I was posting some of this in another thread, but realized that it would probably make more sense on its own.
I have a 16" MBP M1 Max of my own, and happen to have access to a brand-new Dell Precision 5560 mobile workstation that is pretty darned close to the closest Dell offers in its lineup today, so I thought it would be fun to run some comparative benchmarks to see how a top-of-the-line laptop workstation from Apple compares to a pretty-much-equivalent one from Dell.
Specs:
MacBook Pro 2021 (18,2):
Dell Precision 5560:
The physical and cost breakdown are pretty darned close. Note that these are actual measurements with a ruler and kitchen scale, not from the spec sheet. Thickness is measured from the surface of a table it's sitting on to the top of the case. I did this because the weights were imprecise and Dell fudges the thickness; the computer is kind of a trapezoid shape with curved edges on the underside, and their spec sheet shows the thickness at the front and back, both of which are much less than the thickest point/"on-desk thickness".
Subjectively, the curved edges on the Dell make it feel thinner than the chunky MBP, and the volume is certainly lower. They both feel good, although I much prefer the aluminum palm rests to the Dell's carbon-fiber-looking slightly matte plastic. Closed, the Dell looks amusingly identical to the MBP from the top, although the bottom is much different (covered with vents and looks like a typical Dell underside). The MBP's monitor feels larger and looks better (color and brightness), although it's a hair lower-res (and not a touchscreen).
The Dell does have a dedicated Delete key, which is a huge plus to me. The keyboard otherwise looks suspiciously like an Apple one, except no fingerprint scanner, half-height and reduced-width F-keys, and full sized left and right arrow keys instead of T-shape. It also feels worse to me, as does the not-quite-as-large trackpad.
The Dell is currently more expensive, but my org (thanks to hefty enterprise discounts) payed around what I paid for the MBP with edu discount.
Both computers are specced at basically top-of-line apart from storage; max storage is the same on both. The Dell could be configured with a Xeon CPU (similar performance to the i9) and ECC RAM, but neither would impact performance significantly and there's a relatively modest price differential.
Note that the Dell 7560 has much higher-performance GPU options, and can take more RAM and storage (also it can cost way over $10,000), but I'm making a distinction between "laptop" workstations and "portable" workstations in that the Dell 7560 is way heavier, way thicker, and based on TDP has something like 30 minutes of battery life if the CPU and GPU are running full-bore, and that's if the battery can even supply the GPU running flat out. The 7000 series is moveable but clearly designed to be used plugged in and stationary; the 5000 series and MBP are genuine laptops.
That said, the MBP can comfortably be used on a lap even with the CPU running full-bore for several minutes straight; the Dell gets too hot to hold your hand on under the CPU after a few minutes of heavy CPU use. In ballpark estimates, heavy-number-crunching will drain the MBP battery in about 120 minutes versus 75 minutes for the Dell, although I haven't done precise comparisons. The Dell's fan is impressively quiet, but still much louder than the MBP's, which is virtually silent under almost any workload.
I started out with Geekbench, Cinebench, and GFXBench. Since 3DMark Wild Life and Wild Life Extreme can be run on an M1 Mac, I'd love to compare, but I'm not going to pay $30 for the Windows version to run the test there.
A major caveat to the results on the Dell: It is very clearly thermally limited on the CPU after the first minute or two of heavy load. If I run a 1-minute Cinebench test, I get around 11,800; on a 10-minute test, the results are erratic and between 9000 and 9800 depending on how it's positioned to maximize airflow. The MBP has no such limitations in my testing, at least at room temperature, and could comfortably be kept on a lap during a 10-minute Cinebench run without impacting performance.
Geekbench on the Dell didn't seem to be impacted by the thermal limits, since it's a short test. GFXBench probably is; I put the computer on its side to maximize airflow (way better than you'd get with it on a table).
Geekbench:
Blackmagic Disk Speed:
Cinebench r23:
GFXBench:
So, depending in part on how severely the Dell is throttling, the CPU in these tests is between 10% and 30% slower, roughly speaking. The M1 Max GPU in its worst showing is 10% faster than the Dell, and on many straightforward game-rendering type tests it's a full 3 times faster, which is pretty impressive.
I believe based on some preliminary tests that for Blender GPU rendering, even with the beta version that supports Metal the NVIDIA has a significant advantage (unfortunately the simple Blender Benchmark app doesn't yet have a version that supports GPU rendering). Optimization may be lagging due to it being so new, but in any case that's the one test so far where the AS GPU hasn't topped the Dell offering.
Any other cross-platform benchmark requests (with free software, I'm not actually going to spend money on this)?
I have a 16" MBP M1 Max of my own, and happen to have access to a brand-new Dell Precision 5560 mobile workstation that is pretty darned close to the closest Dell offers in its lineup today, so I thought it would be fun to run some comparative benchmarks to see how a top-of-the-line laptop workstation from Apple compares to a pretty-much-equivalent one from Dell.
