but it also starts to end their SATA solid-state drives (SSDs). Recent releases from WD include the all-new WD Red SN500...
ruthan
Senior Member
Posts: 545
Joined: 2016-05-24
Senior Member
Posts: 545
Joined: 2016-05-24
#5968032 Posted on: 11/26/2021 10:57 AM
Minimum latency should be part of DDRX specs, because its matters so much and transfer speeds, we so similar thing with first Ryzen CPU design, which also had some latency issue between their blocks.. otherwise its other way how to rip off normal not too much informed customers, because they would thing than bigger speed number would mean much superior product.
Otherwise it would be nice to test more memory speeds and add some timing info into charts.
And yeah Quad channel should be finally available on normal platforms not just some premium workstations. I hoped a lot that AMD will use it to make some advantage over Intel, to force Intel to enable the same.
Minimum latency should be part of DDRX specs, because its matters so much and transfer speeds, we so similar thing with first Ryzen CPU design, which also had some latency issue between their blocks.. otherwise its other way how to rip off normal not too much informed customers, because they would thing than bigger speed number would mean much superior product.
Otherwise it would be nice to test more memory speeds and add some timing info into charts.
And yeah Quad channel should be finally available on normal platforms not just some premium workstations. I hoped a lot that AMD will use it to make some advantage over Intel, to force Intel to enable the same.
nizzen
Senior Member
Posts: 1891
Joined: 2005-08-05
Senior Member
Posts: 1891
Joined: 2005-08-05
#5968039 Posted on: 11/26/2021 11:42 AM
Minimum latency should be part of DDRX specs, because its matters so much and transfer speeds, we so similar thing with first Ryzen CPU design, which also had some latency issue between their blocks.. otherwise its other way how to rip off normal not too much informed customers, because they would thing than bigger speed number would mean much superior product.
Otherwise it would be nice to test more memory speeds and add some timing info into charts.
And yeah Quad channel should be finally available on normal platforms not just some premium workstations. I hoped a lot that AMD will use it to make some advantage over Intel, to force Intel to enable the same.
Don't be blind on the Aida 64 numbers. Performance isn't all about low latency in Aida64. Just look at Zen 3 vs Intel 8,9,10,11 and 12. Zen 3 has almost the same gamingperformance as Intel, with 20ns higher. Big cache is helping Zen3 here. More on "zen3d" will help even more. Low latency is king, so is cache
Minimum latency should be part of DDRX specs, because its matters so much and transfer speeds, we so similar thing with first Ryzen CPU design, which also had some latency issue between their blocks.. otherwise its other way how to rip off normal not too much informed customers, because they would thing than bigger speed number would mean much superior product.
Otherwise it would be nice to test more memory speeds and add some timing info into charts.
And yeah Quad channel should be finally available on normal platforms not just some premium workstations. I hoped a lot that AMD will use it to make some advantage over Intel, to force Intel to enable the same.
Don't be blind on the Aida 64 numbers. Performance isn't all about low latency in Aida64. Just look at Zen 3 vs Intel 8,9,10,11 and 12. Zen 3 has almost the same gamingperformance as Intel, with 20ns higher. Big cache is helping Zen3 here. More on "zen3d" will help even more. Low latency is king, so is cache
fellix
Senior Member
Posts: 206
Joined: 2005-08-31
Senior Member
Posts: 206
Joined: 2005-08-31
#5968056 Posted on: 11/26/2021 01:44 PM
Correct. Those synthetic memory tests measure a very narrow scope of the overall system characteristics and doesn't hint at how those numbers factor in the complete performance pipeline, particularly when comparing two different CPU architectures from different hardware vendors. In the case of AIDA64, the memory benchmark uses the most suitable routine for each CPU model, incl. hand-tuned SIMD memory instructions that are rarely employed in a production code. Unless there's something misconfigured in the system, resulting in a dramatic score reduction in this benchmark, a difference of few cycles/nanoseconds is just a wash. Don't panic over it.
Don't be blind on the Aida 64 numbers. Performance isn't all about low latency in Aida64. Just look at Zen 3 vs Intel 8,9,10,11 and 12. Zen 3 has almost the same gamingperformance as Intel, with 20ns higher. Big cache is helping Zen3 here. More on "zen3d" will help even more. Low latency is king, so is cache
Correct. Those synthetic memory tests measure a very narrow scope of the overall system characteristics and doesn't hint at how those numbers factor in the complete performance pipeline, particularly when comparing two different CPU architectures from different hardware vendors. In the case of AIDA64, the memory benchmark uses the most suitable routine for each CPU model, incl. hand-tuned SIMD memory instructions that are rarely employed in a production code. Unless there's something misconfigured in the system, resulting in a dramatic score reduction in this benchmark, a difference of few cycles/nanoseconds is just a wash. Don't panic over it.
kapu
Senior Member
Posts: 5288
Joined: 2006-12-22
Senior Member
Posts: 5288
Joined: 2006-12-22
#5968060 Posted on: 11/26/2021 02:15 PM
Funny, my ancient cpu with ddr3 memory has nearly identical ram performance to the 12900k with ddr4
It only makes sense if You us common divider But impressive results none the less . Kinda funny when DDR3 can catch DDR5
Funny, my ancient cpu with ddr3 memory has nearly identical ram performance to the 12900k with ddr4
It only makes sense if You us common divider
But impressive results none the less . Kinda funny when DDR3 can catch DDR5
Click here to post a comment for this news story on the message forum.
Senior Member
Posts: 3897
Joined: 2017-11-23
Obviously, but the result is still the same