Specs:
MacBook Pro 2021 (18,2):
M1 Max, 10-core
32-core GPU
64 GB memory
1TB storage
12.2.1
Dell Precision 5560:
Intel Core i9-11950H, 8-core
64 GB RAM (2x 32GB sticks)
NVIDIA RTX A2000 Laptop w/4GB
1TB storage
Win10 Enterprise freshly installed, no bloatware
The physical and cost breakdown are pretty darned close. Note that these are actual measurements with a ruler and kitchen scale, not from the spec sheet. Thickness is measured from the surface of a table it's sitting on to the top of the case. I did this because the weights were imprecise and Dell fudges the thickness; the computer is kind of a trapezoid shape with curved edges on the underside, and their spec sheet shows the thickness at the front and back, both of which are much less than the thickest point/"on-desk thickness".
| MacBook Pro 2021 | Dell Precision 5560 |
| 16" 3456x2234 (minus notch) | 15.6" 3840x2400 touchscreen |
| 3x TB4 ports | 3x TB4 ports |
| Magsafe | no dedicated charge port (must use TB4 port) |
| 1x HDMI | none (includes HDMI + USB-A dongle) |
| SD slot | SD slot |
| 18.5mm x 356mm x 248mm, 2158g | 19mm x 344mm x 232mm, 2011g |
| $3899 list, as-configured | $4216 as-configured ($6023 list) |
Subjectively, the curved edges on the Dell make it feel thinner than the chunky MBP, and the volume is certainly lower. They both feel good, although I much prefer the aluminum palm rests to the Dell's carbon-fiber-looking slightly matte plastic. Closed, the Dell looks amusingly identical to the MBP from the top, although the bottom is much different (covered with vents and looks like a typical Dell underside). The MBP's monitor feels larger and looks better (color and brightness), although it's a hair lower-res (and not a touchscreen).
The Dell does have a dedicated Delete key, which is a huge plus to me. The keyboard otherwise looks suspiciously like an Apple one, except no fingerprint scanner, half-height and reduced-width F-keys, and full sized left and right arrow keys instead of T-shape. It also feels worse to me, as does the not-quite-as-large trackpad.
The Dell is currently more expensive, but my org (thanks to hefty enterprise discounts) payed around what I paid for the MBP with edu discount.
Both computers are specced at basically top-of-line apart from storage; max storage is the same on both. The Dell could be configured with a Xeon CPU (similar performance to the i9) and ECC RAM, but neither would impact performance significantly and there's a relatively modest price differential.
Note that the Dell 7560 has much higher-performance GPU options, and can take more RAM and storage (also it can cost way over $10,000), but I'm making a distinction between "laptop" workstations and "portable" workstations in that the Dell 7560 is way heavier, way thicker, and based on TDP has something like 30 minutes of battery life if the CPU and GPU are running full-bore, and that's if the battery can even supply the GPU running flat out. The 7000 series is moveable but clearly designed to be used plugged in and stationary; the 5000 series and MBP are genuine laptops.
That said, the MBP can comfortably be used on a lap even with the CPU running full-bore for several minutes straight; the Dell gets too hot to hold your hand on under the CPU after a few minutes of heavy CPU use. In ballpark estimates, heavy-number-crunching will drain the MBP battery in about 120 minutes versus 75 minutes for the Dell, although I haven't done precise comparisons. The Dell's fan is impressively quiet, but still much louder than the MBP's, which is virtually silent under almost any workload.
I started out with Geekbench, Cinebench, and GFXBench. Since 3DMark Wild Life and Wild Life Extreme can be run on an M1 Mac, I'd love to compare, but I'm not going to pay $30 for the Windows version to run the test there.
A major caveat to the results on the Dell: It is very clearly thermally limited on the CPU after the first minute or two of heavy load. If I run a 1-minute Cinebench test, I get around 11,800; on a 10-minute test, the results are erratic and between 9000 and 9800 depending on how it's positioned to maximize airflow. The MBP has no such limitations in my testing, at least at room temperature, and could comfortably be kept on a lap during a 10-minute Cinebench run without impacting performance.
Geekbench on the Dell didn't seem to be impacted by the thermal limits, since it's a short test. GFXBench probably is; I put the computer on its side to maximize airflow (way better than you'd get with it on a table).
Geekbench:
| MBP M1 Max | Dell Precision 5560 | MBP Advantage | |
| Single-core | 1,781 | 1,624 | +10% |
| Multi-core | 12,597 | 8,951 | +41% |
| OpenCL Compute | 64,473 | 59,345 | +9% |
| Metal Compute | 66,624 | n/a |
Blackmagic Disk Speed:
| MBP M1 Max | Dell Precision 5560 | MBP Advantage | |
| Peak write | ~5350MB/s | ~2403MB/s | +122% |
| Peak read | ~5350MB/s | ~1640MB/s | +226% |
Cinebench r23:
| MBP M1 Max | Dell Precision 5560 | MBP Advantage | |
| multi-core (10-minute run) | 14,061 | 9,855 | +43% |
GFXBench:
| MBP M1 Max (Metal frames) | Dell Precision 5560 (OpenGL frames) | MBP Advantage | |
| Aztec Ruins 1440p High-Tier Offscreen | 19,900 | 5,578 | +257% |
| Car Chase 1080p Offscreen | 23,709 | 12,706 | +157% |
| Manhattan 3.1.1 1440p Offscreen | 24,587 | 18,974 | +30% |
| T-Rex 1080p Offscreen | 121,240 | 41,953 | +189% |
| ALU 2 1080p Offscreen | 68,548 | 60,331 | +14% |
| Driver Overhead 2 1080p Offscreen | 23,299 | 13,224 | +76% |
| Texturing 1080p Offscreen | 279,141 | 66,805 | +318% |
So, depending in part on how severely the Dell is throttling, the CPU in these tests is between 10% and 30% slower, roughly speaking. The M1 Max GPU in its worst showing is 10% faster than the Dell, and on many straightforward game-rendering type tests it's a full 3 times faster, which is pretty impressive.
I believe based on some preliminary tests that for Blender GPU rendering, even with the beta version that supports Metal the NVIDIA has a significant advantage (unfortunately the simple Blender Benchmark app doesn't yet have a version that supports GPU rendering). Optimization may be lagging due to it being so new, but in any case that's the one test so far where the AS GPU hasn't topped the Dell offering.
Any other cross-platform benchmark requests (with free software, I'm not actually going to spend money on this)?
Last edited:
